RÉDACTEUR EN CHEF

Michel Pugeat (Lyon)

RÉDACTRICE EN CHEF ASSOCIÉE
Marie-Christine Vantyghem (Lille)
COMITÉ DE RÉDACTION : Nadine Binart (Paris, Bicêtre), Jean-François Bonneville (Besançon), Nicolas Chevalier (Marseille),
Frédéric Castinetti (Marseille), Nathalie Chabbert-Buffet (Paris), Olivier Chabre (Grenoble), Sophie Christin-Maitre (Paris),
Thomas Cuny (Marseille), Christine Do Cao (Lille), Bernard Goichot (Strasbourg), Dominique Maiter (Bruxelles),
Hervé Monpeyssen (Neuilly-sur-Seine), Pascal Pigny (Lille), Gaetan Prévost (Rouen), Gérald Raverot (Lyon).
COMITÉ SCIENTIFIQUE : Xavier Bertagna (Paris), Albert Beckers (Liège), Christian Boitard (Paris), Thierry Brue (Marseille),
Philippe Chanson (Paris), Karine Clément (Paris), Philippe Bouchard (Paris), Didier Dewailly (Lille), André Lacroix (Montréal),
Hervé Lefebvre (Rouen), Yves Lebouc (Paris), Marc Lombes (Paris), Yves Morel (Lyon), Nelly Pitteloud (Lausanne), Vincent Rohmer
(Angers), Bernard Rousset (Lyon), Martin Schlumberger (Paris-Villejuif), Jean Louis Wemeau (Lille), Jacques Young (Paris).

Annales d’Endocrinologie (ISSN 0003-4266) 2020 (volume 81) Un an ; 6 numéros. Tarif institutionnel France (TTC) : 523 euros. Voir tarifs complets sur le site
www.elsevier-masson.fr/revue/180
Adresser commande et paiement à Elsevier Masson SAS, Service Abonnements, 65, rue Camille-Desmoulins, 92442 Issy-les-Moulineaux cedex : paiement
par chèque, carte de crédit (CB, MasterCard, EuroCard ou Visa : indiquer le n°, la date d’expiration de la carte, le cryptogramme et signer) ou par virement :
La Banque postale, Centre de Paris, n° RIB : 20041 00001 1904540H020 95.
Les abonnements sont mis en service dans un délai de 4 semaines après réception du paiement. Ils partent du premier numéro de l’année. Les numéros de l’année
et les volumes antérieurs doivent être commandés à l’éditeur. Les réclamations pour les numéros non reçus doivent parvenir dans un délai maximal de 6 mois
après la parution. Expédition par voie aérienne incluse.
Responsable de production éditoriale – Laure Etcheverry. E-mail : ANDO-gjp@elsevier.com
Publicité – Jean-Marie Pinson. Tél. : +33 (0) 1 71 16 53 10. E-mail : j.pinson@elsevier.com
Partenariats – Claire Ebersold. Tèl. : +33 (0)1 71 16 51 14. E-mail : c.ebersold@elsevier.com
Abonnements – Tél. : +33 (0)1 71 16 55 99. Site web : http://www.em-consulte.com/infos
Éditeur – Anne-Elisabeth Fournié. E-mail : a.fournie@elsevier.com
Directeur de la publication – Daniel Rodriguez
Les modalités d’abonnement, les sommaires de chaque numéro ainsi que les résumés des articles publiés dans cette revue sont disponibles sur le site internet
d’Elsevier Masson SAS : http://www.em-consulte.com

Imprimé en France par Jouve, 53101 Mayenne. Dépôt légal à parution

ISSN : 0003-4266
 Annales d’Endocrinologie 81 (2020) 473–475
Disponible en ligne sur
ScienceDirect
www.sciencedirect.com
Editorial
Patient requests for improved diagnosis and information in polycystic
ovary syndrome
Requête des patientes pour une amélioration du diagnostic et de l’information
pour le syndrome des ovaires polykystiques
1. English version
Polycystic ovary syndrome (PCOS) is, by definition, based on
a combination of clinical, biological and ultrasound criteria, the
individual variability and reversibility of which make for great con-
fusion for patients and healthcare workers. This confusion is all the
more harmful as PCOS affects 10% to 15% of women of child-bearing
age.
In this issue of the Annales d’Endocrinologie, M. Bouthier et al. [1],
a team of gynecologists from Marseille (France), report a qualitative
study of sharing of experience in seven French-speaking Internet
forums. In this qualitative study, the authors explore the emotional
feelings of patients in real life conditions, and identify the complex
issues of infertile women with PCOS using the forums.
A total of 785 comments from 211 infertile women diagnosed
with PCOS were analyzed. Their main complaints concerned: (i)
delay, with several consultations needed to diagnose PCOS; (ii) a
lack of clear explanatory information about PCOS; and (iii) treat-
ment difficulties, in particular, in controlling frequent overweight.
The authors testify to the negative psychological impact of PCOS
and infertility on daily life. They observe that this negative impact
seems to be heightened by the exchange of knowledge and experi-
ence made possible by online discussion forums.
Using the Internet as a source of research material raises new
issues, for which methodological and ethical recommendations are
still vague. This type of approach offers a snapshot of spontaneous
testimony, which, although not having the value of a prospective
study, was confirmed by recent articles which converge on the same
conclusions.
Gibson-Helm et al. [2] used an online questionnaire to assess
patients’ experience of their diagnosis and the information pro-
vided about PCOS. In this study, 1385 English-speaking women
from 32 countries were recruited from sites in English on-line
between 2015 and 2016. The authors found that more than a third
of respondents spent 2 years trying to find a diagnosis that could
explain their symptoms, and consulted at least 3 and up to 5 differ-
ent medical providers. Only 35% were satisfied with the diagnostic
process, and fewer than 20% considered the information provided
to be appropriate. Adequacy of information correlated positively
with satisfaction with the diagnosis; consultation of ≥ 5 practition-
https://doi.org/10.1016/j.ando.2020.10.001
0003-4266/© 2020 Published by Elsevier Masson SAS.
ers and a period of > 2 years before diagnosis correlated negatively,
independently of patient age and place of origin or residence. The
main difficulties experienced were absence of weight loss, irregular
menstrual cycle, and infertility.
Recently, a study compared monthly Google view volumes
between PCOS-related and fibroid-related research topics [3]. In
the period 2014–2017, monthly absolute search volumes related
to PCOS increased significantly. In this population, PCOS was diag-
nosed by an obstetrician or a gynecologist. More than half of the
patients were dissatisfied with their overall care. More than 90%
searched the Web for information, and 20% joined an online syn-
drome support group or forum. From another perspective, the
authors report that only 4% of hospital residents spent time devel-
oping their skills to recognize, diagnose and treat patients with
underlying endocrine disorders and infertility.
What explanations can we give for these findings of patients’
distress with the methods of diagnosis, information and treatment
of PCOS? And how can we remedy it?
The first reason that comes to mind is the great variability in clin-
ical expression of PCOS, which is diagnosed on three fundamental
criteria: clinical (irregularity of the anovulatory cycle), clinical-
biological (clinical hyperandrogenism (hirsutism) or increased
blood pressure and testosterone level), and radiological (large mul-
tifollicular ovaries) [4]. The combination of these criteria results in
a multitude of clinical and biological expressions. In addition, the
usual association with a metabolic syndrome dominated by insulin
resistance further increases variability. Thus, two subtypes of PCOS
have recently been proposed, on the basis of body mass index
(BMI) and hormonal criteria (LH level and the level of sex-hormone
binding-globulin (SHBG), the main transport protein of sex-steroid
hormones, the production of which in the liver depends on the
metabolic and nutritional environment) [5]. The “metabolic” sub-
type combines increased BMI, blood sugar and hyperinsulinemia
with decreased SHBG and LH; the “reproduction” subtype combines
low BMI and insulinemia with elevation of LH and SHBG. Using
this approach, new genetic variants associated with each PCOS
subtype have been identified [6]. Thus, there are distinct forms of
PCOS, implying different pathophysiology, which may likely differ
in treatment response and long-term outcome.
Pugeat M, Dewailly D
Another explanation concerns the great confusion in com-
munication about consensus-building for the diagnosis of PCOS.
This situation is partly due to the battle of experts who co-
signed the main consensuses and then made multiple amendments,
blurring the message for health professionals and patients, who
hear different messages depending on whether their syndrome
involves menstrual cycle disorder, excess hair or infertility, and
who also discover the metabolic disorders that will accompany
them throughout life.
Can these findings create an opportunity to develop an earlier
diagnostic strategy, and an education and information program
adapted to the demands of PCOS patients?
When participants in Gibson-Helm et al.’s study [2] were asked
“How can we better support women with PCOS?” the answer
was clearly: “Provide widely available educational material”. If the
information material is widely available, what patients are say-
ing is that it should be educational. This is the first message that
the community of endocrinologists and gynecologists, teachers and
members of learned societies must take on board. The educational
form of this information will have to be validated by healthcare
professionals, especially nurses, but also tested and approved by
the patients themselves. The initiative of our Australian colleagues,
who recently developed an app “Ask PCOS” for the use of patients,
which is very informative, is to be commended.
Raising awareness and supporting the training of professionals
in the diagnosis and management of PCOS is the second require-
ment highlighted by patients [1–3,7]. To achieve this objective,
it seems essential to underline the high prevalence of PCOS and
to develop the identification of PCOS at puberty or during pre-
scription of contraceptives and, ideally before any pregnancy plan,
in a prenuptial consultation screening for any pathology of preg-
nancy, including PCOS. Associating the definition of PCOS with
ovarian dysfunction and metabolic disorders of the prediabetes
type would improve identification of PCOS as a health-care prior-
ity and improve the funding of clinical research programs devoted
to screening and education and the prevention of complications of
PCOS, of which infertility and diabetes are the two components.
2. Version française
Le syndrome des ovaires polykystiques (SOPK) repose, dans sa
définition, sur une association de critères cliniques, biologiques et
échographiques dont la variabilité individuelle et la réversibilité
contribue à une grande confusion pour les patientes et les acteurs
de santé. Cette confusion est d’autant plus préjudiciable que le SOPK
concerne 10 à 15 % de la population de femmes en âge de procréer.
Dans ce numéro des Annales d’Endocrinologie, M Bouthier
et al. [1], une équipe de gynécologues de Marseille (France), rap-
porte une étude qualitative conduite sur un partage d’expériences
auto-déclarées sur sept forums Internet francophones. Le choix
d’une recherche qualitative permet selon les auteurs d’explorer le
ressenti émotionnel des patientes dans les conditions de vie réelle,
et d’identifier les problématiques complexes des femmes infer-
tiles SOPK consultant les forums. Un total de 785 commentaires
provenant de 211 femmes infertiles et diagnostiquées SOPK a été
analysé. Les principales doléances de ces patientes concernent :
(i) le retard au diagnostic et la nécessité de plusieurs consul-
tations pour aboutir à la formulation du SOPK ; (ii) le manque
d’informations claires et explicatives concernant le SOPK ; (iii) les
difficultés de traitement et notamment le contrôle d’un poids sou-
vent excessif. Les auteurs témoignent de l’impact psychologique
négatif du SOPK et de l’infertilité sur la vie quotidienne. Ils obser-
vent que cet impact négatif semble être amélioré par les échanges
de connaissances et l’apprentissage expérientiel rendus possibles
par les forums de discussion en ligne.
474
Annales d’Endocrinologie 81 (2020) 473–475
L’utilisation d’internet comme matériel de recherche soulève
de nouveaux enjeux pour lesquels les recommandations
méthodologiques et éthiques sont encore vagues. Ce type
d’approche propose un instantané de témoignages spontanés
qui, bien que n’ayant pas la valeur d’une étude prospective, rejoint
dans ses observations plusieurs articles récents qui convergent
vers les mêmes conclusions.
Gibson-Helm et al. [2], par un questionnaire online, ont
évalué l’expérience vécue des patientes lors du diagnostic et de
l’information fournie sur le SOPK. Dans cette étude, 1385 femmes
anglophones provenant de 32 pays ont été recrutées sur les sites
disponibles sur le web en langue anglaise, entre 2015 et 2016. Les
auteurs ont constaté que plus d’un tiers des femmes interrogées ont
passé 2 ans à rechercher un diagnostic pour expliquer leurs symp-
tômes et ont consulté au moins 3 et jusqu’à 5 prestataires médicaux
distincts. Seulement 35 % des patientes étaient satisfaites de la
démarche diagnostique mais moins de 20 % considéraient adap-
tées les informations apportées. Lorsque les informations étaient
adéquates, la satisfaction du diagnostic était corrélée positivement.
La consultation d’au moins 5 praticiens et un délai supérieur à 2 ans
avant le diagnostic étaient négativement associé au degré de sat-
isfaction du diagnostic quelque soit l’âge, l’origine et le lieu de vie
des patientes. Les principales difficultés ressenties par les patientes
étaient l’absence de perte de poids, l’irrégularité du cycle menstruel
et l’infertilité.
Récemment, une étude a quantifié les volumes de consulta-
tions mensuelles sur Google pour les thèmes de recherche liés au
SOPK en comparaison de ceux liés aux fibromes [3]. Dans la péri-
ode 2014–2017, les volumes mensuels de recherche absolue liés au
SOPK a augmenté de façon significative. Dans cette population, le
diagnostic de SOPK a été établi par un obstétricien ou un gynéco-
logue. Plus de la moitié des patientes étaient insatisfaites des soins
globaux. Parmi les patientes, plus de 90 % ont recherché des infor-
mations sur le Web et 20 % ont rejoint un groupe de soutien ou un
forum en ligne sur le syndrome. D’un autre point de vue, les auteurs
rapportent que parmi les résidents, seuls 4 % consacraient du temps
à développer leurs compétences pour reconnaître, diagnostiquer et
traiter les patients atteints de troubles endocriniens sous-jacents,
ainsi que ceux souffrant d’infertilité.
Quelles explications peut-on apporter à ce constat de désarroi
des patientes devant les modalités de diagnostic, d’information et
de traitement du SOPK ? Comment y remédier ?
La première raison qui s’impose à l’esprit est la grande vari-
abilité de l’expression clinique du SOPK dont le diagnostic repose
sur trois critères fondamentaux : cliniques : irrégularité du cycle
anovulatoire – clinico-biologique : hyperandrogènie clinique (hir-
sutisme) ou élévation de la testostérone – échographique : gros
ovaires multifolliculaires [4]. La combinaison de ces critères offre
une multitude d’expression clinique et biologique. De plus, cette
variabilité est encore augmentée par l’association habituelle d’un
syndrome métabolique dominé par la résistance à l’insuline. Ainsi,
récemment deux-sous types de SOPK ont été proposés sur la base de
l’index de masse corporelle (IMC) et de critères hormonaux, le taux
de LH et le niveau de la sex-hormone binding-globulin (SHBG), prin-
cipale protéine de transport des hormones stéroïdes sexuelles, dont
la production hépatique dépend de l’environnement métabolique
et nutritionnel [5]. Le sous-type métabolique associe élévation
de l’IMC, de la glycémie et de l’hyperinsulinémie avec baisse de
la SHBG et de la LH. Le sous-type reproduction associe IMC et
insulinémie basses, élévation de LH et de la SHBG. En utilisant cette
approche, de nouveaux variants génétiques, associés à chacun des
sous-types de SOPK, ont été identifiés [6]. Ainsi, il existe des formes
distinctes de SOPK ce qui suppose une physiopathologie différente
qui pourraient aussi différer dans leurs réponses au traitement et
dans leur devenir à long terme.
Pugeat M, Dewailly D
Une autre explication qui pourrait être incriminée est la grande
confusion de communication sur la recherche d’un consensus pour
le diagnostic du SOPK. Cette situation incombe en partie à la bataille
d’experts qui ont cosigné les principaux consensus puis apporté de
multiples amendements contribuant à brouiller le message pour les
acteurs de santé et les patients qui entendent différents messages
selon que leur syndrome concerne les troubles du cycle menstruel,
l’excès pilaire ou la fécondité et qui découvre aussi les troubles
métaboliques qui vont les accompagner toute la vie.
Est-ce une opportunité pour développer une stratégie de diag-
nostic plus précoce, et un programme d’éducation et d’information
adaptées à la demande des patientes SOPK.
A la question des participants de l’étude de Gibson-Helm et al.
[1] « Comment pouvons-nous mieux soutenir les femmes atteintes
du SOPK ? » la réponse a été clairement : « Fournir du matériel péd-
agogique largement disponible ». Si le matériel d’information est
largement disponible, ce que disent les patientes, c’est qu’il doit
être pédagogique. C’est la première réponse que la communauté
des endocrinologues et des gynécologues, enseignants et membres
des sociétés savantes, se doit de satisfaire. La forme pédagogique
de cette information devra être validée par les acteurs de santé
et notamment des infirmières mais aussi testée et approuvée par
les patientes eux même à l’occasion de consultations. Il faut saluer
l’initiative de nos collègues australiens qui ont développé récem-
ment une application « Ask PCOS » à l’usage des patientes, très
informative.
Sensibiliser et soutenir la formation des professionnels sur le
diagnostic et la prise en charge du SOPK est la deuxième exigence
soulignée par les patientes [1–3,7]. Pour atteindre cet objectif, il
paraît primordial de souligner la prévalence élevée du SOPK et de
développer l’identification du SOPK au moment de la puberté ou
lors de la prescription de contraceptif, et idéalement avant tout
projet de grossesse, dans une consultation prénuptiale dédiée à
toute pathologie de la grossesse, y compris le SOPK. Associer à la
définition du SOPK le dysfonctionnement ovarien et les désordres
métaboliques de type prédiabète, permettrait d’identifier mieux
le SOPK comme une priorité du système de santé et de financer
des programmes de recherche clinique consacrés au dépistage,
à l’éducation et la prévention des complications du SOPK dont
l’infertilité et le diabète sont les deux composantes.
475
Annales d’Endocrinologie 81 (2020) 473–475
Déclaration de liens d’intérêts
Les auteurs déclarent ne pas avoir liens d’intérêts.
References
[1] Authier M, Normand C, Jego M, Gaborit B, Boubli, Courbiere B. Qualitative study
of self-reported experiences of infertile women with polycystic ovary syndrome
through on-line discussion forums. Annales d’Endocrinologie. Available online
19 August 2020. DOI 10.1016/j.ando. 2020.07.1110.
[2] Gibson-Helm M, Teede H, Dunaif A, Dokras A. Delayed diagnosis and
a lack of information associated with dissatisfaction in women with
polycystic ovary syndrome. J Clin Endocrinol Metab 2017;102:604–12,
http://dx.doi.org/10.1210/jc.2016-2963.
[3] Hoyos LR, Putra M, Armstrong AA, Cheng CY, Riestenberg CK, Schooler TA,
Dumesic DA. Measures of patient dissatisfaction with health care in polycys-
tic ovary syndrome: retrospective analysis. J Med Internet Res 2020;22:e16541,
http://dx.doi.org/10.2196/16541.
[4] Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group.
Revised 2003 consensus on diagnostic criteria and long-term health risks
related to polycystic ovary syndrome (PCOS). Hum Reprod 2004;19:41–7,
http://dx.doi.org/10.1093/humrep/deh098 [PMID: 14688154].
[5] Pugeat M, Crave JC, Elmidani M, Nicolas MH, Garoscio-Cholet M,
Lejeune H, et al. Pathophysiology of sex hormone binding globulin
(SHBG): relation to insulin. J Steroid Biochem Mol Biol 1991;40:841–9,
http://dx.doi.org/10.1016/0960-0760(91)90310-2 [PMID: 1958579].
[6] Dapas M, Lin FTJ, Nadkarni GN, Sisk R, Legro RS, Urbanek M, et al.
Distinct subtypes of polycystic ovary syndrome with novel genetic associ-
ations: an unsupervised, phenotypic clustering analysis. PLoS Med 2020;17,
http://dx.doi.org/10.1371/journal.pmed.1003132.
[7] Cree-Green M. Worldwide dissatisfaction with the diagnostic process and initial
treatment of PCOS. J Clin Endocrinol Metab 2017;102:375–8.
Michel Pugeat a,b,∗
Didier Dewailly c,d
a Fédération d’endocrinologie, groupement hospitalier
Est, hospices civils de Lyon, 69677 Bron, France
b Université de Lyon1, 69003 Lyon, France
c Faculty of medicine, university of Lille, Lille, France
d Inserm, laboratory of development and plasticity of
the neuroendocrine brain, Jean-Pierre Aubert
Research Centre, UMR-S 1172, Lille, France
∗ Corresponding author.
E-mail addresses: michel.pugeat@chu-lyon.fr (M.
Pugeat), didier.dewailly@orange.fr (D. Dewailly)
 Annales d’Endocrinologie 81 (2020) 476–481

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

Original article

The plasma levels of atrial natriuretic peptide and brain natriuretic

peptide in type 2 diabetes treated with sodium-glucose
cotransporter-2 inhibitor夽,夽夽
Niveaux plasmatiques du facteur natriurétique auriculaire et du peptide
natriurétique de type B dans le diabète de type 2 traité par inhibiteur du
cotransporteur-2 de sodium-glucose
Xiu Feng , Qingwei Gu , Gu Gao , Lu Yuan , Qian Li ∗ , Ying Zhang ∗∗
Department of Endocrinology, Nanjing First Hospital, Nanjing Medical University, 68, Changle Road, Qinhuai District, Nanjing, Jiangsu Province, China

a r t i c l e i n f o a b s t r a c t

Keywords: Purpose. – The aim of this study was to determine the levels of atrial natriuretic peptide (ANP) and
Natriuretic peptide brain natriuretic peptide (BNP) after treatment with sodium-glucose cotransporter-2 (SGLT2) inhibitor
Blood glucose or dipeptidyl peptidase-4 (DPP4) inhibitor in patients with type-2 diabetes inadequately controlled by
Sodium-glucose cotransporter-2 inhibitor
insulin, and to determine whether variation in ANP levels can explain favorable cardiovascular outcome.
Dipeptidyl peptidase-4 inhibitor
Methods. – We enrolled 56 patients, aged 18–80 years, with type-2 diabetes inadequately controlled by
Type 2 diabetes
insulin: i.e., HbA1c level 7.5–10.5% despite at least 8 weeks’ injectable insulin at a stable mean dose of
20–150 IU daily, with or without no more than two oral antidiabetic agents.
Findings. – The 56 patients were randomized between 3 treatment groups: SGLT2 inhibitor (n = 18), DPP4
inhibitor (n = 19) and placebo (n = 19). Patients who received SGLT2 inhibitor or DPP4 inhibitor treat-
ment all showed significantly lower HbA1c levels, fasting blood glucose (FBG) levels and systolic blood
pressure at 24 weeks than controls. SGLT2 inhibitor treatment decreased ANP levels, BNP levels, systolic
blood pressure and weight compared with placebo. Compared to those receiving DPP4 inhibitor, patients
receiving SGLT2 inhibitor showed lower HbA1c levels (7.01 vs. 7.58%; P = 0.03), ANP levels (28.41 vs. 43.03
pg/mL; P = 0.00) and weight (66.14 vs. 71.76 kg; P = 0.04) at 24 weeks after adjusting for baseline values.
The SGLT2 inhibitor group showed higher sodium concentrations than the placebo and DPP4 inhibitor
groups (145.89 vs. 143.89 and 144.79 mmol/L, respectively; P = 0.00 and P = 0.04) at 24 weeks. ANP and
BNP levels did not significantly correlate with HbA1c and blood glucose levels.
Implications. – These results indicated that SGLT2 inhibitors may be superior to DPP4 inhibitors in reducing
risk of cardiovascular disease in diabetic patients. The major study limitation was the small number of
patients per group, which should be enlarged in further research.
© 2020 Published by Elsevier Masson SAS.

r é s u m é

Mots clés : Objectif. – L’intention de cette étude était de : déterminer les niveaux de facteur natriurétique auriculaire
Peptide natriurétique (FNA) et de peptide natriurétique de type B (BNP) après le traitement par des inhibiteurs de cotrans-
Glycémie porteurs 2 du sodium-glucose (SGLT2) ou des inhibiteurs de la dipeptidyl peptidase-4 chez les patients
Inhibiteur du cotransporteur-2 de atteints du diabète de type 2 mal contrôlés par l’insulinothérapie ; et de déterminer si la variation des
sodium-glucose
taux de FNA pourrait expliquer des résultats cardiovasculaires positifs.

夽 Trial registration: IRSCTN 2013L01573. Registered 29 August 2013.

夽夽 Trial registration: IRSCTN 2014L00001. Registered 4 January 2014.
∗ Corresponding author.
∗∗ Corresponding author.
E-mail addresses: show clancy@foxmail.com (X. Feng), gqw9299@163.com (Q. Gu), feifeigalt@163.com (G. Gao), happy yuanlu@126.com (L. Yuan), shygu@njmu.edu.cn
(Q. Li), zhangying932@163.com (Y. Zhang).

https://doi.org/10.1016/j.ando.2020.07.1113
0003-4266/© 2020 Published by Elsevier Masson SAS.
 X. Feng et al. / Annales d’Endocrinologie 81 (2020) 476–481 477

Inhibiteur de la dipeptidyl peptidase-4
Diabète de type 2
Méthodes. – Nous avons recruté 56 patients, entre 18 et 80 ans, atteints du diabète de type 2 mal contrôlé
par l’insulinothérapie : niveau de HbA1c à 7,5–10,5 % après au moins 8 semaines d’injection quotidienne
d’insuline d’une dose table de 20 à 150 IU, accompagné de zéro à deux antidiabétiques oraux.
Résultats. – Les 56 patients ont été randomisés entre trois groupes de traitements : inhibiteurs SGLT2
(n = 18), inhibiteurs DPPA (n = 19) et placebo (n = 19). Les patients qui ont reçu le traitement par inhibiteurs
SGLT ou par inhibiteurs DPPA ont tous montré à 24 semaines un taux de HbA1c significativement plus
bas ainsi qu’une glycémie à jeun (FBG) et une pression artérielle systolique significativement plus basses
par rapport aux valeurs contrôles. Le traitement par inhibiteurs SGLT2 diminue le taux de HbA1c, la
glycémie, la pression artérielle systolique et le poids par rapport au groupe placebo. Par rapport aux
patients recevant des inhibiteurs DPPA, ceux recevant des inhibiteurs SGLT2 montrent un taux plus faible
de HbA1c (7,01 % contre 7,58 % ; p = 0,03) et de FNA (28,41 pg/mL contre 43,03 pg/mL ; p = 0,00) ainsi
qu’une perte de poids (66,14 kg contre 71,76 kg ; p = 0,04) à 24 semaines après ajustement sur des valeurs
seuils. Le groupe des inhibiteurs SGLT2 montre une concentration en sodium plus élevée que les groupes
placebo et inhibiteurs DPPA (145,89 mmol/L contre 143,89 et 144,79 mmol/L, respectivement ; p = 0,00 et
p = 0,04) à 24 semaines. Les niveaux de FNA et de BNP ne sont pas corrélés de manière significative avec
le taux de HbA1c et la glycémie.
Conclusion. – Ces résultats indiquent que les inhibiteurs SGLT2 pourraient être plus efficaces que les
inhibiteurs DPPA dans la réduction des risques de maladies cardiovasculaires chez les patients diabétiques.
La limite principale de l’étude a été le nombre de patients par groupe, qui devrait être augmenté pour des
recherches plus approfondies.
© 2020 Publié par Elsevier Masson SAS.

1. Introduction
Diabetes is well known as one of the risk factors on the incidence
of cardiovascular disease [1]. There are plenty types of lowering-
glucose drugs to choose. However, some drugs were reported to
increase the risk of cardiovascular events, e.g. rosiglitazone [2].
Sodium-glucose cotransporter-2 (SGLT2) inhibitors and dipeptidyl
peptidase 4 (DPP4) inhibitors are relatively novel classes of oral
hypoglycemia drugs developed in the treatment of type 2 dia-
betes. SGLT2 inhibitors have an insulin-independent mechanism
for the correction of hyperglycemia through decreasing renal glu-
cose reabsorption, resulting in increased urinary glucose excretion,
declined plasma glucose, a mild osmotic diuresis, and a net loss
of calories [3]. DPP4 inhibitors decrease blood glucose levels via
inhibiting the degradation of the incretins glucagon-like peptide-
1 (GLP-1) and gastric inhibitory polypeptide (GIP), which leads to
reducing the secretion of glucagon, increasing secretion of insulin
and delaying gastric emptying [4]. SGLT2 inhibitors were reported
to reduce cardiovascular and all-cause mortality in patients with
T2DM and established cardiovascular disease [5].
Atrial natriuretic peptide (ANP) and brain natriuretic peptide
(BNP) are the cardiac natriuretic peptides (NP), true “car-
diometabolic” hormones well known for their renal, endocrine and
cardiovascular activities leading to diuresis, natriuresis and vasodi-
lation.
ANP is predominantly synthesized, stored in preformed gran-
ules, and released from atrial cardiomyocytes. BNP is produced
in atrial and ventricular cardiomyocytes, and both can be largely
secreted when ventricular pressures increasing [6]. ANP and BNP
exert hemodynamic and tubular renal actions, leading to decrease
sodium reabsorption and arterial blood pressure. ANP and BNP are
structurally and functionally similar to each other [7] and exert
their actions through interaction with their cell-surface receptors:
the cGMP-signaling receptor NPRA and a natriuretic peptide clear-
ance receptor (NPRC) [8], which acts predominantly as a clearance
receptor for both ANP and BNP [9]. In addition to clearance through
NPRC, NPs are inactivated by neutral endopeptidases located within
renal tubular cells and the vasculature, as well as insulin degrad-
ing enzyme and dipeptidyl peptidase-IV, and may also be passively
excreted into the urine [10]. In practice, there is increasing aware-
ness to the significance of BNP in the pathophysiology, diagnosis
and prognosis of cardiovascular disorders beyond heart failure [11].
The aim of this study was to evaluate and compare the changes in
plasma ANP and BNP levels after SGLT2 inhibitor or DPP4 inhibitor
treatments in patients with type 2 diabetes inadequately controlled
with insulin.
2. Patients and methods
This study was performed in the Department of Endocrinology,
Nanjing First Hospital, affiliated to Nanjing Medical University. It
also was approved by the appropriate independent ethics commit-
tees and regulatory authorities and was conducted in accordance
with the ethical principles of the Declaration of Helsinki and
International Conference on Harmonisation Good Clinical Practice
guidelines. After the purpose and procedures to the study were fully
explained, all subjects provided informed consent before enrolling
in the study.
We enrolled 56 patients with hyperglycemia inadequately con-
trolled with insulin aged 18∼80 years with type 2 diabetes mellitus
between October 2011 and December 2014. After taking a sta-
ble mean dose of ≥ 20 IU and ≤ 150 IU injectable insulin daily and
combining with or without no more than two oral antidiabetic
agents for at least 8 weeks, patients who still experienced inad-
equate glycemic control, as defined by HbA1c levels of 7.5∼10.5%,
were recruited. The exclusion criteria were as follows: (1) treat-
ment with more than two oral antidiabetic agents within 6 weeks of
enrollment and administration of SGLT2 inhibitors, DPP4 inhibitors
and/or GLP1 analogues before randomization; (2) history of dia-
betes insipidus; (3) type 1 diabetes mellitus; (4) history of diabetic
ketoacidosis or diabetic coma; (5) severe uncontrolled hyper-
tension defined as systolic blood pressure ≥ 180 mmHg and/or
diastolic blood pressure ≥ 110 mmHg; (6) history or current diag-
nosis of significant comorbid diseases, such as cardiovascular,
hepatic and renal diseases; (7) pregnancy or possible pregnancy;
(8) positive test for glutamic acid decarboxylase autoantibodies.
These eligible participants were randomly assigned (ratio
1:1:1) to receive dapagliflozin (10 mg) plus conventional therapy,
saxagliptin (5 mg) plus conventional therapy or placebo plus con-
ventional therapy (insulin with or without oral antidiabetic agents,
except for SGLT2 inhibitors, DPP4 inhibitors and GLP1 analogues)
for 24 weeks, and, after 4 weeks of treatment, patients lack-
ing glycemic control (fasting blood glucose > 11.1 mmol/L) were
478 X. Feng et al. / Annales d’Endocrinologie 81 (2020) 476–481
Table 1
Baseline demographics and clinical characteristics.
Characteristic Placebo (n = 18)
Age, years 60.7 (7.0)
Duration, years 14.4 (5.9)
Sex: male, n (%) 9 (50)
Weight, kg 70.5 (11.8)
BMI 24.31 (3.51)
Heart rate, bpm 74.8 (7.9)
Systolic blood pressure, mmHg 123.4 (17.4)
Diastolic blood pressure, mmHg 75.2 (8.4)
HbA1c, percent 8.52 (0.60)
Fasting plasma glucose, mmol/l 8.92 (2.01)
QRS, mV 75.8 (6.5)
QT, ms 375.6 (24.6)
appropriate to increase the insulin dose properly, based on their
particular symptoms.
Blood samples were taken from the antecubital vein after 10
hours of overnight fasting. These blood samples were centrifuged
within 30 min at 1500 rpm for 10 min and were subsequently stored
in aliquots without preservatives at −80 ◦ C. ANP and BNP lev-
els were measured by commercially available human ELISA kits
(ANP: catalogue numbers MLI023526, Enzyme-linked Biotechnol-
ogy Co., Ltd., Shanghai, China; BNP: catalogue numbers MLI025108,
Enzyme-linked Biotechnology Co., Ltd., Shanghai, China). The ANP
intra-assay coefficient of variation is 6.6% and inter-assay coeffi-
cient of variation is 7.6%. The BNP intra-assay coefficient of variation
is 7.1% and inter-assay coefficient of variation is 8.4%. Sodium con-
centrations, fasting plasma glucose levels, and HbA1c levels were
assayed using routine methods.
3. Statistical analysis
Normally distributed and continuous variables were expressed
as mean (SD). Unpaired Student’s t-tests were used to compare all
the variables measured for each treatment group and to determine
significant differences between groups. Analyses of continuous
outcomes were based on a covariance (ANCOVA) model, with treat-
ment as a fixed effect and baseline as the covariate, and were
also used to estimate differences between placebo and treatment
groups. Qualitative data were described as percentages and ana-
lyzed using Chi2 (␹2 ) test as indicated. The P-value reported was
two-sided and values less than 0.05 was considered statistically
significant. The association between ANP and BNP levels and other
parameters was determined using the Pearson’s correlation coef-
ficient. Non-normally distributed variables were log-transformed
for further evaluation. All analyses were performed using the SPSS
software (Version 20.0, SPSS Inc, USA).
4. Results
In all, 56 people with T2DM were enrolled and randomly
assigned to the three groups (n= 18, 19 and 19 patients in placebo,
dapagliflozin, and saxagliptin group, respectively). There were no
significant differences in baseline characteristics between three
groups (Table 1).
In the SGLT2 inhibitor group, ANP levels were significantly
reduced to 37.51 (22.65) pg/mL at 12 weeks, 28.41 (19.61) pg/mL
at 24 weeks compared with 46.09 (17.83) pg/mL in the placebo
group (P < 0.001). In the DPP4 inhibitor group, ANP levels were
slightly increased to 40.60 (17.84) pg/mL at 12 weeks, 43.03 (17.57)
pg/mL at 24 weeks, and there was no significant difference between
DPP4 inhibitor and placebo groups. Comparing the SGLT2 inhibitor
groups with the DPP4 inhibitor group for ANP levels, a statistical
difference between them was observed (P < 0.001) (Fig. 1a). BNP
levels within the SGLT2 inhibitor group were also more significantly
SGLT2 inhibitor (n = 19) DPP4 inhibitor (n = 19) P-value
58.3 (8.0) 60.7 (7.1) 0.524
9.4 (6.9) 12.4 (6.2) 0.161
10 (52.6) 9 (47.4) 0.949
71.4 (9.5) 73.5 (9.5) 0.667
24.82 (3.14) 25.93 (2.08) 0.18
68.9 (10.7) 74.4 (10.8) 0.134
124.7 (12.8) 131.4 (12.4) 0.196
78.2 (8.0) 77.6 (9.5) 0.542
8.31 (0.74) 8.48 (0.44) 0.52
9.17 (1.86) 9.82 (2.09) 0.375
104.1 (11.9) 97.7 (8.7) 0.108
397.1 (26.5) 387.2 (20.7) 0.134
reduced to 12.03 (5.16) pg/mL at 12 weeks, and 9.20 (4.42) pg/mL
at 24 weeks compared with 16.58 (9.91) pg/mL in the placebo
group (P < 0.05). In DPP4 inhibitor group, BNP levels increased to
11.80 (9.23) pg/mL at 12 weeks, 11.84 (8.54) pg/mL at 24 weeks,
while there was no significant difference between DPP4 inhibitor
and placebo groups (Fig. 1b). The sodium concentrations of the
SGLT2 inhibitor group was increased at 12 weeks compared with
the other groups, but it increased to 145.9 (2.1) mmol/L at 24 weeks
compared with 143.9 (2.6) mmol/L in placebo group (P < 0.01) and
144.8 (2.3) mmol/L in DPP4 inhibitor (P < 0.05). There was no sig-
nificant change in sodium concentrations in the DPP4 inhibitor
group at both 12 and 24 weeks compared with placebo treat-
ment (Fig. 1c). Decreasing in fasting blood glucose concentrations
and HbA1c levels was both distinct in SGLT2 inhibitor and DPP4
inhibitor group, at 12 weeks, fasting blood glucose levels in SGLT2
inhibitor group and DPP4 inhibitor group were separately reduced
to 7.78 (1.20) mmol/L and 8.53 (1.70) mmol/L compared with 9.83
(1.63) mmol/L in the placebo group (P < 0.001 and P < 0.01), when at
24 weeks, fasting blood glucose levels in SGLT2 inhibitor group and
DPP4 inhibitor group were correspondingly reduced to 7.48 (1.05)
mmol/L and 7.96 (1.48) mmol/L compared with 9.36 (1.55) mmol/L
in the placebo group (P < 0.001 and P < 0.001) (Fig. 1d). HbA1c levels
in SGLT2 inhibitor group and DPP4 inhibitor group were respec-
tively reduced to 7.19 (0.86) % and 7.64 (0.56) % compared with 8.48
(0.83) % in the placebo group (P < 0.001 and P < 0.001) at 12 weeks,
and at 24 weeks, The level of HbA1c in the SGLT2 inhibitor group
reduced to 7.01 (0.70) % while DPP4 inhibitor group 7.58 (0.53) %
compared with placebo group 8.82 (0.68) % (P < 0.001 and P < 0.001)
and there was a statistical difference between the two treatment
groups (P < 0.05) (Fig. 1e). SGLT2 inhibitor treatment also signif-
icantly decreased systolic blood pressure to 122.6 (10.5) mmHg
compared with placebo treatment increased systolic blood pres-
sure to 133.9 (17.1) mmHg (P < 0.01) at 24 weeks, DPP4 inhibitor
treatment substantially decreased systolic blood pressure to 129.8
(14.7) compared with placebo treatment (P < 0.05) at 24 weeks
(Fig. 1f). What’s more, SGLT2 inhibitors had a significant influ-
ence on weight loss, at 24 weeks, this treatment reduced patients’
mean weight to 66.14 (7.82) kg, while placebo treatment increased
weight to 70.94 (11.81) kg (P < 0.001), and DPP4 inhibitor treatment
reduced weight to 71.76 (9.84) kg (P < 0.05) (Fig. 1g).
There were no statistical significant differences in diastolic
blood pressure, QRS and QT among the three groups. Multivariate
linear regression analysis indicated that ANP and BNP levels were
not related to plasma sodium, blood glucose or HbA1c levels.
5. Discussion
According to the physiology of ANP and BNP, there are several
reasons for SGLT2 inhibitors lowering ANP and BNP levels. First,
SGLT2 inhibitors act on proximal tubule to reabsorb sodium and
water, resulting in increased urine volume thereby decreased blood
 X. Feng et al. / Annales d’Endocrinologie 81 (2020) 476–481 479

Fig. 1. Change from baseline in atrial natriuretic peptide (ANP) levels, brain natriuretic peptide (BNP) levels, sodium levels, HbA1c levels, fasting blood glucose (FBG)
concentrations, systolic blood pressure and weight in the placebo group, the sodium glucose co-transporter 2 (SGLT2) inhibitor group and the DPP4 inhibitor group at
baseline, 12 weeks and 24 weeks. *: P < 0.05. **: P < 0.01.

volume, which could reduce the preload and afterload of ventricle
with subsequentl decreases in ANP and BNP secretion [12]. Second,
cell-surface receptors of ANP and BNP are expressed most abun-
dantly in proximal tubules cells. When dapagliflozin inhibits SGLT2
in the early proximal tubules, it could also block the cGMP-signaling
receptor NPRA and, in compensation, increase combination of the
natriuretic peptide clearance receptor NPRC, resulting in ANP and
BNP decreased levels. Third, ANP effect on renal sodium absorption
is also associated with an inhibition of SGLT2 by directly increas-
ing intracellular cyclic guanosine monophosphate (cGMP) and/or
through the release of nitric oxide [13]. The present study involved
patients treated with an SGLT2 inhibitor, which may result in
reduced ANP levels via theses negative feedback regulation mech-
anisms. Fourth, ANP and BNP are partly degraded by neprilysin
(neutral endopeptidases) [14], and neprilysin is not only respon-
sible for degradation of NPs, but also of a range of other vasoactive
peptides, including Ang II [15]. SGLT2 inhibitors reduce the sodium
and water reabsorption in the proximal tubule and may activate
the renin–angiotensin–aldosterone system (RAAS), thus increase
the level of angiotensin-II (Ang II). The level of neprilysin may be
upregulated by increased Ang II, subsequently clear more ANP and
BNP.
We thought DPP4 inhibitors might decline ANP and BNP lev-
els similarly to SGLT2 inhibitors. There are several explanations on
these unexpected results. First, ANP and BNP degradation could be
inactivated by DPP4 [14,16], and this could be prevented by addi-
tion of DPP4 inhibitors. Second, DPP4 inhibitors can normalize a
rise of Ang II in hypertension rats [17], which may down-regulate
the level of neprilysin, and subsequently reduce the degradation of
ANP and BNP. All together, these mechanisms could contribute to
a moderate increase in fANP and BNP levels under DPP4 inhibitor
treatment. In addition, some studies suggest that DPP4 inhibitors by
activating the sympathetic nervous system and interfering with the
RAAS, may influence blood pressure [18] increasing ANP and BNP.
480 X. Feng et al. / Annales d’Endocrinologie 81 (2020) 476–481
Therefore, DPP4 inhibitors by a complex combination of effects may
explain a slightly lower benefit in ANP and BNP levels than SGLT2
treatment but yet better than placebo.
Since EMPA-REG OUTCOME trial shows that SGLT2 inhibitor
can reduce hospitalization for heart failure and all-cause mor-
tality in high-risk individuals with type 2 diabetes [5], there are
considerable discussion and debate around the potential mecha-
nisms through which these benefits were achieved [19]. Theories
have been put forward that SGLT2 inhibitor may favorably affect
myocardial loading conditions through optimization of preload (via
osmotic diuresis and natriuresis) and afterload (by reduction of
blood pressure) [20]. Some people reckon that SGLT2 inhibition
may offer cardiac protection via improving myocardial substrate
utilization and efficiency [21,22]. Shi et al. [23] show that the
cardiovascular protective effect of empagliflozin stems from the
decreased expression of BNP and ANP through zebrafish heart fail-
ure model research elucidating.
In the clinical setting, analysis of serum levels of ANP and BNP is
used for diagnosis and prognosis of congestive heart failure (CHF)
in both diabetic and nondiabetic subjects. In the Hoorn Study [24],
a slight increase in BNP levels was a strong predictor of both left
ventricular hypertrophy and diastolic left ventricular dysfunction,
and this association was particularly evident in T2DM patients.
Recent studies have demonstrated that NT-proBNP predicts cardio-
vascular events in T2DM patients and has a greater predictive value
than other well-established cardiac risk markers [25–27]. A 10-year
longitudinal study performed on asymptomatic high-risk diabetic
patients found that NT-proBNP ≥ 38 pg/mL was the only indepen-
dent predictor of asymptomatic coronary artery disease [28]. In
the present study, the SGLT2 inhibitor reduced ANP and BNP lev-
els respectively by ∼11.51 pg/mL to 28.41 pg/mL and ∼4.63 pg/mL
to 9.20 pg/mL. Besides, we also observed SGLT2 inhibitors to lower
the systolic blood pressure, which could reduce the afterload of left
ventricle. These reductions may influence left ventricular function
and coronary artery disease, based on previous findings, such that
SGLT2 inhibitors may benefit the cardiovascular systems of T2DM
patients.
The present research has some limitations that should be con-
sidered. The major limitation concerns the little number of patients
included within each group in this study. This small sample size
could be enlarged in further research. Furthermore, fewer observa-
tional value on the coronary heart disease were used in this study
and there is not enough evidence to support the conclusion. These
are all the points we need to pay attention to in our future research
work.
6. Conclusion
We showed that both SGLT2 inhibitors and DPP4 inhibitors dis-
tinctly decreased blood glucose levels and systolic blood pressure,
while SGLT2 inhibitors achieved better glycemic control than DPP4
inhibitors. SGLT2 inhibitors may get more weight loss than DPP4
inhibitors. Moreover, SGLT2 inhibitors can also reduce ANP and BNP
levels, yet DPP4 inhibitors had no significant influence on them.
These results indicated that SGLT2 inhibitors may be superior to
DPP4 inhibitor to help to significantly reduce the risk of develop-
ment of the cardiovascular disease in patients with diabetes.
Ethics and consent to participate
The investigations were carried out in accordance with the Dec-
laration of Helsinki, including written informed consent from all
participants. The study was approved by the ethics committee of
the Nanjing First Hospital (Nanjing, China).
Availability of data and materials
The data are subject to national data protection laws, and restric-
tions were imposed by the Ethics Committee of Nanjing First
Hospital to ensure data privacy of the study participants. Please
contact the corresponding author Li Qian in case of further ques-
tions.
Disclosure of interest
The authors declare that they have no competing interest.
Funding
This study was supported by a grant from Jiangsu Natural
Science Foundation (BK20171121), Jiangsu Planned Projects of
Postdoctoral Research Funds, the Peak of Six Personnel in Jiangsu,
and the Nanjing Medical Science and Technique Development
Foundation.
Authors’ contributions
L Q and Z Y designed this study. F X, G QW wrote the manuscript.
Z Y, G G, Y L contributed data. L Q drafted the statistical analysis plan.
L DM performed the statistical analysis. All authors revised and read
the final version of the manuscript and approved submitting and
publication.
Acknowledgements
We thank AstraZeneca provided the dapagliflozin, saxagliptin
and matching placebo.
References
[1] Association, American, Cardiovascular disease and risk management: stan-
dards of medical care in diabetes-2019. Diabetes Care 2019;42:S103–23.
[2] Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction
and death from cardiovascular causes. N Engl J Med 2007;356:2457–71.
[3] Neumiller JJ, White Jr JR, Campbell RK. Sodium-glucose co-transport inhibitors:
progress and therapeutic potential in type 2 diabetes mellitus. Drugs
2010;70:377–85.
[4] Mulvihill EE, Drucker DJ. Pharmacology, physiology, and mechanisms of action
of dipeptidyl peptidase-4 inhibitors. Endocr Rev 2014;35:992–1019.
[5] Zinman B, Wanner C, Lachin JM, Fitchett D, Bluhmki E, Hantel S, et al., EMPA-REG
OUTCOME Investigators. Empagliflozin, cardiovascular outcomes, and mortal-
ity in type 2 diabetes. N Engl J Med 2015;373:2117–28.
[6] Yoshimura M, Yasue H, Okumura K, Ogawa H, Jougasaki M, Mukoyama M, et al.
Different secretion patterns of atrial natriuretic peptide and brain natriuretic
peptide in patients with congestive heart failure. Circulation 1993;87:464–9.
[7] Sudoh T, Kangawa K, Minamino N, et al. A new natriuretic peptide in porcine
brain. Nature 1988;332:78–81.
[8] Potter LR, Abbey-Hosch S, Dickey DM. Natriuretic peptides, their receptors, and
cyclic guanosine monophosphate-dependent signaling functions. Endocr Rev
2006;27:47–72.
[9] Rubattu S, Sciarretta S, Morriello A, Calvieri C, Battistoni A, Volpe M. NPR-C:
a component of the natriuretic peptide family with implications in human
diseases S, Morriello A, et al. J Mol Med (Berl) 2010;88:889–97.
[10] Volpe M, Rubattu S, Burnett Jr J. Natriuretic peptides in cardiovascular diseases:
current use and perspectives. Eur Heart J 2014;35:419–25.
[11] Diez J. Chronic heart failure as a state of reduced effectiveness of the natriuretic
peptide system: implications for therapy. Eur J Heart Fail 2017;19:167–76.
[12] Gruden G, Landi A, Bruno G. Natriuretic peptides, heart, and adipose tissue:
new findings and future developments for diabetes research. Diabetes Care
2014;37:2899–908.
[13] Gonzalez Bosc LV, Elustondo PA, Ortiz MC, Vidal NA. Effect of atrial natriuretic
peptide on sodium-glucose cotransport in the rat small intestine. Peptides
1997;18:1491–5.
[14] McMurray JJ, Packer M, Desai AS, Gong J, Lefkowitz MP, Rizkala AR, et al.
PARADIGM-HF Investigators and Committees. Angiotensin-neprilysin inhi-
bition versus enalapril in heart failure. N Engl J Med 2014;371:993–1004,
http://dx.doi.org/10.1056/NEJMoa1409077.
[15] Judge P, Haynes R, Landray MJ, Baigent C. Neprilysin inhibition in chronic kidney
disease. Nephrol Dial Transplant 2015;30:738–43.
 X. Feng et al. / Annales d’Endocrinologie 81 (2020) 476–481
[16] Brandt I, Lambeir AM, Ketelslegers JM, VanderheydenM, Scharpe S, De Meester
I. Dipeptidyl-peptidase IV converts intact B-type natriuretic peptide into its
des-SerPro form. Clin Chem 2006;52:82–7.
[17] Kawase H, Bando YK, Nishimura K, oyama M, Monji A, Murohara T. A
dipeptidyl peptidase-4 inhibitor ameliorates hypertensive cardiac remodeling
via angiotensin-II/sodium-proton pump exchanger-1 axis. J Mol Cell Cardiol
2016;98:37–47.
[18] Zhang J, et al. DPP-4 inhibitors as potential candidates for antihypertensive
therapy: improving vascular inflammation and assisting the action of tradi-
tional antihypertensive drugs. Front Immunol 2019;10:1050.
[19] Abdul-Ghani M, Del Prato S, Chilton R. Defronao RA.SGLT2 inhibitors and
cardiovascular risk: lessons learned from the EMPA-REG OUTCOME Study. Dia-
betes Care 2016;39:717–25.
[20] Sattar N, McLaren J, Kristensen SL, Preiss D, McMurray JJ. SGLT2 Inhibition and
cardiovascular events: why did EMPA-REG Outcomes surprise and what were
the likely mechanisms? Diabetologia 2016;59:1333–9.
[21] Ferrannini E, Mark M, Mayoux E. CV protection in the EMPA-REG OUTCOME
Trial: a “thrifty substrate” hypothesis. Diabetes Care 2016;39:1108–14.
[22] Lopaschuk GD, Verma S. Empagliflozin’s fuel hypothesis: not so soon. Cell
Metab 2016;24:200–2.
481
[23] Shi XJ, Verma S, Yun J, Brand-Arzamendi K, Singh KK, Liu X, et al. Effect of
empagliflozin on cardiac biomarkers in a zebrafish model of heart failure: clues
to the EMPA-REG OUTCOME trial? Mol Cell Biochem 2017;433:97–102.
[24] Kroon MH, van den Hurk K, Alssema M, Kamp O, Stehouvwe CD, Henry RM,
et al. Prospective associations of B-type natriuretic peptidewith markers of left
ventricular function in individuals with and without type 2 diabetes: an 8-year
follow-up of the Hoorn Study. Diabetes Care 2012;35:2510–4.
[25] Sharma A, et al. Clinical and biomarker predictors of expanded heart failure
outcomes in patients with type 2 diabetes mellitus after a recent acute coronary
syndrome: insights from the EXAMINE trial. J Am Heart Assoc 2020;9:e012797.
[26] Rorth R, et al. The prognostic value of troponin T and N-terminal pro B-type
natriuretic peptide, alone and in combination, in heart failure patients with
and without diabetes. Eur J Heart Fail 2019;21:40–9.
[27] Fraty M, et al. Prognostic value of plasma MR-proADM vs. NT-proBNP for heart
failure in people with type 2 diabetes: the SURDIAGENE prospective study.
Diabetologia 2018;61:2643–53.
[28] Cosson E, Nguyen MT, Pham I, Pontet M, Nitenberg A, Valensi P. N-terminal pro-
B-type natriuretic peptide: an independent marker for coronary artery disease
in asymptomatic diabetic patients. Diabet Med 2009;26:872–9.
 Annales d’Endocrinologie 81 (2020) 482–486

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

Original article

Differentiated thyroid carcinoma in sporadic and familial

presentations of acromegaly: A case series
Cancer différencié de la thyroïde dans les formes familiales et sporadiques de
l’acromégalie : une série de cas cliniques
Amelia Rogozinski a,∗ , Adrian F. Daly b , Adriana Reyes a , Alejandra Furioso a ,
Albert Beckers b , Alicia Lowenstein a
a
Endocrinology division, Hospital J.M. Ramos Mejia, Buenos Aires, Argentina
b
Service d’Endocrinologie, Centre Hospitalier Universitaire de Liège, Liège Université, 4000 Liège, Belgium

a r t i c l e i n f o a b s t r a c t

Keywords: Background. – In acromegaly, chronic growth hormone (GH) and insulin-like growth factor-1 (IGF-1)
Acromegaly exacerbate comorbidities in multiple organs. Differentiated thyroid carcinoma (DTC) has been reported
Thyroid cancer as being a comorbid condition in acromegaly. Acromegaly is usuallysporadic, but 5% of cases may be
FIPA
genetic. The most frequent inheritable form of acromegaly is related to germline mutations in the aryl
AIP
hydrocarbon receptor interacting protein (AIP) gene. Epidemiological data on the relationship between
active acromegaly, its familial forms and DTC are sparse. We present the investigation of a FIPA family
(familial isolated pituitary adenoma) with homogeneous acromegaly and 6 sporadic acromegaly patients
with DTC.
Patients and methods. – A study of 59 acromegaly patients assessed thyroid nodules on ultrasound and
fine-needle aspiration biopsy following the ATA 2015 criteria. We diagnosed 7 differentiated thyroid
carcinomas. Resected thyroid carcinoma tissues were stained using an anti-AIP antibody. Analysis of
germline and tumor-derived DNA for variants in the AIP and MEN1 genes were performed in the FIPA
kindred.
Results. – We describe one FIPA patient and 6 sporadic acromegaly cases with DTC. The FIPA family
(AIP mutation negative) consisted of two sisters, one of whom had a DTC with intermediate risk and
incomplete structural response to therapy. In our study, DTC in sporadic acromegaly had a low recurrence
rate (6/6), and excellent response to therapy (6/6). Immunohistochemistry for AIP showed similar or
increased staining intensity in DTC versus normal thyroid tissue.
Conclusion. – In our cohort of sporadic and familial forms of acromegaly with DTC, AIP did not appear to
influence thyroid cancer progression.
© 2020 Published by Elsevier Masson SAS.

r é s u m é

Mots clés : Contexte. – Lors d’une acromégalie, la présence chronique d’hormones de croissance (GH) et de somatomé-
Acromégalie dine C (insulin-like growth factor-1 ou IGF-1) exacerbe l’existence de comorbidités dans plusieurs
Cancer thyroïdien organes. Le cancer différencié de la thyroïde (CDT) a été signalé comme une comorbidité de l’acromégalie.
FIPA L’acromégalie est généralement sporadique mais, dans 5 % des cas elle peut être d’origine génétique.
AIP
La forme héréditaire la plus fréquente est liée à des mutations germinales du gène codant la protéine
d’interaction du récepteur d’aryl hydrocarbone (gène AIP). Les données épidémiologiques sur la relation
entre la forme sporadique de l’acromégalie, sa forme familiale et le CDT sont rares. Nous présentons
l’étude d’une famille atteinte de FIPA (adénome hypophysaire familial isolé) présentant une acromégalie
homogène et 6 cas d’acromégalie sporadique avec un DTC.

∗ Corresponding author at: Endocrinology division, Hospital J.M. Ramos Mejia, Buenos Aires, Argentina.
E-mail address: ameliasu2000@yahoo.com.ar (A. Rogozinski).

https://doi.org/10.1016/j.ando.2020.05.004
0003-4266/© 2020 Published by Elsevier Masson SAS.
 A. Rogozinski et al. / Annales d’Endocrinologie 81 (2020) 482–486 483

Patients et méthodes. – Une étude de 59 patients atteints d’acromégalie a évalué des nodules thyroïdiens
par ultra-son et par ponction biopsie en suivant les critères ATA (American Thyroid Association) de 2015.
Nous avons diagnostiqué 7 cancers différenciés de la thyroïde. Les tissus réséqués des cancers thyroïdiens
ont été colorées à l’aide d’un anticorps anti-AIP. Nous avons analysé de l’ADN germinal et tumoral afin de
rechercher l’existence de mutations des gènes AIP et MEN1 au sein de la famille FIPA.
Résultats. – Nous avons observé un patient FIPA et 6 patients atteints d’acromégalies sporadiques avec
CDT. La famille FIPA (mutation AIP négative) était constituée de deux sœurs, l’une atteinte d’un CDT avec
un risque intermédiaire et une réponse structurale incomplète au traitement. Dans notre étude, les CDT
des acromégalies sporadiques montraient un faible taux de récidive (6/6) et une réponse très conclu-
ante au traitement (6/6). L’immunohistochimie des AIP a montré une intensité du marquage similaire ou
augmentée lors d’un CDT par rapport à un tissu thyroïde sain.
Conclusion. – Lors de notre étude cohorte des formes sporadiques et familiales de l’acromégalie avec CDT,
il semblerait que le gène AIP n’influence pas le développement du cancer de la thyroïde.
© 2020 Publié par Elsevier Masson SAS.

1. Introduction 2. Patients and methods

Acromegaly is caused by chronic growth hormone (GH) hyper-
secretion usually from a pituitary adenoma [1]. GH and insulin-like
growth factor 1 (IGF1) are major mediators of cellular and
tissue growth and metabolism and also have mitogenic and
anti-apoptotic properties. The pathological effects of chronic
GH/IGF-1 hypersecretion in acromegaly include overgrowth and
changes in physical appearance, cardiovascular disease, arthritis
and metabolic disturbances [2]. Together, these pathological effects
lead to increased mortality in uncontrolled acromegaly, mainly due
to cardio-respiratory disease.
There is a long-running controversy as to whether increased
risks of benign or malignant neoplasms exist in acromegaly due
to the mitogenic/proliferative effects of GH/IGF-1. One of the most
frequent associations of other tumors in acromegaly is that of thy-
roid nodules, which occurred in around third of cases in the largest
acromegaly series, although thyroid cancer is less frequent [3]. We
previously reported a prospective study evaluating the prevalence
of thyroid nodular disease in 34 acromegalic patients in whom
ultrasound (US) was routinely performed under ATA guidelines, fol-
lowed by surgery where indicated [4,5]. We found a high prevalence
of nodular thyroid disease overall, 23/34 (67.6%) and differentiated
thyroid carcinoma (DTC), was present in 11% of our acromegaly
population.
Acromegaly usually occurs in a sporadic setting, but up to 5%
of cases may be due to inherited genetic causes [6,7]. The most
frequent and best characterized inheritable form of acromegaly
is related to germline mutations in the aryl hydrocarbon recep-
tor interacting protein (AIP) gene [8,9]. About 20% of AIP mutation
carriers develop a pituitary adenoma, and typically these are large,
aggressive somatotropinomas that occur at a young age and are
relatively treatment-resistant [10]. AIP mutations usually occur in
the setting of familial isolated pituitary adenoma (FIPA) kindreds or
as pituitary gigantism, whereas associations with tumorigenesis in
other tissues has only rarely been noted [11]. Only a few individual
cases of thyroid cancer have been reported in AIP mutation positive
kindreds [12,13].
There are limited epidemiological data on the relationship
between active acromegaly (sporadic or FIPA) and DTC. Mian et al.
reported 12 cases of DTC in acromegaly patients in which no
pathological germline mutations in AIP were seen; immunohis-
tochemistry for AIP was not altered in the thyroid cancer tissue
as compared with surrounding normal tissue [14]. To explore this
topic further we studied the clinical, genetic and histological char-
acteristics of DTC in a series of seven acromegaly patients that
presented sporadically or in the setting of FIPA.
We undertook a study of sporadic or familial acromegaly
patients with DTC at the Endocrine Division, Hospital JM Ramos
Mejia, Buenos Aires, Argentina. The population consisted of
patients diagnosed from 2010 to 2017 in a consecutive group of
59 acromegalic patients.
All the patients were evaluated by thyroid US and FNA biopsy in
thyroid nodules following ATA 2015 criteria. The clinical records
were searched for endocrine investigations, family history of
tumors (endocrine and otherwise) and treatment details. In seven
patients with a positive history of acromegaly and DTC, we
extracted details on tumor diagnosis, staging and treatment, in
addition to histopathology analysis.
2.1. Immunohistochemistry
Formalin-fixed, paraffin embedded blocks of resected thy-
roid carcinoma were identified and thin sections were prepared
for staining using an anti-AIP antibody as previously described
[13,15,16].
2.2. Genetic analyses
Analysis of germline and tumor derived DNA for variants in
the AIP and MEN1 genes was performed as previously described
[17]. In addition, studies were also performed for deletions in the
AIP/MEN1 genes using an MLPA kit (SALSA MLPA Probe Kit P244,
MRC-Holland).
3. Results
We identified 59 acromegaly patients treated at our institution,
of which seven (11.9%) had a confirmed histopathological diag-
nosis of DTC. Six sporadic acromegaly patients had differentiated
thyroid carcinoma (DTC): five were papillary thyroid carcino-
mas (PTC) and one follicular thyroid carcinoma (FTC). All patients
were female: median age: 50 years (range: 31–68 years), median
maximum thyroid nodule diameter 15 mm (range: 9–33 mm),
median follow-up since DTC diagnosis: 44 months (range: 19–70
months). The patients were TNM (8th ed.) Stage I, with an initially
low risk of recurrence/persistent disease, according the American
Joint Committee on Cancer (AJCC) and American Thyroid Asso-
ciation (ATA) risk systems respectively (Table 1). In all cases,
acromegaly remained active in spite of multimodal therapy. In
addition to acromegaly and DTC the patients presented other
neoplasms: Hürthle cell adenoma (Patient 3), thyroid follicular
484 A. Rogozinski et al. / Annales d’Endocrinologie 81 (2020) 482–486
Table 1
Clinical features of differentiated thyroid carcinoma in seven acromegaly patients.
Patient number 1 2
Gender F F
Age (years) 63 38
Thyroid nodule max. diameter (mm) 14 9 30
Thyroid histology PTC PTC+FA
TNM 8th Edition I I
Initial therapy TT TT+RAI
Recurrence risk ATA Low Low
Follow up (month) 19 48
Therapy response Excellent Excellent
F: female; ND: not determined; PTC: papillary thyroid carcinoma; FA: follicular adenoma; FC: follicular carcinoma; TT: total thyroidectomy; RAI: radioactive iodine ablation;
CLND: central lymph node dissection.
a
FIPA case.
adenoma (Patient 2), atypical endometrial hyperplasia and colonic
polyp (Patient 2), and breast carcinoma (Patients 1, 3). Genetic
screening was not performed in the sporadic acromegaly patients
as none had features associated with AIP mutations and the rate of
mutation positivity in this group of patients is known to be very
low [6].
Immunohistochemistry for AIP was performed in DTC tissue
from 5/6 cases. AIP staining was similar to or increased in intensity
as compared with normal thyroid tissue, although no consistent
pattern of alteration in AIP staining intensity was seen.
One of the acromegaly patients came from a previously unrec-
ognized FIPA family. A 55-year-old female presented in 2012
for evaluation of zoledronic acid use in the treatment of bone
metastasis from breast cancer. Her family history was positive for
a sister with acromegaly and a benign nodular goiter, and her
mother had breast cancer. She was hypertensive (blood pressure:
150/90 mmHg) and had a body mass index of 30. On examination
the patient had a widened and thickened nose, macroglossia, prog-
nathism, enlarged hands and feet, oily thick skin and nodular goiter.
Initial investigations showed normal glucose, HbA1c , calcium, phos-
phate, PTH, thyroid function, cortisol, LH/FSH, and prolactin. Her
estradiol was < 25 pg/mL. Her IGF-1 was 855 ng/dL (3.6 × upper
limit of normal), while her nadir GH post-oral glucose tolerance test
was 3.1 ng/mL, which were diagnostic for acromegaly. A pituitary
MRI (Fig. 1), however, demonstrated no convincing evidence of an
adenoma or hyperplasia; an empty sella was seen. To rule out an
ectopic source leading to elevated GH, a CT scan of the thorax and
abdomen was performed, which was normal (serum GHRH test-
ing was not performed). Investigation of her thyroid enlargement
using ultrasound showed an isoechoic solid nodule in the right lobe
and isthmus of 3.3 × 1.8 cm in size. On FNA, a Bethesda VI papillary
thyroid carcinoma (PTC) was diagnosed.
The patient was treated with zoledronic acid for bone metas-
tases and continued with chemotherapy for her breast cancer. For
acromegaly, she was started on Lanreotide Autogel 90 mg/28 days,
which led to IGF-1 normalization in three months. In the interim,
the patient underwent a total thyroidectomy plus modified lateral
node dissection. Histopathology showed a PTC of 2.5 × 1.8 cm on
the right lobe and 6/6 metastatic nodes (Stage II TNM (8th edition)).
The pre-ablation thyroglobulin (TG) was 8.25 ng/mL and thyroglob-
ulin antibody (TGab) was undetectable. Following risk stratification
according to the ATA guidelines, she was considered as an interme-
diate recurrence risk. She underwent recombinant TSH-stimulated
radioactive iodine ablation (150 mCi). Given the FIPA diagnosis,
genetic testing for mutations or deletions in AIP and MEN1 were
performed, but no pathological genetic variants were identified in
DNA derived from blood or thyroid tumor tissue.
In early 2013 a follow-up ultrasound showed a medially located
node of 7.2 × 3.1 mm in size. Laboratory test results were as follows:
TSH: 0.19 ␮UI/mL, TG 3.6 ng/mL, TGab: negative (under LT4 ther-
apy). FNA of the node was performed which confirmed metastatic
3 4 5 6 7a
F F F F F
68 31 53 48 55
16 ND 8 33
PTC FC PTC PTC PTC
I I I I II
TT+RAI TT TT+RAI TT+RAI TT+CLND+RAI
Low Low Low Low Intermediate
70 Lost 60 24 24
Excellent Excellent Excellent Excellent Structurally incomplete
PTC on cytology; the TG in aspiration washout was 243 ng/ml. A
nodal level VI surgical dissection was performed and revealed a
lymph node (2 × 1.4 × 1.3 cm) that was positive for PTC and extra-
nodal extension to soft tissue. On dynamic stratification it was
considered that she had structural persistence/recurrent disease
with extra-nodal extension, giving a high-risk of recurrence [18].
No further surgical treatment of the metastatic PTC was performed,
as the patient experienced progression of her breast cancer and died
the following year.
Among the six sporadic acromegaly patients, all had excellent
responses to therapy and were disease-free in follow up for their
DTC. Apart from the FIPA patient described above, one other patient
died due to progression of breast cancer.
4. Discussion
Acromegaly is associated with multiple comorbidities among
which benign and malignant neoplasms may occur with increased
incidence [19,20]. Among the cancer types potentially associated
with acromegaly are colorectal, breast, and prostate cancer. In the
current cohort of patients with acromegaly we identified seven
that had coexistence of acromegaly and thyroid cancer (11.9%). In
addition to acromegaly and DTC, the patients presented other neo-
plasms: Hürthle cell adenoma, thyroid follicular adenoma, atypical
endometrial hyperplasia, colonic polyp, and breast carcinoma.
DTC arising from thyroid follicular epithelial cells accounts for
the vast majority of thyroid cancers. Papillary carcinoma com-
prises about 85% of cases. The basic goals of initial therapy for
patients with DTC are to improve overall and disease-specific sur-
vival, reduce the risk of persistent/recurrent disease and permit
accurate disease staging and risk stratification [5]. Reports on the
indication for thyroid US in acromegaly are controversial. Based
on our previous experience in thyroid nodules and carcinomas
[4], we have instituted prospective thyroid ultrasonograpy in all
newly diagnosed acromegaly patients. In this way we have prob-
ably detected early low-risk and microcarcinoma patients, such as
patients 2 and 6. In our study DTC had a low recurrence rate in
our sporadic acromegaly patients (6/6 were disease-free), and they
had an excellent response to therapy for DTC (6/6). All had active
acromegaly in spite of multimodal treatment, which suggests that
in this cohort at least, acromegaly did not confer a more aggressive
thyroid cancer phenotype. In keeping with the results of Mian et al.,
female sex predominated in patients with DTC and acromegaly [14].
Familial isolated pituitary adenoma syndrome (FIPA) is the most
common familial cause of acromegaly [6]. AIP mutations lead to a
well characterized form of aggressive acromegaly at an early age
and the most common forms of presentation are as part of FIPA fam-
ilies or with pituitary gigantism [10]. The FIPA patient we describe
here had a clinical profile that was not typical of AIP-related disease,
in that she was middle aged rather than young at onset, did not have
 A. Rogozinski et al. / Annales d’Endocrinologie 81 (2020) 482–486
Fig. 1. Panel A. MRI at diagnosis showing absence of visible pituitary tumor tissue in an AIP mutation negative FIPA patient with biochemical acromegaly. Panel B. High
power magnification image illustrating AIP immunohistochemistry of papillary thyroid cancer in an AIP mutation negative individual. Higher AIP staining intensity is seen
in tumoral as compared with adjacent thyroid tissue.
a macroadenoma and had an excellent response to somatostatin
analogs. Indeed, about 50% of FIPA kindreds with homogeneous
acromegaly are negative for AIP mutations [11], and the recently-
described involvement of GPR101 duplications in FIPA cohorts with
pituitary gigantism are exceptionally rare and did not fit her clinical
picture [21].
GH-producing adenomas of the pituitary gland are usually
macroadenomas [3]. Co-existence of acromegaly and empty sella is
an unusual finding that has been reported in individual cases and in
small series. For example, Molitch et al. reported two patients with
an empty sella who presented with active acromegaly [22]. Infre-
quently there are cases of ectopic acromegaly due to GH-secreting
tumors located outside of the pituitary fossa; the majority of these
lesions are located in the sphenoid sinus [23]. In our case, there
was an empty sella without an evident adenoma. No evidence of
an ectopic GH source or a GHRH secreting carcinoma was seen on
imaging. Although an empty sella could indicate a previous event
such as apoplexy, in this case there was no evidence of this from the
patient’s history. Trans-sphenoidal surgery is the preferred initial
treatment for acromegaly, with the use of somatostatin analogues
for patients for whom surgery is declined or inadvisable. In this case
the lack of a visible tumor meant that we elected to treat our patient
initially with a somatostatin analog as we felt it would have a bet-
ter outcome [24]. The patient responded well to medical treatment
and had IGF-1 control within three months.
Germline AIP mutations lead to one of the best characterized
subtypes of acromegaly, one that occurs often in a FIPA setting, and
leads to young-onset macroadenomas that are difficult to treat [10].
None of the sporadic acromegaly patients with DTC had criteria sug-
gestive of AIP related acromegaly and were not tested genetically, as
is generally recommended [6,25]. The patient with FIPA also proved
to be AIP mutation/deletion negative, as is the case in about 50% of
FIPA kindreds with homogeneous acromegaly [9]. Due to recent
485
findings of thyroid cancer in individual AIP mutation carriers, we
were interested in studying AIP staining characteristics, as there
are few data available in the literature. Like in the Italian series
of Mian et al. [14], among our acromegaly cohort DTC was over-
whelmingly seen in women. We found no relationship between
acromegaly disease activity and DTC responsiveness and the AIP
immunohistochemical staining results showed similar intensity to
tissue-control normal samples, again confirming the results of Mian
et al. [14]. Unlike in recent work from our group in AIP mutation
affected patients, there was no loss of AIP staining was seen on
immunohistochemistry of thyroid cancer samples from acromegaly
patients [13,26]. In summary, in our cohort of sporadic and familial
forms of acromegaly with concurrent DTC, addition of AIP sequenc-
ing and AIP immunohistochemistry did not add clinically useful
information on the etiology of thyroid cancer.
Credit author statement
Amelia Rogozinski (Conceptualization, formal analysis; investi-
gation, writing-original draft, writing-review and editing); Adrian
F. Daly (validation, formal analysis, investigation, writing-original
draft, writing-review and editing), Adriana Reyes (Data curation,
investigation; writing-review and editing); Alejandra Furioso (Data
curation, investigation, writing-review and editing); Albert Beckers
(Supervision, resources, writing-review & editing); Alicia Lowen-
stein (Supervision, resources, writing-review & editing).
Funding sources
The work was supported by in part by grants from the FIRS
2016–2018 (CHU de Liège) and from the JABBS Foundation (to
Albert Beckers).
486 A. Rogozinski et al. / Annales d’Endocrinologie 81 (2020) 482–486
Disclosure of interest
The authors declare that they have no competing interest.
Acknowledgement
We thank Dr. Albert Thiry for discussions on AIP immunohisto-
chemistry.
References
[1] Melmed S. Medical progress: acromegaly. N Engl J Med 2006;355:2558–73,
http://dx.doi.org/10.1056/NEJMra062453.
[2] Colao A, Ferone D, Marzullo P, Lombardi G. Systemic complications of
acromegaly: epidemiology, pathogenesis, and management. Endocr Rev
2004;25:102–52.
[3] Petrossians P, Daly AF, Natchev E, Maione L, Blijdorp K, Sahnoun-Fathallah
M, et al. Acromegaly at diagnosis in 3173 patients from the Liège
Acromegaly Survey (LAS) Database. Endocr Relat Cancer 2017;24:505–18,
http://dx.doi.org/10.1530/ERC-17-0253.
[4] Rogozinski A, Furioso A, Glikman P, Junco M, Laudi R, Reyes A, et al. Nódu-
los tiroideanos na acromegalia. Arq Bras Endocrinol Metabol 2012;56:300–4,
http://dx.doi.org/10.1590/S0004-27302012000500004.
[5] Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ, Nikiforov YE, et al.
2015 American Thyroid Association Management Guidelines for adult patients
with thyroid nodules and differentiated thyroid cancer: the American Thyroid
Assciation Guidelines Task Force on Thyroid Nodules and Differentiated Thy-
roid Cancer. Thyroid 2016;26:1–133, http://dx.doi.org/10.1089/thy.2015.0020.
[6] Vandeva S, Daly AF, Petrossians P, Zacharieva S, Beckers A. Genetics in
endocrinology: somatic and germline mutations in the pathogenesis of pitu-
itary adenomas. Eur J Endocrinol 2019, http://dx.doi.org/10.1530/EJE-19-0602.
[7] Gadelha MR, Kasuki L, Korbonits M. The genetic background of acromegaly.
Pituitary 2017;20:10–21, http://dx.doi.org/10.1007/s11102-017-0789-7.
[8] Fajardo-montañana C, Daly AF, Riesgo-suárez P, Gómez-vela J, Tichomirowa
MA, Camara-gómez R, et al. AIP mutations in familial and sporadic pituitary
adenomas: local experience and review of the literature. Endocrinol y Nutr
2009;56:369–77.
[9] Daly AF, Vanbellinghen J-F, Sok KK, Jaffrain-Rea M-L, Naves LA, Guitelman MA,
et al. Aryl hydrocarbon receptor-interacting protein gene mutations in familial
isolated pituitary adenomas: analysis in 73 families. J Clin Endocrinol Metab
2007;92:1891–6, http://dx.doi.org/10.1210/jc.2006-2513.
[10] Daly AF, Beckers A. Familial isolated pituitary adenomas (FIPA) and mutations in
the aryl hydrocarbon receptor interacting protein (AIP) gene. Endocrinol Metab
Clin North Am 2015;44:19–25, http://dx.doi.org/10.1016/j.ecl.2014.10.002.
[11] Beckers A, Aaltonen LA, Daly AF, Karhu A. Familial isolated pituitary ade-
nomas (FIPA) and the pituitary adenoma predisposition due to mutations
in the aryl hydrocarbon receptor interacting protein (AIP) gene. Endocr Rev
2013;34:239–77, http://dx.doi.org/10.1210/er.2012-1013.
[12] Urbani C, Russo D, Raggi F, Lombardi M, Sardella C, Scattina I, et al. A novel
germline mutation in the aryl hydrocarbon receptor-interacting protein (AIP)
gene in an Italian family with gigantism. J Endocrinol Invest 2014;37:949–55,
http://dx.doi.org/10.1007/s40618-014-0123-4.
[13] Daly A, Rostomyan L, Betea D, Bonneville JF, Villa C, Pellegata NS, et al.
AIP-mutated acromegaly resistant to first-generation somatostatin analogs:
long-term control with pasireotide LAR in two patients. Endocr Connect
2019;8:367–77, http://dx.doi.org/10.1530/EC-19-0004.
[14] Mian C, Ceccato F, Barollo S, Watutantrige-Fernando S, Albiger N, Regazzo D,
et al. AHR over-expression in papillary thyroid carcinoma: clinical and molec-
ular assessments in a series of Italian acromegalic patients with a long-term
follow-up. PLoS One 2014:9, http://dx.doi.org/10.1371/journal.pone.0101560.
[15] Jaffrain-Rea M-LM-LL, Rotondi S, Turchi A, Occhi G, Barlier A, Peverelli E,
et al. Somatostatin analogues increase AIP expression in somatotropino-
mas, irrespective of Gsp mutations. Endocr Relat Cancer 2013;20:753–66,
http://dx.doi.org/10.1530/ERC-12-0322.
[16] Salvatori R, Daly AF, Quinones-Hinojosa A, Thiry A, Beckers A. A clinically
novel AIP mutation in a patient with a very large, apparently sporadic soma-
totrope adenoma. Endocrinol Diabetes Metab Case Rep 2014;2014:140048,
http://dx.doi.org/10.1530/EDM-14-0048.
[17] Daly AF, Cano DA, Venegas-Moreno E, Petrossians P, Dios E, Castermans
E, et al. Aip and men1 mutations and aip immunohistochemistry in pitu-
itary adenomas in a tertiary referral center. Endocr Connect 2019;8:338–48,
http://dx.doi.org/10.1530/EC-19-0027.
[18] Roh JL, Park JW, Jeong J, Gong G, Cho KJ, Choi SH, et al. Extranodal exten-
sion of lymph node metastasis as a prognostic indicator of recurrence
and survival in papillary thyroid carcinoma. J Surg Oncol 2017;116:450–8,
http://dx.doi.org/10.1002/jso.24713.
[19] Ruchala M, Wolinski K. Health-related complications of acromegaly—risk
of malignant neoplasms. Front Endocrinol (Lausanne) 2019;10:268,
http://dx.doi.org/10.3389/fendo.2019.00268.
[20] Boguszewski CL, da Boguszewski MCS. Growth hormone’s links to cancer.
Endocr Rev 2019;40:558–74, http://dx.doi.org/10.1210/er.2018-00166.
[21] Trivellin G, Daly AF, Faucz FR, Yuan B, Rostomyan L, Larco DO, et al. Gigantism
and acromegaly due to Xq26 microduplications and GPR101 mutation. N Engl
J Med 2014;371:2363–74, http://dx.doi.org/10.1056/NEJMoa1408028.
[22] Molitch ME, Hieshima GB, Marcovitz S, Jackson IM, Wolpert S. Coexisting pri-
mary empty sella syndrome and acromegally. Clin Endocrinol 1977;7:261–3.
[23] Ramírez C, Hernández-Ramirez LC, Espinosa-De-Los-Monteros AL, Franco JM,
Guinto G, Mercado M. Ectopic acromegaly due to a GH-secreting pituitary ade-
noma in the sphenoid sinus: a case report and review of the literature. BMC
Res Notes 2013:6, http://dx.doi.org/10.1186/1756-0500-6-411.
[24] Sasagawa Y, Hayashi Y, Tachibana O, Oishi M, Fukui I, Iizuka H, et al.
Clinical characteristics of acromegalic patients with empty sella and
their outcomes following transsphenoidal surgery. Pituitary 2017;20:403–8,
http://dx.doi.org/10.1007/s11102-017-0798-6.
[25] Caimari F, Hernández-Ramírez LC, Dang MN, Gabrovska P, Iacovazzo
D, Stals K, et al. Risk category system to identify pituitary ade-
noma patients with AIP mutations. J Med Genet 2018;55:254–60,
http://dx.doi.org/10.1136/jmedgenet-2017-104957.
[26] Coopmans EC, Muhammad A, Daly AF, de Herder WW, van Kemenade FJ, Beck-
ers A, et al. The role of AIP variants in pituitary adenomas and concomitant thy-
roid carcinomas in the Netherlands: a nationwide pathology registry (PALGA)
study. Endocrine 2020, http://dx.doi.org/10.1007/s12020-020-02303-7.
 Annales d’Endocrinologie 81 (2020) 487–492

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

Original article

Qualitative study of self-reported experiences of infertile women with

polycystic ovary syndrome through on-line discussion forums
Vécu des femmes infertiles atteintes d’un syndrome des ovaires polykystiques :
étude qualitative sur les forums de discussion en ligne
Magali Authier a,∗ , Caroline Normand a , Maëva Jego a , Bénédicte Gaborit b , Léon Boubli c ,
Blandine Courbiere c,d
a
Département de médecine générale, faculté des sciences médicales et paramédicales – Timone, Aix-Marseille Université, 27, boulevard Jean-Moulin, 13385
Marseille, France
b
Département d’endocrinologie, diabète et maladies métaboliques, AP-HM, Aix Marseille Université, INSERM, INRA, C2VN, Marseille, France
c
Pôle femmes-parents-enfants, centre clinico-biologique d’AMP – CECOS, AP-HM Hôpital Nord/Hôpital La Conception, 147, boulevard Baille, 13005
Marseille, France
d
Aix-Marseille Université, CNRS, IRD, IMBE, Avignon Université, Marseille, France

a r t i c l e i n f o a b s t r a c t

Keywords: Objectives. – To highlight the self-reported experiences and disease perceptions of infertile women with
PCOS polycystic ovary syndrome (PCOS).
Polycystic ovary syndrome Methods. – A qualitative study using an inductive method was conducted on infertile women with PCOS
Infertility
who shared their self-reported experiences on French-speaking on-line forums.
Experience
Results. – 785 comments by 211 women on 7 forums were analyzed. Women complained of late diagnosis
Forums
and lack of information regarding PCOS. PCOS and infertility showed negative psychological impact on
daily life. This impact appeared to be alleviated by the sharing of knowledge and experience enabled by
these forums.
Conclusion. – The self-reported experience of infertile women with PCOS is interesting for health practi-
tioners. The psychological impact of PCOS and perceptions of illness appear to be improved by sharing
experiences between women with PCOS, suggesting a beneficial support role of online discussion forums.
© 2020 Elsevier Masson SAS. All rights reserved.

r é s u m é

Mots clés : Objectifs. – Explorer l’expérience personnelle et les perceptions de leur maladie par des femmes infertiles
SOPK ayant un syndrome des ovaires polykystiques (SOPK).
Syndrome des ovaires polykystiques Méthodes. – Étude qualitative conduite chez des femmes infertiles, ayant un syndrome des ovaires
Infertilité polykystiques, partageant leur expérience personnelle sur des forums Internet francophones.
Vécu
Résultats. – À travers 7 forums en ligne, 785 commentaires provenant de 211 femmes ont été analysés. Les
Forums Internet
femmes se plaignaient d’un retard diagnostique et d’un manque d’informations concernant le syndrome
des ovaires polykystiques. Nous avons observé un impact psychologique négatif du SOPK et de l’infertilité
sur la vie quotidienne de ces femmes. Cet impact semblait être amélioré par les échanges de connaissances
et le partage d’expériences permis par les forums Internet.
Conclusion. – L’expérience personnelle des femmes infertiles atteintes de SOPK est intéressante pour les
professionnels de santé. L’impact psychologique du SOPK et les perceptions de la maladie semblent être
améliorés par le partage d’expérience entre les femmes atteintes de SOPK, suggérant un rôle de soutien
bénéfique des forums de discussion en ligne.
© 2020 Elsevier Masson SAS. Tous droits réservés.

∗ Auteur correspondant.
E-mail address: magaliauthier@gmail.com (M. Authier).

https://doi.org/10.1016/j.ando.2020.07.1110
0003-4266/© 2020 Elsevier Masson SAS. All rights reserved.
488 M. Authier et al. / Annales d’Endocrinologie 81 (2020) 487–492
1. Introduction
Polycystic ovary syndrome (PCOS) is a syndrome combining
dysovulation, clinical and/or biological hyperandrogenism, and/or
polycystic ovaries on ultrasound (2 out of 3 criteria affirm diagno-
sis, according to the Rotterdam criteria) [1–4]. PCOS is the most
common endocrine disorder in women of reproductive age, affect-
ing 5–10% [5,6]. The consequences of PCOS, such as obesity, acne,
hirsutism, infertility, and metabolic syndrome, have a significant
impact on the quality of life of women with PCOS [7–9]. Increased
cardiovascular risk, hypertension, type 2 diabetes mellitus, dyslip-
idaemia and somatic complaints such as hirsutism are often given
priority by physicians, while the very frequent psychological suf-
fering is often misunderstood [10,11].
The web is a place for free, open and anonymous exchange,
liberating people from psychological barriers such as inhibiting
expression or emotions that might occur face-to-face [12–14]. In
this context, the Internet is an accessible source of information and
is commonly used for health research. Web users search for med-
ical information in addition to that provided by their physicians
[15,16].
The objective of our study was to explore the self-reported expe-
riences and disease perceptions of infertile PCOS women through
Internet discussions forums.
2. Methodology
The criterion for inclusion was to be a woman suffering from
both PCOS and infertility. PCOS was a self-reported diagnosis by
women on the Internet forums. Women were considered infertile
according to these criteria:
• if the woman signified a failure to conceive after more than 12
months of regular unprotected sexual intercourse (WHO defini-
tion);
• if the woman, in the context of a pregnancy project, noted the use
of a cyclical treatment for inducing periods or if she underwent
an ovulation induction treatment;
• if the woman considered herself infertile, even in the absence of
the criteria above.
This research did not require prior consent from the ethics com-
mittee, as it was conducted exclusively on the basis of public,
anonymous and retrospectively collected data.
We conducted a qualitative study with an inductive thematic
content analysis inspired by the grounded theory developed by
Glaser and Strauss [17]. We chose to conduct qualitative research,
which aims to describe in-depth phenomena in the context of
everyday life, while quantitative research aims to quantify or com-
pare elements to identify general rules [18]. Contrary to most
studies, the bibliographical analysis was carried out after the data
analysis to avoid predetermined ideas during the analysis.
The first stage of the study was to analyse hundreds of com-
ments on French-speaking Internet forums concerning infertile
females with PCOS. The open-access French-speaking forums were
selected using the Google search engine using the following key-
words: “forum” or “opinion” and “PCOS”. The forums studied
were allodocteur, journal des femmes santé-médecine, doctissimo,
mamanpourlavie, magicmaman, pmafertilité and forum.fiv. The
search was deliberately large and did not contain keywords related
to infertility to reduce the risk of the exclusion of discussions by
the search engine, as the selection of relevant discussions and com-
ments were made by the author. The analysis of the comments led
to the emergence of a classification system of codes and categories.
Simultaneously, a second physician performed a double reading
(MA and CN), which made it possible to validate the codes found
in common to reduce the subjectivity of the study. Data satura-
tion was reached after approximately 600 comments. The analysis
of approximately 150 additional comments confirmed the absence
of the emergence of new codes. The transcription of the data was
carried out by the software N Capture® . Data analysis was per-
formed using the qualitative research software NVivo 12 Pro (QSR
International® , Victoria, Australia). The predominant themes were
exploited and clarified to form the content of this article. The orig-
inal quotes were in French and have been translated into English
by the author.
The second stage was to carry out a literature review to compare
the elements found with the bibliographical data.
3. Results
A total of 785 comments were analysed from 211 women on
seven Internet forums. Important psychological suffering with anx-
iety and emotional distress was observed, which appears to be
related to low knowledge of the disease.
3.1. Diagnostic delay and a syndrome that is hard to understand
The level of understanding of PCOS appeared to be low among
infertile Internet users with PCOS. A delay in diagnosis was noted
by many of the women. This is suggested to be the cause of having
multiple consultations and switching doctors at the beginning of
the search for an explanation of their problems.
Despite its frequency, the women were not aware of anyone
who had PCOS, except for those with a family history of PCOS. As
a result, a strong feeling of aloneness emerged from the women in
the analysis. These elements suggested that PCOS was a disease not
known by the general population.
The women requested explanations to understand their pathol-
ogy. However, the quantity and complexity of information given
during the medical consultation seemed difficult to integrate in the
beginning, compounded by the multiple orders for additional tests
on specific days of the cycle. Sometimes, in contrast, they felt like
they had no information about the disease when they came out of
the consultation.
A lack of understanding of mechanisms of the disease and the
mode of action of the drugs prescribed, including metformin, hin-
dered adherence and compliance to treatment. One of the patients’
reflexes was to consult the Internet in search for additional infor-
mation about their disease or the treatments. At first, they gave
the impression of addressing worrying explanations that they did
not fully understand and reported that those explanations were
a source of fear and panic. Some comments even conveyed false
information, such as the notion that PCOS could be due to an overly
strong desire to have a child, that it could be due to stress, that
it was a malformation, or that the disease was due to eggs stick-
ing around the ovary after ovulation. However, in a second round,
Internet users appreciated receiving further explanations about
their particular cases on medical websites or from other women.
They were very relieved to find people with similar experiences
on the forums, which reduced their feelings of loneliness. Some
of them expressed the wish to get together and to be able to meet
other Internet users to interact face to face. The sharing experiences
were highly appreciated, satisfying their desire for information and
reducing apprehension and uncertainty (Table 1).
3.2. Psychological impact related to PCOS
The most constant symptomatology appeared to be an anxi-
ety disorder and emotional distress. Internet users were initially
stressed after the diagnosis of PCOS. This fear seemed to be linked
 M. Authier et al. / Annales d’Endocrinologie 81 (2020) 487–492
Table 1
Excerpts about diagnostic delay and a syndrome hard to understand (translated from French).
Categories
Diagnostic delay
Isolation feeling
Lack of understanding
not only to the unknown and the lack of knowledge about this
disease but also to the information they gathered during the consul-
tation of the announcement or on the Internet, in particular, the fear
of not being able to have children. They were then worried about
further tests, such as pain during hysterosalpingography and the
side effects of ovulation induction treatments such as clomiphene
citrate. Each symptom was analysed and caused anxiety about its
meaning, with the women often imagining the worst: metrorrhagia
as a sign of miscarriage or ectopic pregnancy and abdominal pain as
a sign of ovarian hyperstimulation. The absence of symptoms also
seemed to be disturbing to them and meant in their minds that the
treatment had failed. For those who were able to achieve a preg-
nancy, anxiety continued with the fear of miscarriage or ectopic
pregnancy.
The long medical history and infertility of these women, the
failure of the treatments and the miscarriages were sources of dis-
couragement, disappointment and frustration. For some, the goal of
getting pregnant became a real obsession. All of this contributed to
guilt and low self-esteem associated with not being able to become
pregnant. Psychological distress was very frequently found, along
with sadness and crying, which went as far as the expression of
suicidal ideation for one woman.
In addition to the medical consultations, the women had to
undergo many tests (repeated ultrasounds and blood tests), which
required a significant amount of time in their everyday lives. In
addition, there was a feeling of the medicalization of their pri-
vate life and a loss of intimacy. Their intimacy was affected by
the scheduling of sexual intercourse, which, when combined with
the stress of their medical situation, led to harmful psychological
pressure (Table 2).
3.3. Difficulties experienced during the medical course
3.3.1. Difficulties specific to PCOS
The issue of weight was central among the difficulties felt by
PCOS infertile women. Guilt-blaming, pessimistic, and even mock-
ing remarks were reported, implying poor involvement of women
in their weight loss process. This sort of phrasing was very badly
experienced by the women, who felt very stressed and even more
discouraged after the medical consultations. In addition, the med-
ical history of these women, who were given different medical
opinions, often resulted in questioning the physician’s assessment.
Even the truthfulness of the diagnosis of PCOS was doubted because
of previous diagnostic uncertainty. All these elements seemed to
contribute to a real division between some physicians and women.
The physician had to be supportive, reassuring, confident and opti-
mistic. These points, normally obvious, did not seem to correspond
489
Example of verbatim Number of
verbatim
“In 2005, my husband and I decided to have a child...I stopped the pill 21
and 2 years later, still nothing. We went from physician to physician
and to a general practitioner, an endocrinologist, and several
gynaecologists, with the answer ‘Ma’am, you are too fat!!!’ Therefore,
[I was] in tears after each visit...until I was sent to THE gynaecologist,
who looks at me and says, ‘You must have PCOS’. I had an ultrasound
like 2 days later, showing 20 cysts on each ovary...”
“I felt alone with my husband in my misadventure, but as I read all 12
your stories, I realized that it [PCOS] affects a lot more people than I
thought.”
“What nobody’s been able to explain to me is where this dystrophy 72
comes from”;
“I went to a gynaecologist to treat an infection and she told me I had
PCOS with no further explanation.”
to the experiences of these women, pointing to possible areas for
improvement in the relationships between some physicians and
women with PCOS.
Additional tests must be performed at the beginning of the men-
strual cycle. A major problem then appeared that was not always
anticipated by the physicians: women with PCOS most often have
oligo-anovulation and therefore amenorrhea or spaniomenorrhea.
Several months often passed before these tests were administered
if no treatment was given to induce periods.
The proposal of a care strategy explaining the different suc-
cessive steps of the treatment that could be used allowed them
to project themselves and to have a life plan. Active participa-
tion motivated these women. The main issues in which they could
become involved were diet, lifestyle, and weight loss for those who
were overweight. Empowering them on the success of treatment
improved their involvement and decreased their feelings of passiv-
ity and expectation (Table 3).
3.3.2. Bad experiences of infertility not specific to women with
PCOS
“Waiting” was the main burden. Infertile women began wait-
ing for a spontaneous pregnancy. When they decided to seek help,
they often faced long delays in appointments, up to several months.
Once the disease was finally diagnosed, the right treatment was
put in place, menstruation and ovulation resumed, and a new wait
began with the hope of a positive pregnancy test. Every day was
experienced as endless.
Most of the time, Internet users were deep into the feeling of
uncertainty. It was impossible for them to project themselves into
the years to come, or even into the next months. The interpreta-
tion of their symptoms was difficult because of the similarity of
the signs of pregnancy with the side effects of ovulation induction
treatments.
Women needed support and encouragement. Their partners
were often less stressed, able to step back and able to be more
positive and optimistic, which could have a reassuring effect.
Conversely, some women suffered from their husband’s lack of
involvement and felt that he did not adhere to the care plan, which
increased their doubts and anxiety. In addition to the feeling of
misunderstanding by their relatives, the same women had great dif-
ficulty expressing themselves towards their family, most of whom
were not aware of their pathology.
They were also seeking reinsurance on the forums. Messages of
encouragement and positive testimonies from other women on the
web seemed to have a motivating and stimulating effect, providing
hope and comfort (Table 4).
490 M. Authier et al. / Annales d’Endocrinologie 81 (2020) 487–492

Table 2
Excerpts about psychological impact related to PCOS (translated from French).

Categories Example of verbatim Number of verbatim

Anxiety disorder “It’s scary I know, I was on the lookout for these pains, [...] in case of 92
ectopic pregnancy or miscarriage[...] but it’s normal, it’s ligament pain.
However, dear, it’s only the beginning, the time seems long and
frightening, you see, that’s what I was talking about, that we think that
once we have our++ we will be relieved, but NOT at all! Because
afterwards it’s the anguish of “what if it doesn’t hold”. [...] I’m
refraining from going to the emergency room on the pretext of pain so
I can get an ultrasound”;
“So I’m a little scared to gain weight with all these treatments... The life
with PCOS is not fun, but you have to deal with it...”
Emotional distress “For me, PCOS eats my life away”; 193
“I always try to hope, but it’s hard...I’m severely depressed”
Privacy implications “It is a real obstacle course, there are still exams to be passed [...], and 89
papers, and an unfailing availability for both of us”;
“PCOS and its symptoms, weight gain, moods, then the infernal spiral
of doctors, medically-assisted procreation, treatments...”

Table 3
Excerpts about difficulties specific to PCOS (translated from French).

Categories Example of verbatim Number of verbatim

Weight issue “One day, one of the 5 gynaecologists I met told me that if I did not lose 29
at least 30 kg (66 pounds) I would not be able to have a child”
“And then they came up with this sentence that makes me so angry:
Lose weight!”
Questioning physician’s assessment “She told me that I have PCOS despite my periods being regular, I want 50
your opinion”
Menstrual cycle “I had nothing [no period] for 9 months and they came back only 11
adjustment because I was given Duphaston [dydrogestrone], otherwise I might
still be waiting.
They like to torture us”
Women involvement “And in the meantime I give all I can about healthy living and sport 3
because it is all I can do”

Table 4
Excerpts about difficulties about infertility not specific to PCOS (translated from French).

Categories Example of verbatim Number of verbatim

Expectation and “PCOS is a mess, you never know if it worked or not. Furthermore, with 170
uncertainty the treatments we have, the side effects cast doubt on us... In short, it
is a puzzle.”
Family support “My wonderful husband supported me and cheered me up.” 32
Difficulty to express “My mother-in-law who would like too much to be grandmother but I 19
do not know how to tell her it is too hard”
Internet forum support “You are in the right place to come and talk about your problems, your 115
feelings, we are all here to support each other and it feels good. For my
part, I was lucky that the treatments worked but I stay on the forum to
support you and to tell you the rest, because for those who know me,
they have been there for me from the beginning”

4. Discussion
We observed a strong psychological impact among PCOS infer-
tile women who share their experience in discussion forums. We
pointed out their difficulty in understanding what PCOS is.
First, there could be a bias regarding certain inclusion criteria:
the self-reported nature of the diagnosis of PCOS and the criterion
of infertility reported by women, which is subjective.
The difficulty in interpreting the comments was the presence
of many common features to infertile women and PCOS women.
One limitation of this study is related to the specificity of the PCOS
population among infertile women and to potential significant bias
induced by our methodology. We relied on a qualitative approach,
considering here that improved understanding of complex issues
is more important than generalizability of results [18]. The women
consulting the Internet forums are probably unconvinced by their
management and are possibly the ones who understand the least
of the medical information received by their physician regarding
PCOS. For these reasons, they may not be representative of the
general population. However, we believe that these women who
need to share their self-experience in Internet forums deserve to
be considered.
4.1. Interest of online communities in chronic pathologies
The French population conducting health research on the web is
mainly composed of middle-aged women. The majority of women
who expressed themselves on the PCOS forums were the most
often unsatisfied or anxious, which constitutes a selection bias. In a
study of populations conducting medical research on the Internet,
Renahy et al. reported that patients were not satisfied with medical
consultations due to a lack of explanations for 75% of them and a
lack of listening for 65% of them. Patients’ recourse is therefore to
obtain medical information via the Internet. However, the reliabil-
ity of the information retrieved is highly variable, and patients often
have difficulty selecting relevant sources [15]. Forums are a way to
 M. Authier et al. / Annales d’Endocrinologie 81 (2020) 487–492
gather and share experience that could increase the understand-
ing of pathology, including women with PCOS [19]. However, the
answers to the questions that Internet users ask each other were
not always accurate, constituting a limitation to the improvement
in knowledge. First, the Internet could be a source of anxiety regard-
ing the information found [16]. Nevertheless, appropriation of their
illness benefited from the use of the Internet, where they could
learn passively, by reading information, and actively, by interact-
ing in online communities. Knowledge and understanding of PCOS
appeared to decrease women’s anxiety, increase their involvement
and compliance, and thus improve their quality of life [20]. On the
other hand, PCOS has never been the subject of national information
campaigns in France, suggesting that this disease is poorly known
by the general public, even though the cardiovascular morbidity of
PCOS is a real public health issue [21].
4.2. Listening and acknowledging difficulties
More than one-third of women with PCOS are reported to have
a depressive syndrome, with a five-fold increase in the risk of
depression in obese women compared to those of normal weight
[22,23]. Infertility alone is also a cause of suffering and stress for the
individual and the couple, with a significant increase in the preva-
lence of depressive syndrome compared to fertile women [9,24].
PCOS women also suffer physically from a poor body image linked
to the consequences of the disease, such as hirsutism and over-
weight. These two aspects have been taken into consideration, and
an assessment of the psychological state is systematically recom-
mended for women with PCOS [25–27]. Despite this, a study by
Ellis et al. showed that psychological assessment was not carried
out in practice, in contrast to the screening for somatic patholo-
gies such as diabetes and high blood pressure in the follow-up [11].
Comprehensive psychological therapies have shown good results
on psyche and procreation, suggesting the importance of improv-
ing psychological well-being in the success of the pregnancy project
[28].
Two studies analysed the link between psychological support
and Internet forums. The first study examined the social support
provided by online communities to Japanese mothers for child-
care. Miyata et al. showed that mothers who actively participated
in forums received more support, with an increase in self-esteem
and a decrease in the risk of depression with short- and long-term
benefits on their social relationships in real life [12]. Sharf et al.
examined an online community of women with breast cancer. They
showed that the information exchanged enabled women to make
more informed decisions and provided a support network that was
available at any time and from any place, unlike medical consulta-
tions [13,29].
The qualitative method does not allow comparison between two
populations. However, it might be interesting to explore forum
users with another cause of infertility or PCOS women with or
without the variable “forum user”.
4.3. Diagnostic difficulties and heterogeneity of medical practices
Initial diagnostic wandering regarding PCOS is a major concern
in the experience of women with PCOS. Gaps exist in this area, as
we observed in the analysis of the comments. In an online sur-
vey of 1385 women with PCOS, Gibson–Helm et al. examined their
satisfaction with the information received at diagnosis and their
concerns at the time of the survey. They showed similar results,
highlighting a diagnostic delay and an inadequacy of the informa-
tion given about PCOS reported by women. In particular, more than
a third of the women interviewed reported a delay of more than
[10] Bellver J, Rodríguez-Tabernero L, Robles A, Muñoz E, Martínez F, Landeras J,
two years, and at least three health professionals consulted before
the diagnosis of PCOS was established [30]. A retrospective study
491
conducted by March et al. was looking for criteria of PCOS through
interrogation, clinical examination and further tests in 728 women
from the general population, aged from 27 to 34 years. They showed
that, according to the Rotterdam criteria, 68% of women with PCOS
were undiagnosed [31]. The results of this study are probably con-
sistent but should be interpreted with caution because of the small
size and it monocentric nature. If the diagnosis of PCOS is not made
despite a medical complaint attributable to the disease (hirsutism,
cycle disorders, infertility), it seems legitimate to question the lack
of sufficient medical training on this pathology in France, as was
found in the United States by Hoyos et al. and Chemerinski et al.
[32,33].
The international consensus on the diagnosis and management
of PCOS, published in 2018, should be better known and widely
disseminated among health professionals, particularly general
practitioners, who should be the central element in the manage-
ment of women with PCOS [25–27]. All health professionals should
be aware of the impact of PCOS on sex life, body image and relation-
ships. Moreover, they have the mission to inform the women that
the metabolic syndrome will not disappear, resulting in difficulty
losing weight despite their efforts, and prepare them for a poten-
tial delay in getting pregnant. Better knowledge of this consensus
would make it possible to homogenize practices and correct the
divergence factor between physicians, which is a source of anxiety
and confusion for patients [34].
5. Conclusion
Infertile women with PCOS who consult Internet forums appear
to suffer from low understanding of the disease, affecting their
quality of life. The medical information provided to these women is
perceived as inadequate or insufficient. Moreover, anxiety and psy-
chological distress, which are very common in women with PCOS,
are frequently omitted by health practitioners, despite recent rec-
ommendations for their systematic research. The use of discussion
forums appears to be beneficial, representing a source of support
and experiential knowledge sharing.
Disclosure of interest
The authors declare that they have no competing interest.
References
[1] Rotterdam ESHRE/ASRM-Sponsored, PCOS., Consensus Workshop Group.
Revised 2003; consensus on diagnostic criteria and long-term health risks
related to polycystic ovary syndrome. Fertil Steril 2004;81:19–25.
[2] Abbott DH, Dumesic DA, Levine JE. Hyperandrogenic origins of polycystic
ovary syndrome – implications for pathophysiology and therapy. Expert Rev
Endocrinol Metab 2019;31:1–13.
[3] Sakumoto T, Tokunaga Y, Tanaka H, Nohara M, Motegi E, Shinkawa T, et al.
Insulin resistance/hyperinsulinemia and reproductive disorders in infertile
women. Reprod Med Biol 2010;9:185–90.
[4] Cardozo E, Pavone ME, Hirshfeld-Cytron JE. Metabolic syndrome and oocyte
quality. Trends Endocrinol Metab 2011;22:103–9.
[5] Bozdag G, Mumusoglu S, Zengin D, Karabulut E, Yildiz BO. The prevalence and
phenotypic features of polycystic ovary syndrome: a systematic review and
meta-analysis. Hum Reprod 2016;31:2841–55.
[6] Lizneva D, Suturina L, Walker W, Brakta S, Gavrilova-Jordan L, Azziz R. Cri-
teria, prevalence, and phenotypes of polycystic ovary syndrome. Fertil Steril
2016;106:6–15.
[7] Hahn S, Janssen OE, Tan S, Pleger K, Mann K, Schedlowski M, et al. Clinical and
psychological correlates of quality-of-life in polycystic ovary syndrome. Eur J
Endocrinol 2005;153:853–60.
[8] Cousineau TM, Domar AD. Psychological impact of infertility. Best Pract Res
Clin Obstet Gynaecol 2007;21:293–308.
[9] Deeks AA, Gibson-Helm ME, Teede HJ. Anxiety and depression in polycystic
ovary syndrome: a comprehensive investigation. Fertil Steril 2010;93:2421–3.
et al. Polycystic ovary syndrome throughout a woman’s life. J Assist Reprod
Genet 2018;35:25–39.
492 M. Authier et al. / Annales d’Endocrinologie 81 (2020) 487–492
[11] Ellis E, Gibson-Helm M, Boyle JA. Polycystic ovary syndrome in Central
Australia: diagnosis and screening of cardiometabolic risk and emotional well-
being. Aust J Gen Pract 2018;4:227–32.
[12] Miyata K. Social support for Japanese mothers online and offline. In: Well-
man B, Haythornthwaite C, editors. The Internet in everyday life. Oxford, UK:
Blackwell; 2002. p. 520–48.
[13] Sharf BF. Communicating breast cancer on-line: support and empowerment on
the Internet. Women Health 1997;26:65–84.
[14] Kramish Campbell M, Meier A, Carr C, Enga Z, James AS, Reedy J, et al.
Health behavior changes after colon cancer: a comparison of findings from
face-to-face and on-line focus groups. Fam Community Health 2001;24:
88–103.
[15] Renahy E, Parizot I, Chauvin P. Determinants of the frequency of online
health information seeking: results of a web-based survey conducted in
France in 2007. Inform Health Soc Care 2010;35:25–39 [For further details, see
https://www.inserm.fr/sites/default/files/2017-11/Inserm RapportThematique
EnqueteHabitudeRechercheInformationsSanteInternet 2007.pdf].
[16] Beck F, Richard J-B, Nguyen-Thanh V, Montagni I, Parizot I, Renahy E. Use of the
Internet as a health information resource among french young adults: results
from a nationally representative survey. J Med Internet Res 2014;16:e128.
[17] Glaser B, Strauss A. The discovery of grounded theory: strategies for qualitative
research. Chicago: Aldine; 1967.
[18] Im E-O, Chee W. Practical guidelines for qualitative research using online
forums. Comput Inform Nurs 2012;30:604–11.
[19] Holbrey S, Coulson NS. A qualitative investigation of the impact of peer to
peer online support for women living with polycystic ovary syndrome. BMC
Women’s Health 2013;13:51.
[20] Drentea P, Moren-Cross JL. Social capital and social support on the web: the
case of an Internet mother site. Sociol Health Illn 2005;27:920–43.
[21] Teede H, Deeks A, Moran L. Polycystic ovary syndrome: a complex condition
with psychological reproductive and metabolic manifestations that impacts on
health across the lifespan. BMC Med 2010;8:41.
[22] Greenwood EA, Pasch LA, Shinkai K, Cedars MI, Huddleston HG. Clinical course
of depression symptoms and predictors of enduring depression risk in women
with polycystic ovary syndrome: results of a longitudinal study. Fertil Steril
2018;111:147–56.
[23] Annagür BB, Kerimoglu ÖS, Tazegül A, Gündüz Ş, Gençoglu BB. Psychiatric
comorbidity in women with polycystic ovary syndrome. J Obstet Gynaecol Res
2015;41:1229–33.
[24] Cwikel J, Gidron Y, Sheiner E. Psychological interactions with infertility among
women. Eur J Obstet Gynecol Reprod Biol 2004;117:126–31.
[25] Teede HJ, Misso ML, Costello MF, Dokras A, Laven J, Moran L, et al. Recommen-
dations from the international evidence-based guideline for the assessment
and management of polycystic ovary syndrome. Fertil Steril 2018;110:364–79.
[26] Teede HJ, Misso ML, Boyle JA, Garad RM, McAllister V, Downes L, et al.
Translation and implementation of the Australian-led PCOS guideline: clini-
cal summary and translation resources from the International Evidence-based
Guideline for the Assessment and Management of Polycystic Ovary Syndrome.
Med J Aust 2018;209:S3–8.
[27] Teede HJ, Misso ML, Costello MF, Dokras A, Laven J, Moran L, et al. Erra-
tum. Recommendations from the international evidence-based guideline for
the assessment and management of polycystic ovary syndrome. Hum Reprod
2018;34:38.
[28] ZareMobini F, Kazemi A, Farajzadegan Z. A comprehensive mental health care
program for women with polycystic ovary syndrome: protocol for a mixed
methods study. Reprod Health 2018;15:46.
[29] Im E-O, Lee BI, Chee W, Dormire S, Brown A. A national multiethnic online
forum study on menopausal symptom experience. Nurs Res 2010;59:26–33.
[30] Gibson-Helm M, Teede H, Dunaif A, Dokras A. Delayed diagnosis and a lack of
information associated with dissatisfaction in women with polycystic ovary
syndrome. J Clin Endocrinol Metab 2017;102:604–12.
[31] March WA, Moore VM, Willson KJ, Phillips DIW, Norman RJ, Davies MJ. The
prevalence of polycystic ovary syndrome in a community sample assessed
under contrasting diagnostic criteria. Hum Reprod 2010;25:544–51.
[32] Hoyos LR, Putra M, Armstrong AA, Cheng CY, Riestenberg CK, Schooler TA,
et al. Measures of patient dissatisfaction with health care in polycystic ovary
syndrome: retrospective analysis. J Med Internet Res 2020;22:e16541.
[33] Chemerinski A, Cooney L, Shah D, Butts S, Gibson-Helm M, Dokras A. Knowledge
of PCOS in physicians-in-training: identifying gaps and educational opportu-
nities. Gynecol Endocrinol 2020;6:1–6.
[34] Lua ACY, How CH, King TFJ. Managing polycystic ovary syndrome in primary
care. Singapore Med J 2018;59:567–71.
 Annales d’Endocrinologie 81 (2020) 493–499

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

Original article

Impact of aerobic training with and without whole-body vibration

training on metabolic features and quality of life in non-alcoholic fatty
liver disease patients夽,夽夽
Impact de l’entraînement aérobie avec et sans entraînement aux vibrations du
corps entier sur les caractéristiques métaboliques et la qualité de vie chez les
patients atteints de stéatose hépatique non alcoolique
Tülay Çevik Saldiran a,∗ , Fatma Karantay Mutluay b , İlker Yağci c , Yusuf Yilmaz d
a
Department of physiotherapy and rehabilitation, faculty of health science, Bitlis Eren University, Ahmet-Eren boulevard, Rahva street, 13000 Bitlis, Turkey
b
Department of physiotherapy and rehabilitation, Istanbul Medipol University, Istanbul, Turkey
c
Department of physical therapy and rehabilitation medicine, Marmara University, Istanbul, Istanbul, Turkey
d
Department of gastroenterology medicine, Marmara University, Istanbul, Istanbul, Turkey

a r t i c l e i n f o a b s t r a c t

Keywords: The present study examined the effectiveness of adding exercises with whole-body vibration (WBV)
Acceleration to aerobic training in terms of metabolic features and quality of life. Patients with non-alcoholic fatty
Exercise liver disease (NAFLD), confirmed on imaging, underwent an 8-week individualized exercise program
Metabolic features
randomized between aerobic training with and without WBV. Training was performed at 60–80% heart
Non-alcoholic fatty liver disease
Well-being
rate workload for 165 min/week. The WBV amplitude was 2–4 mm and the training frequency was 30 Hz,
for 15 min. Assessments were carried out on surrogate scores of steatosis and fibrosis including transient
elastography (FibroScan), metabolic features (biochemical analysis) and quality of life (SF-36). Insulin
resistance was markedly reduced (−2.36; 95% CI: −4.96 to −0.24; P: 0.049) in aerobic training with WBV.
The decrease in serum aspartate transaminase was significantly greater in aerobic training without WBV
(−14.81; 95% CI: −23.36 to −6.25; P: 0.029). There were no significant differences between groups for
the other metabolic features (P < 0.05). All quality of life well-being domains improved in both groups
(P < 0.05). Given this reduction in insulin resistance, WBV can usefully be added to aerobic training.
However, WBV did not provide further benefits in improving metabolic properties or quality of life.
© 2020 Elsevier Masson SAS. All rights reserved.

r é s u m é

Mots clés : La présente étude a examiné l’efficacité de l’ajout d’exercices avec vibrations globales du corps (WBV)
Accélération à l’entraînement aérobie en termes de caractéristiques métaboliques et de qualité de la vie. Les
Bien-être patients présentant une stéatose hépatique non alcoolique (NAFLD), confirmée en imagerie médi-
Caractéristiques métaboliques cale et par les données cliniques, ont subi un programme d’exercice, individualisé de 8 semaines,
Exercice
randomisé entre un entraînement aérobie avec ou sans WBV. L’entraînement a été effectué avec
Stéatose hépatique non alcoolique
une charge de travail correspondant à une fréquence cardiaque de 60 à 80 % pendant 165 min par
semaine. L”amplitude WBV était de 2 à 4 mm avec une fréquence de 30 Hz, pendant 15 min. Des éval-
uations ont été effectuées sur les scores de substitution de la stéatose et de la fibrose, y compris
l’élastographie transitoire (FibroScan), les caractéristiques métaboliques (analyse biochimique) et la
qualité de vie (SF-36). La résistance à l’insuline a été nettement réduite (IC −2,36 : −4,96 à −0,24 ;
p : 0,049) lors de l’entraînement aérobie avec le groupe WBV. La diminution des transaminases aspar-
tiques sériques était significativement plus élevée dans l’entraînement aérobie sans WBV (IC −14,81 :

夽 The study was conducted at Marmara University.

夽夽 Trial registration number: NCT03461562.
∗ Corresponding author.
E-mail addresses: tlyfztcvk@gmail.com, tc.saldiran@beu.ed.tr (T. Çevik Saldiran).

https://doi.org/10.1016/j.ando.2020.05.003
0003-4266/© 2020 Elsevier Masson SAS. All rights reserved.
494 T. Çevik Saldiran et al. / Annales d’Endocrinologie 81 (2020) 493–499
−23,36 à −6,25 ; p : 0,029). Il n’y avait pas de différence entre les groupes dans les autres caractéristiques
métaboliques (p < 0,05). Tous les domaines du « bien-être » et de la qualité de vie se sont améliorés
dans les deux groupes (p < 0,05). Compte tenu de la réduction de la résistance à l’insuline, le WBV est
une option additionnelle à l’entraînement aérobique. Cependant, la WBV n’a pas apporté d’avantages
supplémentaires pour améliorer les propriétés métaboliques etla qualité de la vie.
1. Introduction
Non-alcoholic fatty liver disease (NAFLD) is a major cause of
chronic liver disease worldwide. Insulin resistance is the main
factor responsible for the development of steatosis in hepato-
cytes in NAFLD [1]. It has been shown in previous studies that
insulin resistance management can be provided by aerobic and/or
strengthening exercise training [2–4]. However, there is no clarity
about the most effective exercise training in NAFLD. It was reported
that the majority of the patients could not attend the exercise pro-
grams regularly without complaint of early fatigue [5,6].
The whole-body vibration (WBV) has been shown to be the sub-
ject of research in recent years with the responses on the endocrine
system [6,7]. In these studies, it has been shown that the immune
modulator response revealed with WBV exercises in a short-term
training period [4,6,7]. The WBV has been reported to selectively
stimulate larger alpha motor neurons in muscle activation to pro-
duce more power in less time with less energy [8,9].
Abnormal changes in metabolic properties in the biochemi-
cal analysis are the main factors leading individuals with NAFLD
to further examination. Although these changes are not always
parallel with the steatosis level and disease progression, it has
been reported that they should be questioned in individual follow-
up [10,11]. Considering the multifactorial pathogenesis of NAFLD,
it was mentioned that the quality of life evaluations should be
included [12]. It is important to note that impairment of quality
of life in patients with NAFLD was less pronounced in the literature
although current guidelines state that the lifestyle change created
by exercise programs will have positive results on the quality of life
[13].
In this direction, the hypothesis of our research was that the
WBV exercises, which was added to the aerobic training, could
improve the metabolic features and quality of life of patients with
NAFLD with perceived effort reduction and immunomodulatory
effects. In this way, it was sought to answer the question of whether
the WBV addition to aerobic training programs in NAFLD patients
would have more positive effects on metabolic properties and qual-
ity of life. The aim of the study was to investigate the effects of two
different exercise approaches on metabolic features and quality of
life of patients with NAFLD.
2. Methods
2.1. Study design
The study was designed as a randomized clinical research with
two active training groups. Patients with NAFLD were enrolled
between March 2018 and August 2018 after written informed
consent. The trial was registered (ClinicalTrials.gov Identifier:
NCT03461562) and the research protocol conforms to the eth-
ical guidelines of the Declaration of Helsinki as reflected in a
priori approval by the institution’s human research committee
(Nr: 10840098-604.01.01-E.42834). Two active treatment groups
were taken to the training programs in the physiotherapy depart-
ment. The surrogate scores of liver stiffness (LSM) which could
© 2020 Elsevier Masson SAS. Tous droits réservés.
be in line with stage 4 fibrosis/cirrhosis ≥ 12.5 kPa [14], uncon-
trolled diabetes history (HbA1c > 9%), uncontrolled hypertension
[15], presence of a disease-preventing administration to aerobic
exercise or WBV training; cardiovascular, neurological, orthope-
dic disease history were exclusion criteria. NAFLD diagnosis was
made by detecting liver steatosis on ultrasound (US), excluding sec-
ondary causes for liver steatosis (significant alcohol consumption,
liver diseases which may cause steatosis and drugs) and exclu-
sion of other chronic liver diseases which may accompany. Patients
who were between the ages of 18 and 65 years have surrogate
scores of steatosis (CAP) ≥ 238 dB/m [14] and daily alcohol con-
sumption < 20 g were included in the study. Seventy-six patients
with NAFLD were evaluated. Study flow was summarized in the
supplemental file. Block randomization was made by forming 10
random numbers between 1 and 2 using the MedCalc 11.5.1 pack-
age program [16]. The 32 patients were randomly classified into
two groups. The physiotherapy assistant generated the random
allocation sequence, enrolled participants, and assigned partici-
pants to interventions.
2.2. Training procedures
2.2.1. Exercise tolerance test
The gold standard in evaluating the physiologic adaptation
occurring with aerobic training is maximal oxygen consumption,
which is a marker of aerobic capacity [17]. The oxygen consump-
tion was assessed by pedaling to exhaustion on an electronically
braked cycle ergometer using the exercise tolerance test (ETT) sys-
tem (Opticare Software, Ergoline Ergoselect 2 Model 600) [18]. The
test protocol was programmed with a preliminary workload as 10
watts (W) and workload per minute continued for 12 min with
a 10 W progressive increase per minute and continuous gradient
protocol at 60 rpm constant pedal speed. The test was ended when
the pedal speed fell under 40 rpm [19]. The maximal oxygen con-
sumption (VO2 max ), maximal power output (W), and heart rate
responses recorded. Patients with absolute and relative contraindi-
cation for the protocol did not take the test. Based on the test ending
criteria (chest pain, resembling angina, confusion, dizziness, intol-
erable dyspnea, leg cramps or extreme muscle fatigue) and other
undesired clinical effects, the test was ended [20]. Exercise toler-
ance test was applied before and at the end of the 8-week training
period.
2.2.2. Aerobic exercise training with whole-body vibration
Aerobic exercise program was performed on a bicycle ergome-
ter, including 5 min not overloaded, active training with 30 min and
not overloaded for 5 min as a cooling period. The target Heart Rate
(HR) for exercise training was 60–80% of the patients’ HR reserve
and was calculated as (0.6 and 0.8 × [max HR − resting HR] + resting
HR) based on the HR response from the baseline exercise tolerance
test. Speed or grade was adjusted as needed to maintain the exer-
cising of HR within the target range. Weekly training workload was
increased by 5%. HR monitors were used for the standardization of
exercise intensity. According to the ETT results, patients exercised
for 40 min per session at 60–80% of heart rate reserve.
 T. Çevik Saldiran et al. / Annales d’Endocrinologie 81 (2020) 493–499 495

Table 1 2.3.2. Biochemical assessment

The list of exercises with/without whole-body vibration.
Biochemical analyses were made in an internationally accred-
Exercises Repetition Time ited laboratory, using standard methods. The evaluations were
(Sec) made by a blinded researcher. The venous blood sample was
Lateral step Right-left 60 taken at the morning after fasting at least 10 hrs for bio-
Rest 60 chemical analysis including alanine aminotransferase (ALT),
Squat 10 45 aspartate aminotransferase (AST), gamma-glutamyl transferase
Rest 60
(GGT), alkaline phosphatase (ALP), total cholesterol (TC), high-
Calves 20 60
Rest 60 density lipoprotein (HDL-cholesterol), low-density lipoprotein
Lunge Right/left 60 (LDL-cholesterol) triglyceride (TG) containing lipid profile, total
Rest 60 bilirubin (BIL-T), direct bilirubin (BIL-D), glycated hemoglobin
Semi-squat 10 45
(HbA1c), fasting insulin and glucose. Homeostatic Model Assess-
Rest 60
Front-raise 10 × 2 120 ment of Insulin Resistance (Homa-IR) was calculated with “fasting
Rest 60 glucose (mmol/L) × fasting insulin (mU/L)/22.5 formula” [23].
Biceps curl 10 × 2 120
Rest 60
Squat 10 45
Rest 60 2.3.3. Quality of life assessment
Step Right-left 30
The quality of life in patients with NAFLD was assessed with
Rest 60
Squat 10 45 standardized form of SF-36 (Short Form-Health Survey, Version
1). Particularly, in the adult population, the SF-36 questionnaire
sec: second.
was the most frequently used health-related quality of life tool.
It includes eight individual health-related quality of life domains
(physical functioning, physical role, emotional role, vitality, men-
The exercises with whole-body vibration (WBV) were per- tal health, social functioning, pain, general health). For calculation
formed on a vertical-sinusoidal vibration platform (Power Plate® of all these domains, patients’ responses (all on Likert scales of vari-
Pro5; PowerPlate, Amsterdam, Netherlands) for 15 min. The WBV ous sizes, from 2 to 6) were averaged and then transformed to range
exercises are explained in Table 1. In the WBV session, the vibra- from 0 to 100 with higher scores representing better health [13].
tion amplitude intensity was 2–4 mm and the training frequency The SF-36 questionnaire data were collected and recorded with
was 30 Hz. The patients with NAFLD were rested for 60 seconds patient notifications.
between the exercises. These values were stable for 8 weeks. All One patient, assigned to the aerobic training plus exercises with-
training sessions were carried out under supervision. out WBV group, withdrew early during the intervention period.
Thirty-one patients (12 males and 19 females) with NAFLD were
included in the study. Patients underwent an 8-week individual-
2.2.3. Aerobic exercise training without whole-body vibration
ized combined training program consisting of aerobic training plus
Aerobic exercise training was applied by following the same pro-
exercises with/without WBV for 55 min per day, 3 times a week.
gram as previously described. The standard exercises were applied
At the beginning and at the end of 8-week (one day after the last
on the same whole-body vibration platform after aerobic exercise
session), the FibroScan (CAP/LSM, VCTE) examinations carried out.
training with the device button in minoff position, without vibra-
The biochemical measurement (metabolic features) analyses were
tion application.
made and recorded within 48 h.
No adverse event was recorded during the study and a pulse
oximeter and oxygen support were kept ready all the time.

2.4. Statistical analysis

2.3. Outcome measures
2.3.1. Vibration controlled transient elastography examination
Transient elastography was performed using the Fibroscan 402
(Echosens) by a blinded specialist according to a standardised pro-
tocol [21]. Patients were informed that they should be fasting for
at least three hours before coming for evaluation. The patient was
laid on the examination table in the supine position and the arm
was abducted after placing the right hand under the head. Ultra-
sonic shear waves were transferred to the liver from the right-hand
side using M (medium, 25–65 mm, 3.5 Mhz) or XL (extra-large,
35–75 mm, 2.5 Mhz) probe at a certain speed and volume. Probe
selection was made based on the automatic suggestion of the
device. Using probes, advancement-expansion speed and duration
of waves on the tissue were shown on a graph with the software
program. The operation was continued until at least 10 valid mea-
surements were acquired in a certain area from every participant
[22]. Measurements were done with the M probe and transition
to the XL probe if no appropriate reading was obtained. Cases
with ≥ 60% success rate and an IQR/Median < 30% were considered
valid [21]. Hepatic steatosis (controlled attenuation parameter;
CAP) and stiffness (liver stiffness measurement; LSM) surrogate
scores recorded.
Continuous data were tested for normality using the
Shapiro–Wilk test. Comparisons of baseline variables were made
using independent sample t tests. Between-group comparisons
were made using Anova. Within-group changes were assessed by
paired-sample t tests or the non-parametric alternative (Wilcoxon
signed-rank test) for non-normally distributed data. Statistical
power was based on the change in the hepatic steatosis level.
We selected a sample size of 14 to provide a statistical power
of 80% to detect a difference of 25.9% in hepatic steatosis [24]. A
per-protocol analysis was adopted [25]. All analyses were per-
formed using IBM SPSS Statistics software (version 19, NY, USA)
and data are expressed as mean ± SD or 95% confidence interval
(CI). P-values < 0.05 were considered statistically significant.
3. Results
The baseline characteristics of the patients were shown in
Table 2. Preliminary steatosis and stiffness levels of the patients
were over the norm value in both groups (P > 0.05). There was no
difference in the initial assessment between the groups (P > 0.05).
496 T. Çevik Saldiran et al. / Annales d’Endocrinologie 81 (2020) 493–499

Table 2
Baseline characteristics of patients with NAFLD.

Aerobic exercises with WBV group Aerobic exercises without WBV group P

Diagnosis duration (y) 6 (2–24) 4 (1.25–12) 0.599a

Age (y) 45.07 ± 9.11 43.75 ± 8.62 0.682a
Weight (kg) 89.03 ± 14.38 86.19 ± 13.72 0.579
Body mass index (kg/m2 ) 32.74 ± 4.78 33.17 ± 4.91 0.809
b
Waist circumference (cm) 109.60 ± 11.06 106.84 ± 8.17 0.434
Hip circumference (cm) 111.27 ± 9.67 111.66 ± 10.31 0.914
Sex/male 6 (40.0%) 6 (37.5%) 0.886a
Diabetus mellitus
Exist 7 (46.67%) 11 (68.75%) 0.986a
Hypertansion
Exist 4 (26.66%) 3 (18.75%) 0.984a
Drug use 0.347a
Exist 5 (33.3%) 8 (50.0%)
Alcohol useb < 20 g/day 0.266a
Exist 5 (33.3%) 2 (12.5%)
None 10 (66.7%) 13 (%81.3)
Quit 0 (0.0%) 1 (6.3%)
Smoking use 0.379a
Exist 4 (26.7%) 6 (37.5%)
None 6 (40.0%) 8 (50.0%)
Quit 5 (33.3%) 2 (12.5%)
FibroScan (VCTE)
CAP (100–400 dB/m) 321.13 ± 39.06 320.56 ± 33.48 0.965
LSM (2.5–70 kPa) 7.25 ± 2.43 6.02 ± 2.16 0.145
Metabolic features
AST (IU/L) 27.80 ± 9.03 37.56 ± 20.24 0.097
ALT (IU/L) 44.93 ± 16.27 59.19 ± 44.89 0.256
GGT (IU/L) 40.33 ± 19.26 43.69 ± 21.90 0.655
ALP (IU/L) 94.75 ± 53.42 85.75 ± 25.28 0.549
Cholesterol (mg/dL−1 ) 214.67 ± 44.56 207.44 ± 45.19 0.657
TG (mg/dL−1 ) 140.87 ± 65.05 146.88 ± 56.57 0.785
HDL (mg/dL−1 ) 49.73 ± 8.48 48.31 ± 9.57 0.666
LDL (mg/dL−1 ) 143.87 ± 47.27 129.76 ± 36.65 0.359
Ferritin (mg/dL−1 ) 54.80 ± 38.42 62.53 ± 40.92 0.592
Albumin (g/dL−1 ) 4.41 ± 0.33 4.36 ± 0.32 0.629
BIL-T (mg/dL−1 ) 0.63 ± 0.18 0.66 ± 0.24 0.763
BIL-D (mg/dL−1 ) 0.115 ± 0.046 0.119 ± 0.050 0.815
Homa-IR 6.42 ± 5.34 3.73 ± 1.54 0.092
HbA1c (%) 6.01 ± 0.97 6.05 ± 1.09 0.928
Fasting glucose (mg/dL−1 ) 114.73 ± 51.20 105.63 ± 30.21 0.574
Fasting insulin (mU−1 /mL) 20.81 ± 10.70 14.52 ± 5.89 0.050
Energy expenditure
VO2 max (mL/kg/min) 22.86 ± 7.17 19.74 ± 3.21 0.133
MaxPower (W) 113.87 ± 21.65 133.13 ± 54.74 0.214
Heart rate (beats/min) 84.00 ± 10.70 84.81 ± 10.28 0.831

Data presented as mean ± standard deviation or frequency (percentage). VO2 max : maximal oxygen uptake; MaxPower: maximal output power. P denotes a differences in
assessment according to t test.
a
P analyzed by X2 test. P < 0.05 were considered significant.
b
Waist/hip ratio > 0.90 increased risk for morbidity.

3.1. Comparison of the changes in the metabolic features and
energy expenditure
According to intra-group analysis results, the changes in sur-
rogate scores of steatosis (−29.27 dB/m, CI: −48.45 to −10.08),
stiffness (−1.55 kPa, CI: −2.61 to −0.48), total bilirubin (0.17 mg/dL,
CI: 0.07 to 0.26), and direct bilirubin (0.03 mg/dL, CI: 0.01 to 0.05)
was better in the aerobic training with the WBV group compared to
the aerobic training without the WBV group. The decrease in aspar-
tate transaminase results in the aerobic training without the WBV
group (−14.81 IU/L, CI: −23.36 to −6.25) was better versus aero-
bic training with the WBV group (−4.00 UI/L, CI: −6.82 to −1.18).
The improvement in insulin resistance (−2.36 CI: −4.96 to −0.24)
were found significant in the aerobic training with the WBV group.
According to intra-group comparison, both training programs were
effective on maximal oxygen consumption, maximal output power,
and heart rate (P < 0.05). The mean differences suggests that the
aerobic training with the WBV group presented a more signifi-
cant change in the maximal power output (maxPower: 49.33 95%
CI 33.51 to 65.15). The energy expenditure differences between
the two groups were quite small and there were no significant
interaction effect of treatment and time according to inter-group
comparisons (Table 3).
3.2. Comparison of the changes in the quality of life
Patients with NAFLD had significantly lower quality of life and
health utility scores than the general population according to base-
line results. After 8 weeks of training period, all quality of life
domains improved in both groups (P < 0.05). However, the general
health perception development was better in the aerobic training
without the WBV group (P: 0.040), while the improvement in emo-
tional role was higher in the WBV added arm (P: 0.044) (Table 4).
4. Discussion
The most important result of our study, was that a signifi-
cant recovery was observed in the insulin resistance in the WBV
added group. Liver enzymes significantly decreased over the 8-
week training period but the aspartate aminotransferase showed
 T. Çevik Saldiran et al. / Annales d’Endocrinologie 81 (2020) 493–499 497

Table 3
Metabolic and energy expenditure changes observed after 8 weeks of training in both groups.

Aerobic exercises with WBV group Aerobic exercises without WBV group P-value
time-by-group
interaction

: T2-T1 95% CI within-group : T2-T1 95% CI within-group

Fibro scan
CAP (100–400 dB/m) −29.27 ± 37.91a [−48.45 to −10.08] −4.13 ± 38.66 [−23.07 to 14.81] 0.078
LSM (2.5–70 kPa) −1.55 ± 2.10a [−2.61 to −0.48] −0.58 ± 2.03 [−1.57 to 0.41] 0.201
Metabolic features
AST (IU/L) −4.00 ± 5.58a [−6.82 to −1.18] −14.81 ± 17.45a [−23.36 to −6.25] 0.029b
ALT (IU/L) −13.13 ± 16.00a [−21.22 to −5.03] −24.69 ± 33.02a [−40.86 to −8.51] 0.230
GGT (IU/L) −6.73 ± 14.75 [−14.19 to 0.73] −9.88 ± 20.21 [−19.78 to 0.02] 0.627
ALP (IU/L) −7.41 ± 16.61 [−15.81, 0.99] −4.25 ± 20.32 [−14.20 to 5.70] 0.641
Cholesterol (mg/dL−1 ) −4.73 ± 24.13 [−16.94 to 7.48] 10.75 ± 27.69 [−2.81 to 24.31] 0.109
TG (mg/dL−1 ) −19.47 ± 57.86 [−48.75 to 9.81] −4.81 ± 43.00 [−25.87 to 16.25] 0.428
HDL (mg/dL−1 ) −1.27 ± 7.30 [−4.96 to 2.42] 1.75 ± 5.59 [−0.98 to 4.49] 0.205
LDL (mg/dL−1 ) −7.33 ± 38.66 [−26.89 to 12.23] 6.99 ± 24.83 [−5.17 to 19.15] 0.227
Ferritin (mg/dL−1 ) −4.31 ± 29.69 [−19.33 to 10.71] −9.45 ± 20.93 [−19.25 to 0.35] 0.579
Albumin (g/dL−1 ) −0.01 ± 0.32 [−0.17 to 0.15] 0.07 ± 0.25 [−0.05 to 0.19] 0.435
BIL-T (mg/dL−1 ) 0.17 ± 0.18a [0.07 to 0.26] 0.04 ± 0.22 [−0.07 to 0.15] 0.506
BIL-D (mg/dL−1 ) 0.033 ± 0.038a [0.01 to 0.05] 0.004 ± 0.043 [−0.01 to 0.02] 0.692
Homa-IR −2.36 ± 5.15a [−4.96 to −0.24] 0.52 ± 2.68 [−0.79 to 1.83] 0.049b
HbA1c (%) −0.01 ± 0.04 [−0.03 to 0.01] −0.02 ± 0.03 [−0.03 to −0.005] 0.523
Fasting glucose (mg/dL−1 ) −12.60 ± 42.83 [−34.27 to 9.07] 3.44 ± 20.54 [−6.62 to 13.50] 0.097
Fasting insulin (mU−1 /mL) −5.20 ± 10.52 [−10.52 to 0.12] 1.39 ± 9.92 [−3.47 to 6.25] 0.083
Energy expenditure
VO2 max (mL/kg/min) 6.57 ± 3.76 [4.56 to 8.57] 7.03 ± 5.77 [3.83 to 10.22] 0.800
MaxPower (W) 49.33 ± 28.57 [33.51 to 65.15] 45.69 ± 36.36 [26.31 to 65.06] 0.760
Heart rate (beats/min) −5.87 ± 3.05 [−7.56 to −4.18] −4.5 ± 0.23 [−4.62 to −37] 0.747

CAP: controlled attenuation parameter; LSM: liver stiffness measurement; AST: aspartate aminotransferase; ALT: alanine aminotransferase; GGT: gama glutamyl transferase;
ALP: alkaline phosphatase; TG: triacylglycerol; HDL: high-density lipoprotein; LDL: low-density lipoprotein; BIL-T: total bilirubin; BIL-D: direct bilirubin; Homa-IR: home-
ostatic model assessment insulin resistance; HbA1c: glycated hemoglobin; VO2 max : maximal oxygen uptake; MaxPower: maximal output power. : mean change from
baseline to end of trial. T1: initial evaluation; T2: last evaluation. Data presented as mean ± standard deviation with 95% CI.
a
P denotes an improvement in assessment according to pre and post-evaluation (within-group).
b
P Denotes a differences between groups. P < 0.05 were considered significant.

Table 4
Comparison of the quality of life differences between groups before and after trainings.

Quality of life Aerobic exercises with WBV group Aerobic exercises without WBV group P-value
time-by-group
interaction

T1 : T2-T1 95% CI within-group T1 :T2-T1 95% CI within-group

Physical functioning 80.33 ± 15.17 8.67 ± 6.27a

[5.49 to 11.84] 74.19 ± 20.13 13.31 ± 5.19a
[10.76 to 15.85] 0.418
Physical role 63.33 ± 36.43 23.34 ± 5.28a [20.67 to 26.01] 62.50 ± 43.78 18.75 ± 9.02a [14.33 to 23.17] 0.730
Emotional role 46.67 ± 41.46 37.80 ± 16.60a [29.39 to 46.20] 68.81 ± 39.40 4.13 ± 1.38a [3.45 to 4.80] 0.044b
Vitality 54.67 ± 16.85 12.00 ± 5.10a [9.42 to 14.58] 57.19 ± 20.81 15.00 ± 0.32a [14.84 to 15.16] 0.619
Mental health 53.87 ± 18.45 16.80 ± 4.70a [14.42 to 19.17] 65.00 ± 12.35 9.56 ± 2.07a [8.55 to 10.57] 0.243
Social functioning 66.80 ± 19.26 15.87 ± 4.47a [13.60 to 18.13] 61.94 ± 25.89 20.25 ± 3.58a [18.49 to 22.00] 0.570
Pain 72.07 ± 18.32 10.46 ± 2.49a [9.19 to 11.72] 64.38 ± 29.18 16.56 ± 11.53a [10.91 to 22.21] 0.512
General health 49.33 ± 18.41 7.67 ± 3.80a [5.74 to 9.59] 51.88 ± 19.40 22.81 ± 0.21a [22.70 to 22.91] 0.040b

: mean change from baseline to end of trial. T1: initial evaluation; T2: last evaluation. P < 0.05 were considered significant. Data presented as mean ± standard deviation
with 95% CI.
a
P denotes an improvement in assessment according to pre- and post-evaluation (within-group P-value).
b
P Denotes a differences between groups.

better improvement in the aerobic training without the WBV group.
Both training programs were effective on improving physiological
adaptation. Also, an enhancement in health-related quality of life
was reported in both training groups.
According to previous researches, an increase in insulin sen-
sitivity and a decrease in hepatic steatosis was shown with the
aerobic training applied alone, continued for 8 weeks [26], 10 weeks
[27], and 12 weeks [28]. Although, the aerobic training program we
applied in our study was similar to current exercise training sug-
gestions, it was observed that aerobic training without the WBV for
8 weeks was not adequate for the provision of improvement in sur-
rogate scores of steatosis of patients with NAFLD. The only research
examining the WBV effect in patients with NAFLD, was a pilot study
that was conducted by Oh, et al. The researchers reported an 8.7%
decrease in intra-hepatic fat content in agreement with our study
showing steatosis and stiffness improvements [29]. Considering the
cost-effectiveness, we should emphasize that this sort of WBV pro-
gram was not effective enough to be superior to aerobic training
without the WBV in an 8-week training period in surrogate scores
of steatosis and fibrosis in NAFLD patients. So that, it was considered
that more effective results could be acquired through the addition
of WBV in combined exercise programs with longer training and
follow-ups.
Insulin resistance is the factor responsible for the first impact
in liver steatosis [1]. It was reported that even the lowest level
of recoveries observed in insulin resistance were effective on
intra-hepatic fat regression [30]. Based on the previous research,
providing molecular level examination for the basic mechanism
498 T. Çevik Saldiran et al. / Annales d’Endocrinologie 81 (2020) 493–499
on insulin resistance for WBV, it was shown that oxidative stress
was suppressed with 12-week training [31]. Indeed, Shamsoddini
et al. reported that an eight-week aerobic training was effective on
insulin sensitivity in patients with NAFLD [32]. Although not at a
level to make a difference between groups, one of the main rea-
son for better recovery in surrogate scores of steatosis and fibrosis
in the WBV added group may have originated from the decrease
observed in insulin resistance and an increase in bilirubin lev-
els. Bilirubin is an endogenous antioxidant in the liver, playing a
role in disease pathogenesis concerning oxidative stress [33], and
the oxidative stress is one of the main factors in NAFLD devel-
opment and progression [34,35]. Therefore, immune modulator
effect formed by aerobic exercise on metabolic diseases through the
antioxidants [36] increases when the WBV is added and may stimu-
lates regression in steatosis and fibrosis level according to observed
improvements in bilirubin levels. When we evaluated the acquired
results with the energy expenditure changes and follow-up phase
of our study, insulin sensitivity was only improved by WBV, likely
caused by energy expenditure change as physical adaptation. While
both programs were effective in physical adaptation, when we eval-
uate their superiority to each other, the maximal power created by
NAFLD patients were higher in the aerobic training with the WBV
group, although it is not reflected in the difference between groups.
As a secondary cause, this difference might be explained by the
higher exercise intensity. In light of these results, the significant
recovery which was observed in surrogate scores of steatosis and
fibrosis in the WBV added group can be explained through these
mechanisms. The maximal power produced by NAFLD patients
allows verifying in what extent the WBV protocol seems to improve
the insulin sensitivity as a metabolic adaptation so that the WBV
might have fastened the cumulative effect of aerobic exercises. For
these reasons, we could say that the addition of WBV to aerobic
training programs, might induce effective responses that can be
obtained on insulin resistance in an 8-week period. But it should
be known that improvements were independent of liver enzymes
and the obvious results can be achieved in the long-term.
Stated in the guidelines, although liver enzymes are not always
parallel to steatosis level and disease progression, they should be
examined in individual follow-up [10,11]. Khosravi, et al. stated
that compared to other aminotransferases, alanine transaminase is
a biomarker that should be used more commonly to demonstrate
the changes in the liver [37]. As mentioned in the literature, the best
change was observed in alanine transaminase level among liver
enzymes with both combined exercise programs we applied. The
aerobic exercise training with and without the WBV were effec-
tive training methods on liver enzymes. However, the WBV did not
constitute a superior benefit on liver transaminases. So that both
training programs can be used considering serum transaminases
healing in NAFLD patients, but it is not always necessary to add
WBV.
Main glycemic profile constituents evaluated in NAFLD scanning
are fasting insulin, glucose and glycated hemoglobin. Bacchi, et al.
reported that there was no recovery in glycated hemoglobin with
aerobic and strengthening trainings [24]. Contrary to these stud-
ies, it was concluded by Katsagoni, et al. that the aerobic exercise
improved glycated hemoglobin by itself [38]. Similar results to our
study were reported by van der Heijden, et al. in a study examining
the efficiency of aerobic training in NAFLD patients [28]. In line with
this results, it was observed that different results were acquired
through different exercise applications on glycemic markers. An 8-
week aerobic training program with and without the WBV seems
inadequate for effective results in glycemic and also lipid profiles
in patients with NAFLD.
Assessing the quality of life of patients with NAFLD is impor-
tant to accurately estimate the exercise regimens effect on patients’
well-being. Golabi, et al. reported that NAFLD was associated with
impairment of health-related quality of life and they mention that
these impairments resulted in a reduction of patients’ ability to per-
form their daily activities [39]. Our initial data analysis showed that
NAFLD patients experienced significant impairment of their quality
of life. However, after 8 weeks of training programs, patients’ qual-
ity of life improved significantly. Overall, our study results indicate
that aerobic exercises with and without the WBV are particularly
effective in the management of patients with NAFLD with positive
effects on their mental and physical health status. Interestingly,
in the aerobic exercises without the WBV group, patients with
NAFLD had more improvement in their general health perception
than their emotional role domains. Although not exactly clear, the
WBV may provide more energy expenditure in patients with NAFLD
for the development of fatigue. Fatigue may be the reason for less
improvement in physical parameters of quality of life in the WBV
added group. As shown in previous research and both exercise pro-
grams we implemented were effective in improving quality of life
[29,40,41]. However, we should mention that the WBV does not
always need to be included in the combined programs.
The major limitation of this study is the absence of an appro-
priate control group. Another limitation of our study was our lack
of biopsy evaluation. This research was designed as a pilot study.
Studies should be done with a large-scale.
The application of aerobic training with the WBV was effective
in surrogate scores of steatosis and fibrosis, however this improve-
ments were not much better than aerobic training without the WBV
program. The addition of WBV to aerobic training has a positive
effect on insulin resistance in patients with NAFLD. Considering
the reducing insulin resistance the WBV can be added to aerobic
training. The aerobic training with or without the WBV is effective
in the healing of transaminases, however, the addition of WBV does
not constitute an advantage. In terms of liver enzymes, especially
in aspartate aminotransferase, the WBV should not be added to
aerobic training. Aerobic training with and without the WBV was
effective in physical adaptation and improving the quality of life in
many well-being domains however it was understood that there
was no need to add the WBV for better recovery of quality of life. At
this point we could mention that the long-term training programs
should be performed for better results.
Ethical statement
The research protocol conforms to the ethical guidelines of the
Declaration of Helsinki as reflected in a priori approval by the
institution’s human research committee (Nr: 10840098-604.01.01-
E.42834, Istanbul Medipol University, Research Ethic Committee).
Human and animal rights
The authors have not submitted a declaration concerning
human and animal experimentation.
Informed consent and patient details
The authors have not submitted a declaration concerning
informed consent and patient details.
Disclosure of interest
The authors declare that they have no competing interest.
Funding
The authors have not submitted a declaration concerning the
funding of their work.
 T. Çevik Saldiran et al. / Annales d’Endocrinologie 81 (2020) 493–499
Author contributions
Concept/idea/research design: STÇ and MFK.
Drafting the article: STÇ, YY and YI.
Data collection: STÇ, YY and YI.
Data analysis: MFK.
Revising: STÇ and MFK.
Final approval: STÇ, YY, MFK and YI.
Acknowledgement
Thanks all participants and researchers for their contribu-
tions. Written permission obtained from all patients. This research
received no specific grant from any funding agency in the public,
commercial, or not-for-profit sectors. This research has not yet been
presented or published elsewhere.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version at: https://doi.org/10.1016/j.ando.2020.05.003.
References
[1] Buzzetti E, Pinzani M, Tsochatzis EA. The multiple-hit pathogenesis of non-
alcoholic fatty liver disease (NAFLD). Metabolism 2016;65:1038–48.
[2] Guo R, Liong EC, So KF, Fung M-L, Tipoe GL. Beneficial mechanisms of aero-
bic exercise on hepatic lipid metabolism in non-alcoholic fatty liver disease.
Hepatobiliary Pancreat Dis Int 2015;14:139–44.
[3] Prior SJ, Blumenthal JB, Katzel LI, Goldberg AP, Ryan AS. Increased skeletal
muscle capillarization after aerobic exercise training and weight loss improves
insulin sensitivity in adults with IGT. Diabetes Care 2014;37:1469–75.
[4] Rodriguez-Miguelez P, Fernandez-Gonzalo R, Collado PS, Almar M, Martinez-
Florez S, de Paz JA, et al. Whole-body vibration improves the anti-inflammatory
status in elderly subjects through toll-like receptor 2 and 4 signalling pathways.
Mech Ageing Dev 2015;150:12–9.
[5] Gerber L, Otgonsuren M, Mishra A, Escheik C, Birerdinc A, Stepanova M,
et al. Non-alcoholic fatty liver disease (NAFLD) is associated with low level
of physical activity: a population-based study. Aliment Pharmacol Ther
2012;36:772–81.
[6] Cristi C, Collado PS, Márquez S, Garatachea N, Cuevas MJ. Whole-body vibration
training increases physical fitness measures without alteration of inflamma-
tory markers in older adults. Eur J Sport Sci 2014;14:611–9.
[7] Cardinale M, Leiper J, Erskine J, Milroy M, Bell S. The acute effects of different
whole body vibration amplitudes on the endocrine system of young healthy
men: a preliminary study. Clin Physiol Funct Imaging 2006;26:380–4.
[8] Cardinale M, Lim J. Electromyography activity of vastus lateralis muscle
during whole-body vibrations of different frequencies. J Strength Cond Res
2003;17:621–4.
[9] Torvinen S, Kannus P, SievaÈnen H, JaÈrvinen TA, Pasanen M, Kontulainen
S, et al. Effect of a vibration exposure on muscular performance and body
balance. Randomized cross-over study. Clin Physiol Funct Imaging 2002;22:
145–52.
[10] Chalasani N, Younossi Z, Lavine JE, Diehl AM, Brunt EM, Cusi K, et al. The diag-
nosis and management of non-alcoholic fatty liver disease: practice guideline
by the American Association for the Study of Liver Diseases, American Col-
lege of Gastroenterology, and the American Gastroenterological Association.
Hepatology 2012;55:2005–23.
[11] Liver EAftSot, Diabetes EAftSo. EASL-EASD-EASO Clinical Practice Guidelines for
the management of non-alcoholic fatty liver disease. Obes Facts 2016;9:65–90.
[12] Younossi ZM, Stepanova M, Henry L, Racila A, Lam B, Pham HT, et al. A disease-
specific quality of life instrument for non-alcoholic fatty liver disease and non-
alcoholic steatohepatitis: CLDQ-NAFLD. Liver Int 2017;37:1209–18.
[13] Sayiner M, Stepanova M, Pham H, Noor B, Walters M, Younossi ZM. Assessment
of health utilities and quality of life in patients with non-alcoholic fatty liver
disease. BMJ Open Gastroenterol 2016;3:e000106.
[14] Mueller S, Sandrin L. Liver stiffness: a novel parameter for the diagnosis of liver
disease. Hepatic Med Evid Res 2010;2:49.
[15] Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo Jr JL, et al.
The seventh report of the joint national committee on prevention, detection,
evaluation, and treatment of high blood pressure: the JNC 7 report. JAMA
2003;289:2560–71.
[16] Kanık EA, Taşdelen B, Erdoğan S. Klinik denemelerde randomizasyon; 2011.
499
[17] Arena R, Myers J, Williams MA, Gulati M, Kligfield P, Balady GJ, et al. Assessment
of functional capacity in clinical and research settings: a scientific statement
from the American Heart Association Committee on exercise, rehabilitation,
and prevention of the council on clinical cardiology and the council on cardio-
vascular nursing. Circulation 2007;116:329–43.
[18] King AJ, O’Hara JP, Arjomandkhah NC, Rowe J, Morrison DJ, Preston T, et al.
Liver and muscle glycogen oxidation and performance with dose variation of
glucose – fructose ingestion during prolonged (3 h) exercise. Eur J Appl Physiol
2019;119:1157–69.
[19] Ceylan E. Kardiyopulmoner egzersiz testleri. J Clin Exp Investig 2014;5:504–9.
[20] Cerşit S, Cerşit HP. Impact of cardiac rehabilitation on ventricular repolarization
indexes in patients with rheumatid arthritis. J Electrocardiol 2018;51:787–91.
[21] Mikolasevic I, Orlic L, Franjic N, Hauser G, Stimac D, Milic S. Transient elastog-
raphy (FibroScan® ) with controlled attenuation parameter in the assessment
of liver steatosis and fibrosis in patients with nonalcoholic fatty liver disease –
where do we stand? World J Gastroenterol 2016;22:7236.
[22] Vuppalanchi R, Siddiqui MS, Van Natta ML, Hallinan E, Brandman D, Kowd-
ley K, et al. Performance characteristics of vibration-controlled transient
elastography for evaluation of nonalcoholic fatty liver disease. Hepatology
2018;67:134–44.
[23] Muniyappa R, Lee S, Chen H, Quon MJ. Current approaches for assessing insulin
sensitivity and resistance in vivo: advantages, limitations, and appropriate
usage. Am J Physiol Endocrinol Metab 2008;294:E15–26.
[24] Bacchi E, Negri C, Targher G, Faccioli N, Lanza M, Zoppini G, et al. Both resis-
tance training and aerobic training reduce hepatic fat content in type 2 diabetic
subjects with nonalcoholic fatty liver disease (the RAED2 Randomized Trial).
Hepatology 2013;58:1287–95.
[25] Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiol-
ogy 1990;1:43–6.
[26] Davoodi M, Moosavi H, Nikbakht M. The effect of eight weeks selected aerobic
exercise on liver parenchyma and liver enzymes (AST, ALT) of fat liver patients.
J Shahrekord Univ Med Sci 2012;14:84–90.
[27] Chen H, Ma J, Lu B, Ma XL. The effect of whole-body vibration train-
ing on lean mass: a PRISMA-compliant meta-analysis. Medicine (Baltimore)
2017;96:e8390.
[28] van der Heijden GJ, Wang ZJ, Chu ZD, Sauer PJ, Haymond MW, Rodriguez LM,
et al. A 12-week aerobic exercise program reduces hepatic fat accumulation and
insulin resistance in obese, Hispanic adolescents. Obesity 2010;18:384–90.
[29] Oh S, Shida T, Sawai A, Maruyama T, Eguchi K, Isobe T, et al. Acceleration train-
ing for managing nonalcoholic fatty liver disease: a pilot study. Ther Clin Risk
Manag 2014;10:925.
[30] Liu Q, Bengmark S, Qu S. The role of hepatic fat accumulation in pathogenesis
of non-alcoholic fatty liver disease (NAFLD). Lipids Health Dis 2010;9:42.
[31] Huh JY, Mougios V, Skraparlis A, Kabasakalis A, Mantzoros CS. Irisin in response
to acute and chronic whole-body vibration exercise in humans. Metabolism
2014;63:918–21.
[32] Shamsoddini A, Sobhani V, Chehreh MEG, Alavian SM, Zaree A. Effect of aerobic
and resistance exercise training on liver enzymes and hepatic fat in Iranian
men with nonalcoholic fatty liver disease. Hepat Mon 2015;15:e31434.
[33] Salomone F, Li Volti G, Rosso C, Grosso G, Bugianesi E. Unconjugated
bilirubin, a potent endogenous antioxidant, is decreased in patients with
non-alcoholic steatohepatitis and advanced fibrosis. J Gastroenterol Hepatol
2013;28:1202–8.
[34] Manne V, Handa P, Kowdley KV. Pathophysiology of nonalcoholic fatty liver
disease/nonalcoholic steatohepatitis. Clin Liver Dis 2018;22:23–37.
[35] Farzanegi P, Dana A, Ebrahimpoor Z, Asadi M, Azarbayjani MA. Mechanisms
of beneficial effects of exercise training on non-alcoholic fatty liver dis-
ease (NAFLD): roles of oxidative stress and inflammation. Eur J Sport Sci
2019;19:994–1003.
[36] Lancaster GI, Febbraio MA. The immunomodulating role of exercise in
metabolic disease. Trends Immunol 2014;35:262–9.
[37] Khosravi S, Alavian SM, Zare A, Daryani NE, Fereshtehnejad S-M, Daryani NE,
et al. Non-alcoholic fatty liver disease and correlation of serum alanin amino-
transferase level with histopathologic findings. Hepat Mon 2011;11:452.
[38] Katsagoni CN, Georgoulis M, Papatheodoridis GV, Panagiotakos DB, Kon-
togianni MD. Effects of lifestyle interventions on clinical characteristics of
patients with non-alcoholic fatty liver disease: a meta-analysis. Metabolism
2017;68:119–32.
[39] Golabi P, Otgonsuren M, Cable R, Felix S, Koenig A, Sayiner M, et al. Non-
alcoholic fatty liver disease (NAFLD) is associated with impairment of health
related quality of life (HRQOL). Health Qual Life Outcomes 2016;14:18.
[40] Abdelbasset WK, Tantawy SA, Kamel DM, Alqahtani BA, Soliman GS. A ran-
domized controlled trial on the effectiveness of 8-week high-intensity interval
exercise on intrahepatic triglycerides, visceral lipids, and health-related quality
of life in diabetic obese patients with nonalcoholic fatty liver disease. Medicine
2019;98:e14918.
[41] Rezende RE, Duarte S, Stefano JT, Roschel H, Gualano B, de Sá P, et al. Ran-
domized clinical trial: benefits of aerobic physical activity for 24 weeks in
postmenopausal women with nonalcoholic fatty liver disease. Menopause
2016;23:876–83.
 Annales d’Endocrinologie 81 (2020) 500–506

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

Original article

Levothyroxine dose adjustment in hypothyroid patients following

gastric sleeve surgery
Ajustement de la dose de lévothyroxine pour les patients hypothyroïdiens après
une gastrectomie longitudinale
Marlène Richou , Olivier Gilly ∗ , Véronique Taillard , David Paul De Brauwere , Ion Donici ,
Anne Marie Guedj
Centre hospitalier universitaire de Nimes, Nimes, France

a r t i c l e i n f o a b s t r a c t

Keywords: Introduction. – Euthyroid patients show decreased TSH level following sleeve gastrectomy. However,
Levothyroxine studies of levothyroxine absorption after bariatric surgery reported contradictory results and data on
Hypothyroidism levothyroxine dose adjustment according to weight are sparse. The aim of this study was to evaluate
Bariatric surgery
levothyroxine dose adjustment during weight loss following sleeve surgery.
Sleeve gastrectomy
Method. – This retrospective study assessed change in levothyroxine dose in patients undergoing sleeve
gastrectomy at the university hospital center of Nîmes (France) between January 2010 and March 2016.
Patients were receiving standard bariatric surgery follow-up with levothyroxine therapy for hypothy-
roidism.
Results. – Fifty-two of the 271 patients who underwent sleeve gastrectomy (19.2%) were being treated
with levothyroxine. Among these patients, 31 were followed up for 12 months, including 12 who
were followed up for 24 months. Mean weight loss was 35 ± 11 kg at 12 months and 41.8 ± 10.2 kg
at 24 months. Daily levothyroxine dose decreased from 108 [88–144] ␮g/day to 94 [63–125] ␮g/day at
12 months and 69 [44–134] ␮g/day at 24 months, with positive correlation between dose and weight
loss at 12 months (P = 0.03). Weight-adjusted dose was 1.04 [0.81–1.24] ␮g/kg/day at baseline, 1.14
[0.85–1.66] ␮g/kg/day at 12 months, and 0.85 [0.53–2.10] ␮g/kg/day at 24 months, showing no cor-
relation with weight loss. Median TSH level dropped to 1.30 [0.63–2.27] mIU/l at 12 months and 1.48
[1.08–2.42] mIU/l at 24 months.
Conclusion. – Despite a decrease in daily levothyroxine dose correlating with weight loss at 12 months, the
absence of correlation with weight-adjusted dose suggests the involvement of confounding factors such
as poor levothyroxine absorption or altered thyroid function. Further studies are required to elucidate
the absorption of levothyroxine.
© 2020 Elsevier Masson SAS. All rights reserved.

r é s u m é

Mots clés : Introduction. – Les patients euthyroïdiens présentent un niveau de thyréostimuline (TSH) diminué après
Lévothyroxine une gastrectomie longitudinale. Cependant, des études sur l’absorption de la Lévothyroxine après une
Hyperthyroïdie chirurgie de l’obésité rapportent des résultats contradictoires et les données sur l’ajustement de la dose
Chirurgie de l’obésité de lévothyroxine en fonction du poids sont éparses. L’objectif de cette étude est d’évaluer l’ajustement
Gastrectomie longitudinale
de la dose de lévothyroxine au cours de la perte de poids qui suit la gastrectomie longitudinale.
Méthodes. – Cette étude rétrospective évalue les changements de doses de la lévothyroxine chez les
patients bénéficiant d’une gastrectomie longitudinale au centre hospitalier universitaire de Nîmes
(France) entre janvier 2010 et mars 2016. Les patients ont reçu un suivi post-opératoire standard après
une chirurgie de l’obésité ainsi qu’un traitement à la lévothyroxine pour l’hypothyroïdie.

∗ Corresponding author.
E-mail address: olivier.gilly@chu-nimes.fr (O. Gilly).

https://doi.org/10.1016/j.ando.2020.04.011
0003-4266/© 2020 Elsevier Masson SAS. All rights reserved.
 M. Richou et al. / Annales d’Endocrinologie 81 (2020) 500–506 501

Résultats. – Cinquante-deux des 271 patients ayant bénéficié d’une gastrectomie longitudinale (19,2 %)
ont été traités avec de la lévothyroxine. Parmi ces patients 31 ont été suivis pendant 12 mois dont 12 avec
un suivi de 24 mois. La perte de poids moyenne était de 35 ± 11 kg à 12 mois et de 41,8 ± 10,2 kg à 24 mois.
La dose quotidienne de lévothyroxine a été diminuée de 108 [88–144] ␮g/jour à 94 [63–125] ␮g/jour à
12 mois puis, à 69 [44–134] ␮g/jour à 24 mois, avec une corrélation positive entre la dose de lévothy-
roxine et la perte de poids à 12 mois (p = 0,03). La dose ajustée en fonction du poids était de 1,04
[0,81–1,24] ␮g/kg/jour au début de l’étude, de 1,14 [0,85–1,66] ␮g/kg/jour à 12 mois et de 0,85 [0,53–2,10]
␮g/kg/jour à 24 mois, ne montrant aucune corrélation avec la perte de poids. La médiane du niveau de
TSH a diminué jusqu’à 1,30 [0,63–2,27] mUI/L à 12 mois puis 1,48 [1,08–2,42] mUI/L à 24 mois.
Conclusion. – Malgré une corrélation entre la diminution des doses quotidiennes de lévothyroxine et la
perte de poids à 12 mois, l’absence de corrélation entre la perte de poids et les doses ajustées en fonction
du poids suggère la présence de facteurs confondants tels qu’une faible absorption de la lévothyroxine
ou un fonctionnement altéré de la thyroïde. Des études plus poussées sont nécessaires afin de mieux
comprendre l’absorption de la lévothyroxine.
© 2020 Elsevier Masson SAS. Tous droits réservés.

1. Introduction minimum follow-up of 12 months, treated with a substitutive dose

of levothyroxine for peripheral hypothyroidism. Participants were
Obesity is rising in France, in 2012, 32.3% of population was over- excluded if they had a central hypothyroidism, a poor compliance
weight and 15% obese [1]. By the way bariatric surgery is increasing: to levothyroxine, took oral liquid levothyroxine, a follow-up of less
30 513 procedures were carried out in 2011 (44% were sleeve gas- than 12 months after surgery, or were pregnant.
trectomy). Between 10.9% and 12.3% of patients receiving bariatric The recruitment and the follow up of the population is described
surgery are taking levothyroxine [2]. The dose is adapted to the in the Fig. 1 and the characteristics of the population is described
body weight; therefore obese patients require a higher dose due to in the Table 3.
their higher plasmic volume, a longer delay in intestinal absorption
of levothyroxine [3] and the increased of lean body mass [4].
2.2. Progression of patients pursuing a bariatric surgical program
Obese patients have increased levels of TSH [5,6], which reduce
or even return to normal after weight loss through dietary control
Following acceptance at the multidisciplinary team meeting,
[6–8]. Previous studies looking at the variation of thyroid function
surgery was performed under general anaesthetic by the same sur-
following bariatric surgery were focused on euthyroid patients or
geon for all patients by coelioscopy: a complete gastrolysis was
hypothyroid patients treated with levothyroxine. Thyroid function
conducted with removal of the His angle and exposure of the left
seems to improve (TSH reduction or lower levothyroxine dose)
diaphragmatic crus. The stomach was then cut and stapled from
after weight loss following bariatric surgery, to the same extent
4–5 cm above the pylorus to the angle of His. Intraoperative methy-
as weight loss through dietary interventions.
lene blue testing was conducted to ensure the integrity of the seal.
Only three studies have investigated sleeve gastrectomy
Patients were then registered, and followed every 3 months for
[17,20,21] (Tables 1 and 2).
1 year (M3-M6-M9-M12), then every 6 months the following year
In healthy non-operated patients, a normal gastric acid secre-
(M18-M24), and subsequently once per year. At each visit, blood
tion is important for the absorption of levothyroxine [18] who is
samples were taken to test TSH levels to confirm euthyroid status.
largely absorbed in the jejunum-ileum [19]. However, data from
Patients met with their doctor who recorded their body weight,
surgery patients is contradictory: several cases of malabsorption
treatment (levothyroxine dose) and any related events. Vitamin
following bypass have been reported [20–23], although one more
supplements were prescribed if any deficiencies were discovered
recent study instead showed an improvement in pharmacokinetics
and the dose of levothyroxine was modified according to TSH level.
following sleeve gastrectomy and bypass [24].
There is currently little data relating levothyroxine treatment
with thyroid function following sleeve gastrectomy. The aim of this 2.3. Data collection pre-surgery (M0)
study was to investigate the change in levothyroxine dose against
weight loss in hypothyroid patients having undergone sleeve gas- Body weight and height were recorded at the multidisciplinary
trectomy. Weight loss would be expected to result in a decrease in team meeting, which took place on average 2–3 months prior to
daily levothyroxine dose (␮g/day) and a stabilisation of the weight- surgery. There was frequently a difference of ± 2–3 kg between the
adjusted dose ␮g/kg/day. When thyroid function improves, the body weight recorded at the RCP and the body weight recorded on
weight-adjusted dose may decrease, while levothyroxine malab- the day of the surgery, although this difference was no considered to
sorption of may result in an increased weight-adjusted dose. be significant for our study. Levothyroxine dose was recorded from
the consultation or hospital letters preceding surgery to ensure that
there was no variation in the 3 months prior to surgery.
2. Method

2.1. Participants 2.4. TSH measurement

This is a retrospective observational study. Participants were
recruited from patients hospitalised between January 2010 and
March 2016 in the maladies métaboliques et endocriniennes
department of the CHU de Nîmes. Participants included were males
or females over the age of 18 years, having received sleeve gastrec-
tomy at the CHU de Nîmes prior to March 2016 in order to have a
Serum TSH levels were measured by enzyme-linked
immunosorbent assay from January 2010 to March 2013 (Beckman
Coulter® , sensitive to 0.015 mUI/l, reference values 0.40 mUI/l to
4 mUI/l), and by electrochemiluminescent immunometric assay
from March 2013 (Elecys Roche Cobas® , sensitive to 0.014 mUI/l,
reference values 0.40 mUI/l to 3.77 mUI/l).
502 M. Richou et al. / Annales d’Endocrinologie 81 (2020) 500–506

Table 1
Variation of TSH following weight loss due to bariatric surgery in euthyroid patients.

Reference Surgery Number of patients Time after surgery Subjects TSH T3 LT4

Dall’Asta et al. [9] Band 258 6–24 months Euthyroid = ↓LT3 ↑

Lips et al. [8] Band + restriction 23 3 months Euthyroid females ↓ ↓T3 =
Chikunguwo et al. [10] Band + bypass 86 6–12 months Euthyroid ↓ =
MacCuish et al. [11] Bypass 55 4.5–24 months Euthyroid = ↑
Moulin de Moraes et al. [12] Bypass 72 12 months Euthyroid and ↓ ↓T3 =
subclinical
hypothyroid
without treatment
Abu Ghanem et al. [13] Sleeve 38 6–12 months Euthyroid ↓ =

Table 2
Variation in levothyroxine dose following weight loss from bariatric surgery in hypothyroid patients.

Reference Surgery Number of patients Time after surgery Subjects Daily Weight-adjusted
levothyroxine dose dose (␮g/kg/day)
(␮g/J)

Fazylov et al. [14] Bypass 20 1–24 months Hypothyroid = (8) or ↓(7)

females taking ↑(5)
levothyroxine
Raftopoulos et al. [15] Bypass 23 5.5–31 months Hypothyroid taking = (13) or ↓(10)
levothyroxine
Aggarwal et al. [16] Sleeve 19 3–36 months Hypothyroid taking ↓(13) or = (6)
levothyroxine
Fierabracci et al. [17] Bypass + band + sleeve 93 (sleeve 8) 20–36 months Hypothyroid taking ↓ ↑
levothyroxine

Fig. 1. Recruitment and follow-up of patients receiving sleeve gastrectomy.

Table 3
Pre-surgery patient characteristics.

Total = 31 patients 12 patients followed 24 months (38.7%)

Ratio males/females 2/29 1/11

Age average (years) 50.9 ± 9.07 46.2 ± 8.9
Body weight average (kg) 112.1 ± 16.7 117.3 ± 17.2
BMI average (kg/m2 ) 42.6 ± 5.4 43.9 ± 6.3
Number patients BMI < 35(kg/m2 ) 2 (6.5%) 1 (8.3%)
Number patients BMI [35–39] (kg/m2 ) 9 (29%) 3 (25%)
Number patients BMI ≥ 40 (kg/m2 ) 20 (64.5%) 8 (66.7%)
Median daily dose of levothyroxine (␮g/day) 108 [88–144] 125 [84–153]
Median adjusted dose of levothyroxine (␮g/kg/day) 1.04 [0.81–1.24] 1.03 [0.69–1.34]
 M. Richou et al. / Annales d’Endocrinologie 81 (2020) 500–506
Fig. 2. Evolution of body weight.
Fig. 3. Change in daily dose of levothyroxine.
2.5. Judgment Criteria
The main judgment criterion was the change in daily levothy-
roxine dose (␮g/day) between M0 and M12 or M24. Other judgment
criteria were change in weight-adjusted dose (␮g/kg/day) between
M0 and M12 or M24, change in weight-adjusted dose (␮g/kg/day)
between M0 and each follow-up visit at M6-M12-M24 and change
in body weight (kg) between M0 and each follow-up visit at
M6-M12-M24. TSH levels were observed at each hospital visit at
M6-M12-M24 to confirm euthyroid status. The normal level of TSH
value was defined as TSH between 0.40–3.77 mUI/l.
2.6. Data analysis
The correlation between dose alteration and change in body
weight between M0 and M12 or M24 was calculated by Kendall
ranking coefficient with a P ≤ 0.05 considered statistically signifi-
cant. The Kendall coefficient was chosen as the dose was a discrete
variable, not following a normal law, and often presented equiva-
lent values.
3. Results
Between January 2010 and March 2016, 271 patients received
sleeve gastrectomy at the CHU de Nîmes. Fifty-two patients (19.2%)
were taking levothyroxine. Twenty-one (40.4%) patients were
excluded and 31 included in the study (Fig. 1).
The pre-surgery patient characteristics are described in Table 1.
503
The average weight loss was 35 ± 11 kg at 12 months (Fig. 2A)
and 41.8 ± 10.2 kg at 24 months (Fig. 2B).
The median daily dose of levothyroxine dropped for all 31
patients followed for 12 months from 108[88–144] to 94[63–125]
␮g/day (Fig. 3A) at 12 months, and for the 12 patients followed to
24 months, from 125[84–153] to 69[44–134] ␮g/day (Fig. 3B).
Weight-adjusted levothyroxine dose went from 1.04[0.81-1.24]
to 1.14[0.85-1.66] ␮g/kg/day (Fig. 4A) for the 31 patients fol-
lowed for 12 months, and from 1.033[0.69–1.34] to 0.90[0.53–2.10]
␮g/kg/day (Fig. 4B) for the 12 patients followed for 24 months.
A significant positive correlation was found between weight
loss and decrease in daily levothyroxine dose ( Kendall = 0.29;
P = 0.03) (Fig. 5) and between the decrease in BMI and decrease in
daily levothyroxine dose ( Kendall = 0.34; P = 0.01) (Fig. 6) at 12
months. This correlation was not significant in the group followed
for 24 months.
There was no significant correlation between weight loss and
the decrease in weight-adjusted dose of levothyroxine.
4. Discussion
Thyroid function in obese patients is different than in nor-
mal weight patients, the major difference being higher TSH, and
peripheral hormones level in serum [5,6]. This may be due to an
activation of the hypothalamic-pituitary-thyroid axis associated
to the accumulation of fat mass. There is a complex interaction
between adipose tissue and the hypothalamic pituitary thyroid axis
through the action of various adipocytes cytokines. Any authors talk
about “acquired resistance to thyroid hormone” in obesity.
504 M. Richou et al. / Annales d’Endocrinologie 81 (2020) 500–506

Fig. 4. Change in weight-adjusted dose of levothyroxine.

Fig. 5. Correlation between weight loss and daily dose of levothyroxine.

Fig. 6. Correlation between BMI and daily dose of levothyroxine.

 M. Richou et al. / Annales d’Endocrinologie 81 (2020) 500–506
Hypotheses include an adaptive mechanism to compensate for
weight gain, partially inactive TSH or resistance to TSH and/or
abnormal hypothalamic-pituitary regulation with a likely role of
leptin [34]. Indeed studies have shown that after weight loss the
regulation of the hypothalamic-pituitary-thyroid axis is improved.
A decrease in ghrelin is observed after by-pass and sleeve gas-
trectomy [32] as it is mainly secreted in the gastric fundus, which is
removed by this surgery. Ghrelin has appetite stimulating proper-
ties and also plays a role in thyroid function: Emami et al. showed
a negative correlation between ghrelin and TSH and a positive
correlation between ghrelin and T4 and T3 in both hypothyroid
and hyperthyroid patients [33]. A two-way regulation between
the hypothalamic-pituitary and peptides secreted by the digestive
system appears to exist, although mechanisms have not yet been
elucidated [32].
After weight loss in euthyroid bariatric surgery parties, some
studies reported improvement in thyroid function by a decrease of
TSH [8,12,13]. To our knowledge, only one study has analysed the
change of TSH after sleeve gastrectomy (38 patients) and found a
modest yet significant decrease (−0.63mUI/l) in TSH levels that was
not correlated to weight [13].
We know that, after oral administration, 30–90% of levothy-
roxine dose is absorbed within 3 h of ingestion, a majority of the
absorption takes place in the jejunum and ileum. Absorption is
optimal when administered on an empty stomach, reflecting the
importance of gastric acidity in the process.
Following sleeve gastrectomy, our data show that the required
daily dose of levothyroxine decreased with a positive correlation
with weight loss at 12 months between. This correlation was not
retained at 24 months, but most likely due to t the small num-
ber of participants. These results advocate to the importance of
adapting dose to the body weight [3,4]. However, when remov-
ing the “weight effect” by studying the weight-adjusted dose, this
correlation between dose and weight loss was not found.
In all, this suggests that there is no modification of thyroid
function after sleeve gastrectomy. Several individual factors could
induce a variation in adjusted levothyroxine dose. Any treatments
can cause an increase in available levothyroxine. Proton pump
inhibitors (PPI) increase gastric pH through inhibition of the pari-
etal gastric cell H+ /K+ ATPase pump and therefore the dissolution
of levothyroxine [18,25,26]. Iron salts taken orally chelate levothy-
roxine in the digestive tract, preventing its absorption [27]. Calcium
carbonate and cholestyramine also decrease the absorption of
levothyroxine [28]. In our study, 13 patients out of 33 (39.4%)
received PPI at least once over the course of the study; 5 (15.2%)
receiving ferrous sulfate and 8 (24.3%) taking other supplements
containing calcium carbonate (Alvityl® , Azinc® , Protenplus® ). As
not all patients received these treatments, and those that did were
given different dosages and varying lengths of treatment, varia-
tions in absorption could well arise causing knock-on effects on
levothyroxine dose.
Certain digestive disorders can alter levothyroxine absorption:
coeliac disease, gastritis, lactose intolerance and digestive tract
inflammatory disease (Crohn) [28,31]. We do not have a record of
such digestive pathologies in our study.
Eating behaviours change following bariatric surgery. Not only
is there a decrease in quantities consumed, but also a change in the
types of food: a decrease in appetite for high energy food [29] which
could lead to consumption for food affecting absorption of levothy-
roxine. Soya, high fibre foods (walnuts, papaya), coffee [30] and
grapefruit juice can alter levothyroxine absorption by sequestering
it in the digestive tract [28,31]. It was not possible in this retrospec-
tive study to examine the eating habits of patients regarding these
food products.
Patients had different causes of hypothyroidism. Some were
putatively defined as Hashimoto hypothyroid, but without a
505
measure of anti-TPO antibody it was not possible to confirm this
diagnosis, and it is possible that this disease was not explored in
the other patients. It was evocated that thyroiditis group can have
a more fluctuant TSH value during levothyroxine treatment than
post-surgical hypothyroidism. In our study five patients (15.2%) had
received thyroidectomies, four for benign nodules, and one for a
low risk papillary carcinoma. However alteration of levothyroxine
dose for these patients was similar to the principal cohort in this
study.
In hypothyroid patients taking levothyroxine, improved thyroid
function with weight loss could result in a decrease in levothyrox-
ine dose. Analysis of the alteration in levothyroxine dose following
sleeve gastrectomy in 19 patients showed a decrease in daily dose
in the majority of patients (13/19, 68.4%), or a stabilisation of
dose [16]. There was a positive correlation between daily dosage
and percentage of excess weight lost. However this dose was not
adjusted for body weight, and as we have seen above, it is logical
that daily dose would decrease with body weight. Another study
analysed the change in levothyroxine dose following sleeve gas-
trectomy in a small cohort: eight patients out of 93 undergoing
sleeve gastrectomy with a similar result to the main cohort (by-
pass + band + sleeve): there is not a significant correlation between
weight loss and decrease in levothyroxine dose [17]. The correla-
tion with weight-adjusted dose (which rose in this study) was not
examined. In our study there was no correlation between decrease
in weight-adjusted dose and the loss of weight. Weight loss varied
hugely between patients: between eight and 54 kg at 12 months
and between 26 and 58 kg at 24 months. It would be interesting
to study the change in TSH activity and the leptin concentration
in patients compared with this weight loss. Fierabracci et al. saw
an average drop of 60% in leptin following bariatric surgery; but
without correlation with weight loss. In addition, the decrease in
levothyroxine dose was proportional to the decrease in lean body
mass evaluated by absorptiometry [17].
In summary, it appears that the pharmacokinetics of levothy-
roxine are improved following sleeve gastrectomy: Gkotsina et al.
have shown that the AUC (area under curve) of plasmic con-
centration of levothyroxine administered to euthyroid patients
was improved after biliopancreatic derivation and after sleeve
gastrectomy [24]. This paradoxical result is difficult to explain:
levothyroxine is less in contact with bile secretion,therefore less
submited to glucurono-conjugation and therefore better absobed
Modification of the intestinal microbiota and proliferation of diges-
tive mucus is also likely.
The advantage of our study is that the analysis of weight-
adjusted dose changes (␮g/kg/day) allowed us to dissect the effect
of weight.
Nevertheless, our study had certain limitations: as a retrospec-
tive observational study with a small cohort size, there was a lack of
power particularly for the analyses at 24 months. However, the pre-
vious largest study in the literature studying the changes in dose of
levothyroxine following sleeve gastrectomy comprised 19 patients,
compared to 31 in ours. Finally, the adaptation of levothyroxine
over the course of the study was not protocoled; it was based upon
TSH value, but the dose changes were at the discretion of the doctor,
and the value was large between 0.40–3.77 mUI/l.
Similarly, it was difficult to evaluate patient compli-
ance in terms of taking the medication in this observational
study.
Another limit of this study is the only descriptive character; it
will be interesting to have a control group in order to compare the
result, with thyroid function in obese patient with weight loss and
no surgery.
Finally the median adjusted dose of levothyroxine is lower
than expected with value between 1.04 to 1.03 ␮g/kg/day who
correspond more with subclinical hypothyroidism than real
506 M. Richou et al. / Annales d’Endocrinologie 81 (2020) 500–506
hypothyroidism. This fact underlines the difficulty to evaluate the
thyroid function in obese population in general.
In conclusion, after weight loss following sleeve gastrectomy,
the daily dose of levothyroxine decreases, but not regarding
adjusted weight dose. Confusing factors are multiple and prevent
us from proving incontestably a variation in dose adjusted for
weight. This results must encourage to a regular follow-up of TSH
level in the years following bariatric surgery in patients treated
with levothyroxine in order to adapt treatment. Further proto-
coled studies on larger cohorts including absorption observations
are necessary, with control groups in order to confirm this results.
Human and animal rights
The authors have not submitted a declaration concerning
human and animal experimentation.
Informed consent and patient details
The authors have not submitted a declaration concerning
informed consent and patient details.
Disclosure of interest
The authors declare that they have no competing interest.
Funding
The authors have not submitted a declaration concerning the
funding of their work.
Author contributions
The authors have not submitted a declaration concerning their
individual contributions to this work.
CRediT authorship contribution statement
Gilly Olivier: conceptualization, methodology, software. Richou
Marlène: data curation, writing- original draft preparation. Taillard
Véronique: visualization, investigation. Guedj Anne Marie: super-
vision. Donici Ion, De Brauwere David Paul: software, validation.
Gilly olivier: writing- reviewing and editing.
References
[1] Inserm. ObEpi 2012, enquête épidémiologique nationale
le surpoids et l’obésité [Internet]; 2012 http://www.roche.fr/
content/dam/corporate/roche fr/doc/obepi 2012.pdf.
[2] Fierabracci P, Pinchera A, Martinelli S, Scartabelli G, Salvetti G, Giannetti M,
et al. Prevalence of Endocrine Diseases in Morbidly Obese Patients Scheduled
for Bariatric Surgery: Beyond Diabetes. Obes Surg 2011;21:54–60.
[3] Michalaki M, Gkotsina M, Mamali I, Markantes G, Faltaka A, Kalfarentzos F, et al.
Impaired Pharmacokinetics of Levothyroxine in Severely Obese Volunteers.
Thyroid 2011;21:477–81.
[4] Santini F, Pinchera A, Marsili A, Ceccarini G, Castagna M, Valeriano R, et al. Lean
Body Mass Is a Major Determinant of Levothyroxine Dosage in the Treatment
of Thyroid Diseases. J Clin Endocrinol Metab 2005;90:124–7.
[5] Michalaki M, Vagenakis A, Leonardou A, Argentou M, Habeos I, Makri M, et al.
Thyroid Function in Humans with Morbid Obesity. Thyroid 2006;16:73–8.
[6] Pearce E. Thyroid hormone and obesity. Curr Opin Endocrinol Diabetes Obes
2012;19:408–13.
[7] Reinehr T. Obesity and thyroid function. Mol Cell Endocrinol 2010;316:165–71.
[8] Lips M, Pijl H, Klinken J, Groot G, Janssen I, Ramshorst B, et al. Roux-en-Y gas-
tric bypass and calorie restriction induce comparable time-dependent effects
on thyroid hormone function tests in obese female subjects. Eur J Endocrinol
2013;169:339–47.
[9] Dall’Asta C, Paganelli M, Morabito A, Vedani P, Barbieri M, Paolisso G, et al.
Weight Loss Through Gastric Banding: Effects on TSH and Thyroid Hormones
in Obese Subjects With Normal Thyroid Function. Obesity 2010;18:854–7.
[10] Chikunguwo S, Brethauer S, Nirujogi V, Pitt T, Udomsawaengsup S, Chand B,
et al. Influence of obesity and surgical weight loss on thyroid hormone levels.
Surg Obes Relat Dis 2007;3:631–5.
[11] MacCuish A, Razvi S, Syed A. Effect of weight loss after gastric bypass surgery
on thyroid function in euthyroid people with morbid obesity. Clin Obes
2012;2:25–8.
[12] Moulin de Moraes C, Mancini M, de Melo M, Figueiredo D, Villares S, Rascovski A,
et al. Prevalence of subclinical hypothyroidism in a morbidly obese population
and improvement after weight loss induced by Roux-en-Y gastric bypass. Obes
Surg 2005;15:1287–91.
[13] Abu Ghanem Y, Inbar R, Tyomkin V, Kent I, Berkovich L, Ghinea R, et al. Effect
of Sleeve Gastrectomy on Thyroid Hormone Levels. Obes Surg 2014;25:452–6.
[14] Fazylov R, Soto E, Cohen S, Merola S. Laparoscopic Roux-en-Y gastric
bypass surgery on morbidly obese patients with hypothyroidism. Obes Surg
2008;18:644–7.
[15] Raftopoulos Y, Gagné D, Papasavas P, Hayetian F, Maurer J, Bononi P, et al.
Improvement of hypothyroidism after laparoscopic Roux-en-Y gastric bypass
for morbid obesity. Obes Surg 2014;14:509–13.
[16] Aggarwal S, Modi S, Jose T. Laparoscopic Sleeve Gastrectomy Leads to Reduction
in Thyroxine Requirement in Morbidly Obese patients With Hypothyroidism.
World J Surg 2014;38:2628–31.
[17] Fierabracci P, Martinelli S, Tamberi A, Piaggi P, Basolo A, Pelosini C, et al.
Weight Loss and Variation of Levothyroxine Requirements in Hypothyroid
Obese Patients After Bariatric Surgery. Thyroid 2016;26:499–503.
[18] Pabla D, Akhlaghi F, Zia H. A comparative pH-dissolution profile study of
selected commercial levothyroxine products using inductively coupled plasma
mass spectrometry. Eur J Pharm Biopharm 2009;72:105–10.
[19] Hays M. Localization of Human Thyroxine Absorption. Thyroid 1991;1:241–8.
[20] Padwal R, Brocks D, Sharma A. A systematic review of drug absorption following
bariatric surgery and its theoretical implications. Obes Rev 2010;11:41–50.
[21] Bevan J, Munro J. Thyroxine malabsorption following intestinal bypass surgery.
Int J Obes 1986;10:245–6.
[22] Topliss D, Wright J, Volpé R. Increased requirement for thyroid hormone after
a jejunoileal bypass operation. Can Med Assoc J 1980;123:765–6.
[23] Azizi F, Belur R, Albano J. Malabsorption of thyroid hormones after jejunoileal
bypass for obesity. Ann Intern Med 1979;90:941–2.
[24] Gkotsina M, Michalaki M, Mamali I, Markantes G, Sakellaropoulos G, Kalfarent-
zos F, et al. Improved Levothyroxine Pharmacokinetics After Bariatric Surgery.
Thyroid 2013;23:414–9.
[25] Vita R, Saraceno G, Trimarchi F, Benvenga S. Switching Levothyroxine From
the Tablet to the Oral Solution Formulation Corrects the Impaired Absorption
of Levothyroxine Induced by Proton-Pump Inhibitors. J Clin Endocrinol Metab
2014;99:4481–6.
[26] Centanni M, Gargano L, Canettieri G, Viceconti N, Franchi A, Fave G, et al. Thy-
roxine in Goiter, Helicobacter pylori Infection, and Chronic Gastritis. N Engl J
Med 2006;354:1787–95.
[27] Campbell N, Hasinoff B, Stalts H, Rao B, Wong N. Ferrous sulfate reduces
thyroxine efficacy in patients with hypothyroidism. Ann Intern Med
1992;117:1010–3.
[28] Liwanpo L, Hershman J. Conditions and drugs interfering with thyroxine
absorption. Best Pract Res Clin Endocrinol Metab 2009;23:781–92.
[29] Miras A, le Roux C. Bariatric surgery and taste: novel mechanisms of weight
loss. Curr Opin Gastroenterol 2010;26:140–5.
sur
[30] Benvenga S, Bartolone L, Pappalardo M, Russo A, Lapa D, Giorgianni G,
et al. Altered intestinal absorption of L-thyroxine caused by coffee. Thyroid
2008;18:293–301.
[31] Benvenga S. When thyroid hormone replacement is ineffective? Curr Opin
Endocrinol Diabetes Obes 2013;20:467–77.
[32] Langer F, Hoda M, Bohdjalian A, Felberbauer F, Zacherl J, Wenzl E, et al. Sleeve
gastrectomy and gastric banding: effects on plasma ghrelin levels. Obes Surg
2005;15:1024–9.
[33] Emami A, Nazem R, Hedayati M. Is association between thyroid hormones and
gut peptides, ghrelin and obestatin, able to suggest new regulatory relation
between the HPT axis and gut? Regul Pept 2014;189:17–21.
[34] Svare A, Nilsen T, Bjøro T, Åsvold B, Langhammer A, et al., Serum TSH. related
to measures of body mass: longitudinal data from the HUNT Study, Norway.
Clin Endocrinol (Oxf) 2011;74:769–75.
 Annales d’Endocrinologie 81 (2020) 507–510

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

Review

Thyroid disorders and SARS-CoV-2 infection: From

pathophysiological mechanism to patient management
Atteintes thyroïdiennes au cours de l’infection par le SARS-CoV-2 : du mécanisme
physiopathologique à la prise en charge des patients
Philippe Caron
Service d’endocrinologie et maladies métaboliques, pôle cardiovasculaire et métabolique, CHU Larrey, 24, chemin de Pouvourville, TSA 30030, 31059
Toulouse cedex, France

a r t i c l e i n f o a b s t r a c t

Keywords: The World Health Organization (WHO) declared the COVID-19 epidemic to be a global pandemic in March
SARS-CoV-2 2020. COVID-19 is an infection caused by SARS-CoV-2, a coronavirus that utilizes the angiotensin-2 con-
COVID-19 verting enzyme to penetrate thyroid and pituitary cells, and may result in a “cytokine storm”. Based on
Free T3
the pathophysiological involvement of the pituitary-thyroid axis, the current review discusses the diag-
Free T4
nosis of abnormal thyroid function test, and the management of patients presenting with thyrotoxicosis,
TSH
Thyrotoxicosis thyroid-associated orbitopathy and hypothyroidism in the context of SARS-CoV-2 infection.
Orbitopathy © 2020 Elsevier Masson SAS. All rights reserved.
Hypothyroidism

r é s u m é

Mots clés : En mars 2020, l’Organisation mondiale de la santé (OMS) a déclaré l’épidémie COVID-19 une pandémie
SARS-CoV-2 mondiale. L’infection est causée par le SARS-CoV-2, un coronavirus qui utilise l’enzyme de conversion
COVID-19 de l’angiotensine 2 pour pénétrer dans les cellules thyroïdiennes et hypophysaires et qui peut entraîner
T3l un « orage cytokinique ». Sur les bases physiopathologiques de l’infection par le SARS-CoV-2 sur l’axe
T4l
hypophyso-thyroïdien, nous abordons dans cette revue le diagnostic des perturbations fonctionnelles
TSH
thyroïdiennes et la prise en charge des patients avec une thyréotoxicose, une orbitopathie basedowienne
Thyrotoxicose
Orbitopathie et une hypothyroïdie au cours d’une infection par le SARS-CoV-2.
Hypothyroïdie © 2020 Elsevier Masson SAS. Tous droits réservés.

The World Health Organization (WHO) declared the COVID-19
epidemic to be a global pandemic in March 2020. The condition
involves SARS-CoV-2 infection and may result in acute pneumoni-
tis, severe forms of which determine prognosis. Recent clinical
and epidemiological data report varied extra-pulmonary visceral
(cutaneous, neurological, cardiovascular, ophthalmological, etc.)
but also endocrine (pancreatic, pituitary, etc.) involvement, and
particularly thyroid disorders associated with COVID-19.
An understanding of the pathophysiological involvement of
an abnormal pituitary-thyroid axis in SARS CoV-2 infection may
E-mail address: caron.p@chu-toulouse.fr
enable correct interpretation of thyroid function test anomalies and
accurate assessment of thyroid function, particularly in patients
with severe forms requiring Emergency Room (ER) treatment,
allowing appropriate management of thyroid dysfunctions, in par-
ticular thyrotoxicosis and thyroid insufficiency.
1. Pathophysiology of the pituitary-thyroid axis in
SARS-CoV-2 infection
The SARS-CoV-2 coronavirus can cause immune response
hyperactivity involving Th1/Th17 lymphocytes [1,2] leading to
release of pro-inflammatory cytokines (interleukin 1-6, tumor
necrosis factor ␣), and which may cause a “cytokine storm”. In

https://doi.org/10.1016/j.ando.2020.09.001
0003-4266/© 2020 Elsevier Masson SAS. All rights reserved.
508 P. Caron / Annales d’Endocrinologie 81 (2020) 507–510
the acute phase, increased concentrations of interleukins, and in
particular of interleukin 6, lead to thyrotoxicosis, the prevalence
of which correlates with interleukin 6 elevation [3], disruption of
desiodases and thyroid hormone transport proteins, and impaired
pituitary cell TSH secretion, resulting in abnormal thyroid func-
tional parameters. There is a decrease in free T3 concentration,
correlating with an increase in interleukin 6 [4], normal or mod-
erately decreased free T4 and normal or decreased TSH. These
anomalies are described as “low T3 syndrome” or “euthyroid sick
syndrome” (anomalies of thyroid function parameters reported
in non-thyroid-related conditions). These thyroid function test
anomalies are generally transient and do not require specific treat-
ment [5].
During follow-up, the “cytokine storm” may cause immune
system dysregulation, leading to autoimmune disorders such as
anti-phospholipid syndrome, thrombocythemia, hemolytic ane-
mia, Guillain-Barré syndrome, and, in terms of thyroid disorders,
Graves’ disease and, more rarely, chronic autoimmune thyroiditis
[6].
Autopsy studies conducted in the aftermath of the SARS-
CoV epidemic found destruction of thyroid follicular cells [7],
also reported from histological data obtained from patients who
experienced an infectious event related to SARS-CoV-2 [8]. Histo-
logical examination of the thyroid revealed absence of lymphocytic
infiltrates but presence ‘of extensive apoptosis, suggestive of
destructive thyroiditis which may be the causal factor in thyro-
toxicosis [9].
In addition, recent studies demonstrated the presence of
SARS-CoV-2 RNA in serum and plasma from COVID-19 patients,
indicating episodes of viremia [10]. The SARS-CoV-2 virus uses the
angiotensin II converting enzyme, a membrane carboxypeptidase,
as a “receptor” to gain entry into cells. Angiotensin II converting
enzyme is expressed on thyroid follicular and pituitary cells [11],
and this may explain the incidence of thyroid and pituitary disor-
ders reported during COVID-19 episodes [12,13].
2. Abnormal thyroid function test during SARS-CoV-2
infection
Thyroid function tests from 48 patients evaluated in acute-
phase SARS-CoV infection showed decreases in T3 and T4
concentrations in 94% and 46% of patients, respectively. These
anomalies persisted during convalescence [4]. T3 decrease cor-
related with disease severity, and there was also a decrease in
TSH concentration. A study of SARS-CoV survivors reported that
6.7% developed hypothyroidism; one-quarter of these patients pre-
sented peripheral autoimmune thyroiditis (presence of antithyroid
autoantibodies) and three-quarters central thyroid insufficiency,
associated with corticotropic insufficiency in 66% of patients, and
correlated with pituitary disorders that proved to be transient on
long-term follow-up [14].
A recent study of patients with SARS-CoV-2 lung infection
showed a decrease in total T3 and TSH concentrations compared to
controls or to patients with non-COVID-19 pneumonia [15]. These
anomalies, and particularly the decrease in T3 concentration, were
more pronounced in patients with the severe forms resulting in
death than in SARS-CoV-2 survivors [16].
Although, on March 13, 2020, the World Health Organization
(WHO) did not recommend systematic thyroid function test of hos-
pitalized COVID-19 patients [17], it seems useful to perform thyroid
function test in patients admitted to the ER or Intensive Care Unit
(ICU), given the prevalence of thyroid dysfunctions, and also dur-
ing follow-up to detect onset of hypothyroidism in the context of a
pituitary disorder and to diagnose thyrotoxicosis linked to inflam-
matory or destructive thyroiditis [18].
3. Thyrotoxicosis and SARS-CoV-2 infection
During the course of SARS-CoV-2 infection, thyrotoxicosis may
be complicated by thromboembolic episodes and cardiac rhythm
disorders (atrial fibrillation), increasing morbidity and mortality
[19–22].
Thyrotoxicosis may be secondary to various conditions:
• Graves’ disease is an autoimmune thyroid disorder related to the
presence of TSH receptor-stimulating antibodies. Activation of
an autoimmune response, irrespective of context [23] and thus
including during COVID-19 infection, may induce onset or relapse
of Graves’ disease [24]. The risk of SARS-CoV-2 infection is not
higher in diagnosed or treated Graves’ patients. Conversely, in the
elderly with Graves’ disease that is not well controlled by medical
treatment, severe SARS-CoV-2 infection may lead to a thyrotoxic
storm, increasing COVID-19-related mortality. This highlights the
importance of continuing treatment of Graves’ disease during a
COVID-19 episode.
Antithyroid drugs are the first-line treatment in Graves’ disease
[25]. A “block-replace” regimen should be considered for both
children and adult patients [26], to facilitate control of thyrotox-
icosis, minimize hormone level check-ups and teleconsultations
[27] during lockdown. Antithyroid therapy may be complicated
by agranulocytosis, a rare but serious hematological complica-
tion, which can produce infectious signs and symptoms similar
to those of a COVID-19 episode. Agranulocytosis requires imme-
diate discontinuation of antithyroid drugs and complete blood
count [28]. Weekly complete blood counts should be suggested
at the initiation of antithyroid drug treatment, for early detection
of neutropenia and prevention of agranulocytosis. In infectious
episodes, and particularly pulmonary infection due to SARS-CoV-
2, agranulocytosis may exacerbate COVID-19 symptoms;
• subacute inflammatory or destructive thyroiditis in COVID-
19 patients can be secondary to the “cytokine storm”, with
interleukin-6 elevation inducing inflammatory thyroiditis [3],
or alternatively be secondary to the SARS-CoV-2 infection,
which causes destructive thyroiditis [29], as observed dur-
ing other viral infections that induce a thyrotoxicosis episode
(cytomegalovirus, enterovirus, coxsackievirus) [30]. Following
case reports of patients with thyrotoxicosis related to subacute
thyroiditis [29,31–33], a retrospective study of 287 patients hos-
pitalized for COVID-19 infection found higher incidence (20.2%)
of thyrotoxicosis related to inflammatory thyroiditis (in absence
of antithyroid autoantibodies) and correlating with interleukin
6 concentration [3]. Clinical characteristics of thyroiditis during
COVID-19 include: higher incidence in females [22], increased
frequency of heart rhythm disorders (atrial fibrillation) [22], and
silent cervical forms. Lymphocytopenia is a common hemato-
logical anomaly encountered in SARS-CoV-2 infection, and may
decrease lympho-plasmocytic infiltration of the thyroid gland,
thereby decreasing the pain symptoms in the anterior cervical
region observed in some patients [18]; at the hormonal level, “T4
thyrotoxicosis” may result from thyroid cell lysis [4,7] releasing
synthesized thyroid hormones, along with a decrease in desio-
dase activity in the COVID-19 infectious context, responsible
for a decrease in T3 concentration. At the paraclinical level, a
hypoechoic, heterogeneous non-vascularized thyroid gland may
be observed on cervical ultrasound scan [22,31], not fixing on
scintigraphy. Treatment with ␤-blockers should be considered,
depending on the cardiologic context, and corticosteroid ther-
apy (at an initial dose < 0.5 mg/kg/day) should be considered,
depending on the COVID-19 infectious context. Follow-up may
be marked by emergence of transient hypothyroidism [22].
In practice, it is advisable to perform a thyroid function test in
COVID-19 patients, particularly in those admitted to ICU, because
 P. Caron / Annales d’Endocrinologie 81 (2020) 507–510
of the relatively high incidence of SARS-CoV-2-induced thyrotox-
icosis secondary to subacute thyroiditis [18].
4. Graves’ orbitopathy and SARS-CoV-2 infection
Autoimmune orbital disorders are observed in Graves’ dis-
ease and Hashimoto’s thyroiditis. Thyroid-associated orbitopathy
is clinically significant in 30–50% of patients and impairs visual acu-
ity in 5%. Progression of thyroid-associated orbitopathy features an
inflammatory phase followed by the emergence of fibrosis with
sequelae of varying severity (Rundle curve).
Patients in an inflammatory stage of orbitopathy are at greater
risk of COVID-19 infection, which explains why the quarantine
rules for these patients must be strictly respected. Conjunctival
disorders related to SARS-CoV-2 infection may delay diagnosis of
Graves’ orbitopathy and expose patients to a risk of exacerbation,
with eye infections and decreased visual acuity.
At the inflammatory stage and depending on the severity of the
orbitopathy, the following options should be considered [25]: for
mild forms, cease smoking, restore and maintain euthyroidism,
and initiate oral selenium supplementation; for moderate and
severe forms, initiate first-line intravenous corticosteroid therapy,
or oral corticosteroid therapy although this carries a risk of side-
effects (type-2 diabetes, arterial hypertension) and corticotropic
insufficiency; for severe forms with decreased visual acuity, surgi-
cal orbital decompression should be discussed when intravenous
corticosteroid therapy has failed. Surgery for sequelae of thyroid-
associated orbitopathy may be postponed until after the COVID-19
pandemic.
Medical treatment with intravenous and oral glucocorticoids
or immunosuppressive therapy with mycophenolate mofetil, aza-
thioprine, rituximab or tocilizumab are risk factors for SARS-CoV-2
infection, and patients may develop severe forms of COVID-19 [25].
It should be noted that a recent in vitro study of orbital fibro-
blasts isolated from patients with non-inflammatory or mildly
inflammatory Graves’ orbitopathy evaluated chloroquine and
hydroxychloroquine treatment, which have recently been dis-
cussed as potential treatments for specific infectious forms of
COVID-19 [34]. This study reported dose-dependent inhibition of
orbital fibroblast proliferation, adipogenesis and glycoaminogly-
can production, which all contribute significantly to autoimmune
orbital orbitopathy [35]. Thus, according to the authors, chloro-
quine and hydroxychloroquine, respecting standard contraindi-
cations and monitoring criteria, may be a treatment for Graves’
orbitopathy by inhibiting orbital fibroblast autophagy.
5. Hypothyroidism and SARS-CoV-2 infection
In adults, primary hypothyroidism occurs most commonly as
a result of either autoimmune thyroiditis or surgical (total thy-
roidectomy) or radioisotope treatment. Central hypothyroidism
is much more uncommon. Epidemiological studies indicate that
patients with hypothyroidism are not at increased risk of COVID-
19 infection [36,37]. On the other hand, autoimmune thyroiditis
may develop after the “cytokine storm” induced by SARSCoV-2
infection, and may lead to primary hypothyroidism. Conversely,
pituitary disorder during COVID-19 infection may contribute to
pituitary insufficiency, resulting in thyroid and corticotropic insuf-
ficiency, requiring thyroid and corticoadrenal replacement therapy
with levothyroxine and hydrocortisone [14].
Diagnostic and/or therapeutic management of hypothyroidism
during SARS-CoV-2 infection does not require any other particular
consideration. Treatment of pre-existing hypothyroidism should be
continued during COVID-19 infection. The patient should be pro-
vided with the necessary supplies of thyroid hormones in order
509
to pursue treatment throughout lockdown. Levothyroxine dose
should be increased by 30–50% in case of pregnancy, and should
be monitored according to clinical data and TSH concentration,
specifically to prevent the development of hypothyroidism.
Hydroxychloroquine treatment, alone or as part of a combined
therapy [34] has been proposed as a treatment option for some
COVID-19 patients. It may impair thyroxine metabolism and there-
fore requires TSH levels to be monitored to ensure euthyroidism is
maintained [38,39].
In practice, thyroid function needs to be assessed both clini-
cally and hormonally, in the acute phase of a SARS-CoV-2 infection
and during follow-up, in order to initiate levothyroxine replace-
ment therapy if thyroid insufficiency is detected and to discontinue
replacement therapy once pituitary gland disorders, in particular,
resolve during convalescence [40].
6. Conclusion
Based on the pathophysiology of SARS-CoV-2 infection in the
pituitary-thyroid axis and a review of recent articles, we sug-
gest routine assessment of thyroid function in the acute phase
for COVID-19 patients requiring a high level of intensive care, as
they frequently present thyrotoxicosis due to subacute thyroiditis
related to SARS-CoV-2, and during convalescence in order to diag-
nose and adapt levothyroxine replacement treatment in patients
with primary or central hypothyroidism.
Considering the ongoing COVID-19 pandemic, future prospec-
tive studies are needed to increase epidemiological and clinical
knowledge and optimize the management of thyroid disorders in
COVID-19 patients.
Funding
The author did not receive any specific grant from any fund-
ing agency in the public, commercial or not-for-profit sector in the
preparation of this review.
Disclosure of interest
The author declare that they have no competing interest.
References
[1] Wu D, Yang XO. TH17 responses in cytokine storm of COVID-19: an emerging
target of JAK2 inhibitor Fedratinib. J Microbiol Immunol Infect 2020;53:368–70.
[2] Prompetchara E, Ketloy C, Palaga T. Immune responses in COVID-19 and poten-
tial vaccines: Lessons learned from SARS and MERS epidemic. Asian Pac J Allergy
Immunol 2020;38:1–9.
[3] Lania A, Sandri MT, Cellini M, Mirani M, Lavezzi E, Mazziotti G, et al. Thy-
rotoxicosis in patients with Covid-19: the THYRCOV study. Eur J Endocrinol
2020;183:381–7 [20-0335.R2].
[4] Wang W, Ye YX, Yao H. Evaluation and observation of serum thyroid hormone
and parathyroid hormone in patients with severe acute respiratory syndrome.
J Chin Antituberculous Assoc 2003;25:232–4.
[5] Fliers E, Bianco AC, Langouche L, Boelen A. Thyroid function in critically ill
patients. Lancet Diabetes Endocrinol 2015;3:816–25.
[6] Tee LY, Hajanto S, Rosario BH. COVID-19 complicated by Hashimoto’s thyroidi-
tis. Singapore Med J 2020:1–3.
[7] Wei L, Sun S, Xu C-H, Zhang J, Xu Y, Zhu H, et al. Pathology of the thyroid in
severe acute respiratory syndrome. Hum Pathol 2007;38:95–102.
[8] Guan WJ, Ni ZY, Hu Y, Liang WH, Ou CQ, He JX, et al. Clinical characteristics of
coronavirus disease 2019 in China. N Engl J Med 2020;382:1708–20.
[9] Yao XH, Li TY, He ZC, Ping YF, Liu HW, Yu SC, et al. A pathological report of three
COVID-19 cases by minimal invasive autopsies. Zhonghua Bing Li Xue Za Zhi
(Chinese J Pathol) 2020;49:411–7.
[10] Pal R, Banerjee M. COVID-19 and the endocrine system: exploring the unex-
plored. J Endocrinol Invest 2020;43:1027–31.
[11] Li MY, Li L, Zhang Y, Wang XS. Expression of the SARS-CoV-2 cell receptor gene
ACE2 in a wide variety of human tissues. Infect Dis Poverty 2020;9:45.
[12] Wei L, Sun S, Zhang J, Zhu H, Xu, Ma Q, et al. Endocrine cells of the adeno-
hypophysis in severe acute respiratory syndrome (SARS). Biochem Cell Biol
2010;88:723–30.
510 P. Caron / Annales d’Endocrinologie 81 (2020) 507–510
[13] Lazartigues E, Qadir MMF, Mauvais-Jarvis F. Endocrine significance of SARS-
CoV-2’s reliance on ACE2. Endocrinology 2020;161 [bqaa108].
[14] Leow MK-S, Kwek DS-K, Ng AW-K, Ong K-C, Kaw GJ-L, Lee LS-U. Hypocor-
tisolism in survivors of severe acute respiratory syndrome (SARS). Clin
Endocrinol 2005;63:197–202.
[15] Chen M, Zhou W, Xu W. Thyroid function analysis in 50 patients with COVID-19:
a retrospective study. Thyroid 2020.
[16] Chen T, Wu D, Chen H, Yan W, Yang D, Chen G, et al. Clinical characteristics
of 113 deceased patients with coronavirus disease 2019: retrospective study.
BMJ 2020;368:m1091.
[17] World Health Organization. Clinical management of severe acute respiratory
infection (SARI) when COVID-19 disease is suspected: interim guidance; 2020.
[18] Muller I, Cannavaro D, Dazzi D, Covelli D, Mantovani G, Muscatello A, et al. SARS-
CoV-2-related atypical thyroiditis. Lancet Diabetes Endocrinol 2020;8:739–41.
[19] Bellastella G, Maiorino MI, Esposito K. Endocrine complications of COVID-
19: what happens to the thyroid and adrenal glands? J Endocrinol Invest
2020;43:1169–70.
[20] Zhang JJ, Dong X, Cao YY, Yuan YD, Yang YB, Yan YQ, et al. Clinical charac-
teristics of 140 patients infected with SARS-CoV-2 in Wuhan, China. Allergy
2020;75:1730–41.
[21] Shekhar S, Wurth R, Kamilaris CDC, Eisenhofer G, Barrera FJ, Hajdenberg M,
et al. Endocrine conditions and COVID-19. Horm Metab Res 2020;52:471–84.
[22] Brancatella A, Ricci D, Cappellani D, Viola N, SgroD̀, Santini F, et al. Is subacute
thyroiditis an underestimated manifestation of SARS-CoV-2 infection? Insights
from a case series. J Clin Endocrinol Metab 2020;105:dgaa 537.
[23] Paunkovic N, Paunkovic J, Pavlovic O, Paunovic Z. The significant increase in
incidence of Graves’ disease in eastern Serbia during the civil war in the former
Yugoslavia (1992 to 1995). Thyroid 1998;8:37–41.
[24] Marazuela M, Giustina A, Puig-Domingo M. Endocrine and metabolic aspects
of the COVID-19 pandemic. Rev Endocr Metab Disord 2020;9:1–13.
[25] Bartalena L, Chiovato L, Marcocci C, Vitti P, Piantanida E, Tanda ML. Man-
agement of Graves’ hyperthyroidism and orbitopathy in time of COVID-19
pandemic. J Endocrinol Invest 2020;43:1149–51.
[26] Boelaert K, Visser WE, Taylor PN, Moran C, Léger J, Persani L. Endocrinology in
the time of COVID-19: management of hyperthyroidism and hypothyroidism.
Eur J Endocrinol 2020;183:G33–9.
[27] Griffith ML, Bischoff LA, Baum HBA. Approach to the patient with thyrotoxicosis
using telemedicine. J Clin Endocrinol Metab 2020;14:dgaa 373.
[28] Pal R, Bhadada SK. Managing common endocrine disorders amid COVID-19
pandemic. Diabetes Metab Syndr 2020;14:767–71.
[29] Brancatella A, Ricci D, Viola N, SgroD̀, Santini F, Latrofa F. Subacute thyroiditis
after Sars-COV-2 infection. J Clin Endocrinol Metab 2020;105:1–4.
[30] Desailloud R, Hober D. Viruses and thyroiditis: an update. Virol J 2009;6:5.
[31] Asfuroglu Kalkan E, Ates I. A case of subacute thyroiditis associated with Covid-
19 infection. J Endocrinol Invest 2020;43:1173–4.
[32] Ippolito S, Dentali F, Tanda ML. SARS-CoV-2: a potential trigger for subacute
thyroiditis? Insights from a case report. J Endocrinol Invest 2020;43:1171–2.
[33] Ruggeri RM, Campenni A, Siracusa M, Frazzetto G, Gullo D. Subacute thyroiditis
in a patient infected with SARS-COV-2: an endocrine complication linked to the
COVID-19 pandemic. Hormones 2020:1–3.
[34] Lam S, Lombardi A, Ouanounou A. COVID-19: a review of the proposed phar-
macological treatments. Eur J Pharmacol 2020;886:173451.
[35] Guo Y, Li H, Chen X, Yang H, Guan H, He X, et al. Novel roles of chloroquine
and hydroxychloroquine in Graves’ orbitopathy therapy by targeting orbital
fibroblasts. J Clin Endocrinol Metab 2020;105:1906–17.
[36] Bhatraju PK, Ghassemieh BJ, Nichols M, Kim R, Jerome KR, Nalla AK, et al. Covid-
19 in critically ill patients in the Seattle region. N Engl J Med 2020;382:2012–22.
[37] Dworakowska D, Grossman AB. Thyroid disease in the time of COVID-19.
Endocrine 2020;68:471–4.
[38] Al-Bari MAA, Islam MA. Clinically significant drug interaction profiles of chloro-
quine analogues with adverse consequences and risk management. J Sci Res
2015;7:177–95.
[39] Munera Y, Hugues FC, Le Jeunne C, Pays JF. Interaction of thyroxine sodium
with antimalarial drugs. BMJ 1997;314:1593.
[40] Agarwal S, Agarwal SK. Endocrine changes in SARS-CoV-2 patients and lessons
from SARS-CoV. Postgrad Med J 2020;96:412–6.
 Annales d’Endocrinologie 81 (2020) 511–515

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

Consensus

Management of thyroid dysfunctions in the elderly. French Endocrine

Society consensus 2019 guidelines. Short version夽
La version courte du consensus « Management of thyroid dysfunctions in the
elderly »
Bernard Goichota,∗ , Véronique Raverotb , Marc Kleinc , Lavinia Vija Racarud ,
Juliette Abeillon-du Payrate , Olivier Lairezf,g , Remy Leroyh , Anne Cailleuxi , Pierre Wolffj ,
Lionel Groussink , Georges Kaltenbachl , Philippe Caronm
a
Service de médecine interne, endocrinologie et nutrition, hôpitaux universitaires de Strasbourg, hôpital de Hautepierre, 1, avenue Molière, 67098
Strasbourg cedex, France
b
Laboratoire d’hormonologie, LBMMS, Groupement hospitalier est, hospices civils de Lyon, 69677 Bron cedex, France
c
Service EDN, CHU de Nancy, 54500 Vandoeuvre-les-Nancy, France
d
Service de médecine nucléaire, Institut universitaire de cancérologie de Toulouse oncopole, 31059 Toulouse, France
e
Groupement hospitalier est, Fédération d’endocrinologie, hospices civils de Lyon, Lyon, France
f
Fédération des services de cardiologie, CHU de Rangueil, Toulouse, France
g
Centre d’imagerie cardiaque, CHU de Toulouse, Toulouse, France
h
Cabinet d’endocrinologie et diabétologie, 71, rue de La Louvière, 59000 Lille, France
i
Endocrinologist, clinique Mathilde, 7, boulevard de l’Europe, 76100 Rouen, France
j
Espace santé, 8, rue de Lattre de Tassigny, 69350 La Mulatiere, France
k
Inserm U1016, CNRS UMR8104, service d’endocrinologie, université de Paris, Institut Cochin, hôpital Cochin, AP–HP, 75014 Paris, France
l
Pôle de gériatrie, hôpitaux universitaires de Strasbourg, hôpital de la Robertsau, 67000 Strasbourg, France
m
Service d’endocrinologie et maladies métaboliques, CHU de Larrey, 31059 Toulouse, France

The elderly population is classically defined by the World Health
Organization (WHO) and in the international literature, including in
endocrinology, as aged more than 65 years; “geriatrics” begins with
75 year-old patients with multiple pathology. The prevalence of
loss of functional autonomy due to comorbidities increases sharply
after 85 years of age.
In the French population, life-expectancy is increasing by 3
months every year, and the over-65-year-old population is now
18% of the general population, and 9% in over-75-year-old subjects
in France.
1. Q.1. Whom to screen for thyroid dysfunctions?
The symptoms of thyroid dysfunctions are relatively non-
specific in the elderly, who tend to show fewer symptoms overall
(Carlé et al. [1]). A high level of clinical vigilance is required not
to overlook thyroid dysfunction. However, this needs weighing
against the risk of over-prescription of thyroid examinations, which
often reveal transient biological abnormalities that are actually
secondary to other general diseases or medical treatments, and
unrelated to any thyroid pathology, and almost always without any
clinical impact, giving rise to unnecessary treatment and pointless
and costly iterative check-ups.

夽 French Endocrine Society consensus 2019 guidelines published in Ann

Endocrinol (Paris). Vol. 81, No 2–3, Juin 2020 S0003-4266(20)30067-6. doi: Guideline 1.1 Hormonal thyroid exploration is indicated in
10.1016/j.ando.2020.04.010.
∗ Corresponding author.
previously unknown atrial fibrillation, cognitive disorder of
recent onset, unexplained depression or other signs of thyroid
E-mail addresses: bernard.goichot@chru-strasbourg.fr (B. Goichot),
veronique.raverot@chu-lyon.fr (V. Raverot), m.klein@chru-nancy.fr (M. Klein),
dysfunction (Grade 1+++).
vija.l@chu-toulouse.fr (L. Vija Racaru), juliette.abeillon@chu-lyon.fr (J. Abeillon-du Guideline 1.2 Thyroid function screening is not indicated in
Payrat), lairez.o@chu-toulouse.fr (O. Lairez), remydoc2003@gmail.com (R. Leroy), the elderly (Grade 1++).
anne.cailleux@clinique-mathilde.fr (A. Cailleux), docteurwolff@orange.fr (P. Wolff),
lionel.groussin@aphp.fr (L. Groussin), Georges.KALTENBACH@chru-strasbourg.fr
(G. Kaltenbach), caron.p@chu-toulouse.fr (P. Caron).

https://doi.org/10.1016/j.ando.2020.05.002
0003-4266/© 2020 Published by Elsevier Masson SAS.
512 B. Goichot et al. / Annales d’Endocrinologie 81 (2020) 511–515
In case of symptoms prompting thyroid function assessment
that finds normal TSH level, there is no reason to repeat the assay
within 6 months or a year in the absence of new clinical events.
Guideline 1.3 In the absence of new clinical events, there is
no reason to repeat hormonal thyroid exploration that came
out normal (Grade 1+++).
Thyroid function tests are often non-specifically disturbed in
general non-thyroid diseases, especially in acute phase. Thyroid
function evaluation is better avoided in case of intercurrent pathol-
ogy without strong suspicion of thyroid dysfunction if a diagnosis of
thyroid dysfunction would alter the ongoing treatment (for exam-
ple, for recent atrial fibrillation). Thyroid evaluation work-up is
better performed outside of the acute episode. This is especially true
for hospitalized patients. There are no data determining a particular
interval to be respected, but work-up in practice is usually repeated
at a 4–12 month interval, although this depends on the context and
particularly on the progression of the thyroid dysfunction.
Guideline 1.4 Except if diagnosis of thyroid dysfunction
would alter ongoing treatment, thyroid function assessment
is best not performed during acute episode of intercurrent dis-
eases, and especially during unscheduled hospital stay (Grade
1+++).
2. Q.2. How to screen for thyroid dysfunction in the elderly?
Small changes in free T4 (FT4) concentration are accompanied
by much greater changes in TSH levels. This remains true in the
elderly, whatever the relation between the two (linear, sigmoid or
logarithmic) (Rothacker et al. [2]) and is a good reason for assaying
TSH concentration for thyroid exploration in this age group, except
in case of central pathology (hypothalamic or pituitary diseases).
Interpreting TSH in the elderly should always take account of
variations that may be due to medication or non-thyroid patholo-
gies.
Guideline 2.1 Diagnosis of thyroid dysfunctions is based on
TSH assay alone in first line (Grade 1+++).
Guideline 2.2 There is no reason in first line to perform
free T4 or free T3 assay, screen for antithyroid antibodies or
perform thyroid ultrasound scan (Grade 1++).
3. Q.3. What are the TSH reference values in the elderly?
The reference range for a biological parameter is a statistical
notion (range including 95% of subjects without the pathology in
question) and does not in itself determine what is pathological, and
should therefore not be used alone for treatment decision-making.
Guideline 3.1 The lower limit of normal for TSH assay is
usually 0.4 mIU/L. This threshold is relatively unaffected by
age (Grade 2+).
For the upper limit, the TSH distribution is skewed rightward,
and no universal threshold can be set. Given the above-mentioned
reservations, we propose a pragmatic solution: for over-60 year-
old patients, the upper limit of normal range is the patient’s age
(decade) divided by 10: e.g., TSH ≤ 7 for a 70 year-old, or ≤ 8 mIU/L
for an 80 year-old subject.
Guideline 3.2 The upper limit of normal range for TSH
increases with age. For simplicity, it is recommended in clin-
ical practice in over-60 year-old subjects to use the patient’s
age (decade) divided by 10: e.g., ≤ 8 mIU/L for an 80 year-old
subject (Grade 2+).
4. Q.4. What procedure in patients with low TSH level?
4.1. Q.4.1 What diagnostic procedure in patients with low TSH
level?
Guideline 4.1.1 Low TSH level should be systematically
assayed except in the context of clinical emergency (Grade
1++).
The inter-assay interval depends on the clinical situation, initial
TSH value and cardiovascular context.
Guideline 4.1.2 In case of TSH > 0.1 mIU/L, only TSH control
is recommended; in case of TSH < 0.1 mIU/L, control should
associate FT4 assay and FT3 assay if FT4 is normal (Grade 2+).
FT3 assay should be interpreted in the light of the clinical context
and ongoing medications.
4.2. Q.4.2 What procedure if TSH level later normalizes?
Guideline 4.2 There is no need for systematic control of TSH
in the absence of clinical signs (Grade 1++).
Very few patients with isolated low TSH level go on to develop
hyperthyroidism within 5 years, and most recover a normal thyroid
function.
In case of low TSH after iodine overload, any further iodine injec-
tion requires control of thyroid function test (cf. Guideline 4.6.2).
4.3. Q.4.3 What complementary examinations are needed in case
of confirmed thyrotoxicosis?
Thyrotoxicosis consists in the clinical and biological con-
sequences of excess thyroid hormone in tissues. The term
“hyperthyroidism” refers for situations of hyperfunction of the thy-
roid gland. In the elderly, the most frequent cause of thyrotoxicosis
is overdose of levothyroxine.
4.4. Q.4.4 What specificities of hyperthyroidism treatment and
follow-up in the elderly?
Autonomous pathology (multinodular goiter, toxic adenoma) is
more frequent than Graves’ disease. In elderly patients with Graves’
disease, prolonged synthetic antithyroid therapy is often unsuited.
 B. Goichot et al. / Annales d’Endocrinologie 81 (2020) 511–515
Guideline 4.3.1 Etiological assessment depends on the clin-
ical context and treatment objectives, but in most elderly
patients treatment options are relatively unaffected by the eti-
ology of the hyperthyroidism. Work-up in first line comprises
thyroid scintigraphy (Tc99 or I-123 as available) to provide etio-
logic and pre-treatment data (Grade 2+). It should not postpone
treatment initiation.
Guideline 4.3.2 The usefulness of anti-TSH receptor anti-
body assay is to be assessed according to context and the
likelihood of Graves’ disease (Grade 2+).
Guideline.4.3.3 Thyroid ultrasound scan should not be per-
formed in first line, as it is uninformative on etiology and is
associated with a risk of over-diagnosis of thyroid nodule. It
is to be reserved to particular situations: clinical abnormalities
on neck palpation or pre-treatment work-up (notably before
radioiodine therapy), or in the context of amiodarone-related
thyrotoxicosis (Grade 2+).
Definitive treatment is usually to be preferred, notably iodine131 if
not contraindicated.
Guideline 4.4.1 In elderly subjects with permanent atrial
fibrillation, tachy-arrhythmia, ischemic cardiopathy or risk of
acute coronary syndrome, or with cardiovascular symptoma-
tology triggered or aggravated by their hyperthyroidism, the
objective is definitive treatment, with iodine131 ablation in first
line after medical control of thyrotoxicosis (synthetic antithy-
roid drugs) (Grade 2++).
Guideline 4.4.2 In fragile patients without serious car-
diovascular pathology with moderate hyperthyroidism and
contraindications to total thyroidectomy or if iodine131 is
unavailable, long-course low-dose synthetic antithyroid drugs
may be used (Grade 2++).
Guideline 4.4.3 In the absence of contraindications, total
thyroidectomy is indicated in case of large goiter, signs of
compression and/or thyroid cancer (Grade 1++).
In over-65 year-old patients with large multinodular goiter
(>40–60 mL) and TSH < 0.4 mIU/L, radical treatment such as total
thyroidectomy is recommended in the absence of major cardiovas-
cular risk factors. If surgery is contraindicated, repeated low-dose
iodine131 can restore normal thyroid function and reduce goi-
ter volume (by 40% in 6–12 months). Surgery is also indicated
in some other situations: thyroid cancer, primary hyperparathy-
roidism (Kahaly et al. [3], Ross et al. [4], Biondi and Cooper [5]).
4.5. Q.4.5 What procedure in case of persistent low TSH level with
normal FT4?
This situation corresponds to subclinical hyperthyroidism. In
the literature, TSH level sometimes normalizes at a variable delay
of some months to some years, and progression to over or clinical
thyrotoxicosis is rare (Das et al. [6]). There is, however, a risk of
complications: notably cardiovascular with atrial fibrillation.
Guideline 4.5.1 Persistent low TSH level with normal
FT4 and FT3 concentrations requires endocrinologic opinion
(Grade 1++).
Guideline 4.5.2 In case of low TSH level with normal FT4 and
FT3 concentrations, thyroid scintigraphy is recommended to
identify autonomization (toxic adenoma, multinodular goiter)
and assess comorbidities and risk of complications, notably
cardiovascular and bone diseases (Grade 2+).
513
In the absence of evidence from randomized clinical trials, man-
agement is founded on expert consensus. The aim is more to
prevent complications such as atrial fibrillation (AF) than to treat
the disease itself: a large majority of patients are asymptomatic,
and their biological abnormality is revealed fortuitously.
4.6. Q.4.6 Should subclinical hyperthyroidism in the elderly be
treated?
Decision factors to treat or not to treat comprise: patient age,
etiology of hyperthyroidism, TSH level (< or >0.1 mIU/L), and risk
of complications (notably AF). The international guidelines (Euro-
pean Thyroid Association, American Thyroid Association) are not
fully concordant: roughly, the ETA recommends treating all “older”
subjects (>65 years) (Kahaly et al. [3]), while the ATA recommends
treating all over-65 year-old patients with TSH < 0.1 mIU/L (Ross
et al. [4]); for those with moderately decreased TSH level (0.1–0.4
mIU/L), treatment is indicated in case of osteoporosis, atrial fibril-
lation or cardiopathy.
In the absence of interventional studies assessing the
risk/benefit ratio, and taking account of:
• the lack of evidence that treating subclinical hyperthyroidism
significantly reduces the risk of atrial fibrillation;
• the risk of definitive hypothyroidism following treatment for
hyperthyroidism, and;
• several reports that achieving normal thyroid function during
levothyroxine treatment is difficult in the elderly, with high rates
of low or high TSH levels (Canaris et al. [7], Somvaru et al. [8]),
we recommend a cautious attitude weighing risk and benefit for
each individual.
Guideline 4.6.1 The decision between treatment and
surveillance for subclinical hyperthyroidism is to be discussed
with the individual patient and/or family, taking account of
expected benefit (basically, reduced risk of atrial fibrillation)
and possible risk (basically onset of definitive hypothyroidism).
Advanced age, the degree in TSH decrease (arbitrary
threshold generally set at 0.1 mIU/L), etiology of hyper-
thyroidism (autonomous pathologies with greater risk of
progression to over or clinical thyrotoxicosis) and comorbidi-
ties (cardiopathy or osteoporosis) are elements that encourage
the offer of treatment. Their absence prompts discussion with
the patient of regular monitoring (Grade 2+).
Following contrast medium injections, especially in patients
with advanced age, goiter or history of thyroid dysfunction, there
is clear risk either of thyrotoxicosis episodes or of inducing hyper-
thyroidism.
The literature data fail to identify any particular protocol for
preventing iodine overload due to contrast medium injections. A
single prospective randomized study reported preventive benefit
for synthetic antithyroid drugs and sodium perchlorate (Nolte et al.
[9]), providing rapid urinary elimination of the iodine excess and
thus avoiding overload.
A few days’ treatment around the radiology examination, using
synthetic antithyroid drugs or sodium perchlorate (if available)
could be implemented, with empiric dosage and duration, accord-
ing to the 2010 Société française d’endocrinologie – Groupe de
recherche de thyroïde – Société française de radiologie guidelines
for iodized contrast medium injection in thyroid pathology. Thyroid
function should be checked a few weeks after the radiology exam-
ination. If contrast medium is expected to be re-administered, as
514 B. Goichot et al. / Annales d’Endocrinologie 81 (2020) 511–515
in oncologic surveillance, and the risk of thyrotoxicosis is signifi-
cant, iodine131 ablation may be considered some time after contrast
medium injection.
Guideline 4.6.2 In case of known thyroid pathology, and
notably multinodular goiter, and especially if TSH level is
transiently low, prophylactic treatment may be implemented
around iodized contrast medium administration (Grade 2+).
Guideline 4.6.3 When a form of treatment has been chosen,
modalities are the same as for over hyperthyroidism (Grade
2++).
Guideline 4.6.4 If surveillance without treatment is chosen,
it is based on TSH and FT4 monitoring (plus FT3 if FT4 is nor-
mal) at 3 months then every 6 months, during all life. A clinical
check-up should also be made, notably to ensure maintained
sinus rhythm (Grade 2+).
Guideline 4.6.5 Amiodarone is a frequent cause of thyro-
toxicosis (and hypothyroidism) in the elderly. Treatment does
not differ from that in younger patients (Grade 2++). However,
given the potential severity of thyrotoxicosis and the complex-
ity of management, an endocrinologic opinion is indispensable
(Grade 2+++).
Management of thyroid dysfunctions occurring during amio-
darone treatment was the focus of very recent ETA guidelines
(Bartalena et al. [10]). This will not be dealt in this manuscript, as
patient age does not affect diagnosis or treatment modalities.
5. Q.5 What procedure in case of high TSH level in the
elderly?
5.1. Q.5.1 What diagnostic attitude in case of high TSH level?
Guideline 5.1 All high TSH level should be controlled: within
the month in case of clinical signs, and within 3 months in the
absence of symptoms or if the TSH level is < 10 mIU/L (Grade
2+).
In the Framingham study, prevalence of hypothyroidism
(TSH > 10 mIUL) was 4.4% in over-60 year-old patients (5.9% in
women, 2.3% in men) (Sawin et al. [11]).
5.2. Q.5.2 What examinations are mandatory or contributive in
case of high TSH level (checked on repeat assay?
No examinations alter treatment for elderly patients with high
TSH level: antithyroid antibody screening, FT4 assay, lipid profile,
or thyroid ultrasound (except in case of abnormal neck palpation).
On the practical level, no additional examination modifies the
management of elderly patients with high TSH level, in particular:
the search for antithyroid antibodies, the free T4 assay, the lipid
profile, thyroid ultrasound (except abnormality of cervical palpa-
tion). These examinations are useless and are not recommended in
practice in the elderly with hypothyroidism.
Guideline 5.2 No complementary examinations are needed
for high TSH level in elderly patients (Grade 2+).
5.3. Q.5.3 What are the indications for treating hypothyroidism?
Subclinical hypothyroidism has less impact in the elderly than in
younger subjects (Barbesino [12]). Several older studies, with rela-
tively small series, demonstrated normalization of TSH levels over
time in about half of patients with “subclinical hypothyroidism”
(Tabatabaie and Surks [13], Mooijaart et al. [14]).
Guideline 5.3.0 Levothyroxine therapy should be initiated
only when hypothyroidism is confirmed by high TSH level on
two assays (Grade 2+++).
Although formal proof is lacking, the present group recom-
mends levothyroxine therapy when TSH is > 20 IU/L in two samples.
In the absence of proven benefit (Stott et al. [15]), given the risk
of thyrotoxicosis, the group advises against treatment when TSH
is < 10 mIU/L on several successive controls. When TSH is 10–20
mIU/L, treatment should be considered on a case-by-case basis
with the patient and family, depending on context, comorbidities,
clinical signs, expected benefit, baseline TSH and progression over
successive controls.
Guideline 5.3.1 Levothyroxine replacement therapy is indi-
cated in case of TSH > 20 mIU/L in at least two controls (Grade
2+).
Guideline 5.3.2 There is no evidence of favorable
risk/benefit ratio for replacement therapy in case of TSH < 10
mIU/L (Grade 2+).
Guideline 5.3.3 In patients with TSH level between 10–20
mIU/L on several assays, levothyroxine replacement therapy
should be considered on a case-by-case basis, taking account
of the patient’s wishes, expected benefit and TSH progression
(Grade 2+).
5.4. Q.5.4 What are the follow-up modalities during
hypothyroidism treatment in the elderly?
Guideline 5.4.1 Treatment is based on levothyroxine at a
substitutive dose of 1.1–1.3 ␮g/kg/day (Grade 2+).
Guideline 5.4.2 There are no indications for associating
levothyroxine + triodothyronine (T3) (Grade 2+++).
Guideline 5.4.3 Levothyroxine should be introduced pro-
gressively, especially if TSH is significantly elevated and the
patient’s cardiovascular status is unknown (Grade 1+++).
Guideline 5.4.4 Surveillance of levothyroxine treatment for
primary thyroid failure is based on TSH alone. Checks should
be conducted between 6 weeks and 3 months after achieving
replacement dose or any dose adaptation (Grade 1+++).
Guideline 5.4.5 The treatment objective is to have TSH level
within the normal-for-age range (cf. Guideline 3.2), avoiding
risk of overdose if TSH < 1 mIU/L; e.g., the upper limit of normal
is 7 mIU/L for 70–79 years and 8 mIU/L for > 80 years (Grade 2+).
Guideline 5.4.6 Once normal thyroid function is achieved,
annual checks of TSH level are sufficient (Grade 2+).
Guideline 5.4.7 When a treatment known to impact levothy-
roxine bioavailability is introduced or withdrawn, TSH level
should be checked after 6 weeks (Grade 2++).
Guideline 5.4.8 If levothyroxine therapy is not indicated,
TSH level should be assayed every 6 months for 2 years to
assess the kinetics of progression. If TSH level is stable, annual
checks are sufficient (Grade 2+).
 B. Goichot et al. / Annales d’Endocrinologie 81 (2020) 511–515
6. Q.6 Can levothyroxine therapy be withdrawn?
Hypothyroidism is usually definitive. In the elderly, some
patients have often been taking levothyroxine for many years,
frequently at low-doses, and the initial reasons for prescription
may be unknown. It may sometimes be justifiable to try reducing
and withdrawing a non-beneficial treatment, although obviously
making sure that there was no formal indication of levothyrox-
ine treatment (total thyroidectomy, previous iodine131 therapy,
proven auto-immune hypothyroidism), informing the patient and
securing consent. Treatment may then be withdrawn, but a TSH
assay should be check at 4–6 weeks.
Guideline 6 If the indication for levothyroxine therapy is not
clear, withdrawal may be attempted, with the patient’s consent
and TSH check at 4–6 weeks (Grade 2+).
7. Q.7 When should an endocrinologic opinion be sought?
Whether endocrinologic opinion should be sought depends on
the context and is at the physician’s discretion. Given the possible
complexity of situations and treatment considerations, the present
group recommends endocrinologic opinion for all patients with
prolonged low TSH level. Specialist opinion should also be sought
in the following situations:
• at diagnosis:
◦ signs of severity (coronary context, auto-immune poly-
endocrinopathy, over hypothyroidism, myxedema coma, etc.),
◦ discordance between clinical and biological findings,
◦ according to etiology: iatrogenic, central hypothyroidism,
◦ discussion of treatment objectives;
• during follow-up:
◦ doubt regarding treatment objectives,
◦ difficulty in interpreting results (central hypothyroidism),
◦ onset of unfavorable symptoms,
◦ lack of improvement (treatment resistance, drug interactions,
malabsorption, poor adherence, etc.),
◦ poor tolerance,
◦ need to change levothyroxine formulation (e.g., impossibility
of oral administration).
Guideline 7 Endocrinologic opinion should be sought in
proven hyperthyroidism or persistent low THS level. Grade
2++.
515
Disclosure of interest
J. Abeillon: HAC, Sanofi, Uni-pharma.
A. Cailleux: Sanofi, HAC.
P. Caron: Merck-Serono, HAC, Genevrier, Uni-pharma.
B. Goichot: Merck-Serono, Uni-pharma, HAC.
L. Groussin: Merck-Serono, HAC.
M. Klein: Merck-Serono, Genzyme, HAC, Sanofi.
R. Leroy: Merck-Serono, HAC, Sanofi.
V. Raverot: Abbott Diagnostic, Roche Diagnostic, IDS, HAC.
L. Vija: Merck-Serono, Genzyme, HAC, Sanofi.
The other authors declare that they have no competing interest.
References
[1] Carlé A, Pedersen IB, Knudsen N, Perrild H, Ovesen L, Andersen S, et al.
Hypothyroid symptoms fail to predict thyroid insufficiency in old people: a
population-based case-control study. Am J Med 2016;129:1082–92.
[2] Rothacker KM, Brown SJ, Hadlow NC, Wardrop R, Walsh JP. Reconciling
the log-linear and non-log-linear nature of the TSH-Free T4 relationship:
intra-individual analysis of a large population. J Clin Endocrinol Metab
2016;101:1151–8.
[3] Kahaly GJ, Bartalena L, Hegedüs L, Leenhardt L, Poppe K, Pearce SH. 2018
European Thyroid Association Guideline for the management of graves’ hyper-
thyroidism. Eur Thyroid J 2018;7:167–86.
[4] Ross DS, Burch HB, Cooper DS, Greenlee MC, Laurberg P, Maia AL, et al.
2016 American Thyroid Association Guidelines for diagnosis and man-
agement of hyperthyroidism and other causes of thyrotoxicosis. Thyroid
2016;26:1343–421.
[5] Biondi B, Cooper DS. Subclinical hyperthyroidism. N Engl J Med
2018;378:2411–9.
[6] Das G, Ojewuyi TA, Baglioni P, Geen J, Premawardhana LD, Okosieme OE. Serum
thyrotrophin at baseline predicts the natural course of subclinical hyperthy-
roidism. Clin Endocrinol (Oxf) 2012;77:146–51.
[7] Canaris GJ, Manowitz NR, Mayor G, Ridgway EC. The Colorado thyroid disease
prevalence study. Arch Intern Med 2000;160:526–34.
[8] Somwaru LL, Arnold AM, Joshi N, Fried LP, Cappola AR. High frequency of
and factors associated with thyroid hormone over-replacement and under-
replacement in men and women aged 65 and over. J Clin Endocrinol Metab
2009;94:1342–5.
[9] Nolte W, Muller R, Siggelkow H, Emrich D, Hufner M. Prophylactic application of
thyrostatic drugs during excessive iodine exposure in euthyroid patients with
thyroid autonomy: a randomized study. Eur J Endocrinol 1996;134:337–41.
[10] Bartalena L, Bogazzi F, Chiovato L, Hubalewska-Dydejczyk A, Links TP, Van-
derpump M. 2018 European Thyroid Association (ETA) guidelines for the
management of amiodarone-associated thyroid dysfunction. Eur Thyroid J
2018;7:55–66.
[11] Sawin CT, Castelli WP, Hershman JM, McNamara P, Bacharach P. The aging
thyroid. Thyroid deficiency in the Framingham Study. Arch Intern Med
1985;145:1386–8.
[12] Barbesino G. Thyroid function changes in the elderly and their relationship to
cardiovascular health: a mini-review. Gerontology 2019;65:1–8.
[13] Tabatabaie V, Surks MI. The aging thyroid. Curr Opin Endocrinol Diabetes Obes
2013;20:455–9.
[14] Mooijaart SP, Du Puy RS, Stott DJ, Kearney PM, Rodondi N, Westendorp RGJ, et al.
Association between levothyroxine treatment and thyroid-related symptoms
among adults aged 80 years and older with subclinical hypothyroidism. JAMA
2019;322:1977–86.
[15] Stott DJ, Rodondi N, Kearney PM, et al. Thyroid hormone therapy for older adults
with subclinical hypothyroidism. N Engl J Med 2017;376:2534–44.
 Annales d’Endocrinologie 81 (2020) 516–519
Disponible en ligne sur
ScienceDirect
www.sciencedirect.com
Letter to the Editor
Ectopic adrenocortical carcinoma of the ovary: An unex-
pected outcome
Un corticosurrénalome ovarien au devenir exceptionnel
a r t i c l e i n f o
Keywords:
Adrenocortical carcinoma
Ovary
Pregnancy
Mitotane
Prognosis
Mots clés :
Corticosurrénalome
Ovaire
Grossesse
Mitotane
Prognostique
Steroid cell tumours are a rare subgroup of sex cord-stromal
tumours of the ovary. First described by Scully RE [1], this cate-
gory traditionally included Leydig cell tumour, stromal luteoma and
steroid cell tumour, not otherwise specified (NOS) [2]. Steroid cell
tumours are often androgenic and can be associated with Cushing’s
syndrome, mimicking an adrenocortical carcinoma. Indeed, the
ovary may contain ectopic adrenal cells originating from gonadal
and adrenal tissues of the genital ridge [1]. Adrenocortical neo-
plasms of the ovary are exceptionally rare; their prevalence and
prognosis, and the optimal follow-up, are poorly documented. We
report the case of a young girl with a metastatic, ectopic ovarian
adrenocortical neoplasm. In 1991, when she was 4 years of age, the
patient rapidly (within 1 month) developed signs of early puberty
(breasts plus pubic and armpit hair); this prompted her parents
to consult. Laboratory investigation showed increased levels of the
sex hormones estrone, 17-OH-progesterone, and testosterone. Low
levels of pituitary gonadotropins suggested autonomous secretion
of sex hormones from ovarian and/or adrenal origin. At that time, no
pelvic imaging was performed. Triptorelin (a gonadotropin releas-
ing hormone analogue) was prescribed and maintained between
the ages of 4 and 8 years. Menarche occurred at 9 years of age.
Genetic investigations did not identify TP53 gene mutation rulling
out a Li-Fraumeni syndrome. At 15 years of age, the patient reported
abdominal pain and, since a computed tomography (CT) revealed
a mass in the right ovary, aright salpingo-oophorectomy was
performed. Hormonal investigation was not considered. Anato-
mopathological examination revealed a tumour 20 cm in diameter
with large cells, pale and oxyphilic granular cytoplasm, rounded
dystrophic nuclei, and rare mitotic figures. Spans, cords, and mas-
sifs were evident within a vessel-rich endocrinoid stroma. These
structures were in contact with, but did not penetrate, the capsule.
Haemorrhagic foci were observed. Expert analysis by a referral cen-
tre identified an ovarian steroid tumour. There were > 2 mitoses/10
high-power fields but data on Ki67 and the Weiss score were lack-
ing. The patient was regularly examined; no notable event occurred
prior to the age of 17 years, at which time she was started on oral
contraceptives (cyproterone acetate and ethynylestradiol) because
the left ovary was multicystic. Appropriate follow-up was sched-
uled; there was no evidence of malignant transformation.
At 21 years of age, the patient was admitted to our unit
for the first time, with a persistent cough; CT revealed multiple
pulmonary tumours and positron emission tomography (PET) con-
firmed that the tumours were active (Fig. 1A). Biopsy revealed
pulmonary metastasis of the right ovarian carcinoma that was
diagnosed 17 years earlier. Following a multidisciplinary consul-
tation, the decision was made to prescribe adrenolytic medication
(mitotane at an initial dose of 2 g/day). As this was relatively well
tolerated, the dose was rapidly increased to 6 g/day to achieve
a serum level of 14–20 mg/L. Some clinical features were sug-
gestive of Cushing’s syndrome, including obesity (weight: 119 kg,
height: 154 cm, body mass index 50 kg/m2 ), acanthosis nigricans,
and a “buffalo neck”. Plasma cortisol analysis was inappropriate
due to the use of oestroprogestative contraception; unfortunately,
urinary free cortisol data were unavailable. A few months later,
a regular menstrual cycle returned, with notable loss of body
hair. Her weight fell from 119 to 87 kg. Testosterone was unde-
tectable, which remains the case at the time of this report.
Dehydroepiandrosterone-sulfate (DHEAS) serum level remained
normal over the next 10 years. Serum dehydroepiandrosterone-
sulfate (DHEAS) level remained normal over the next ten years.
PET showed that the lung metastases progressively declined; no
other metastasis was identified during regular check-ups (Fig. 1B).
Complete remission was achieved in 2014, 5 years after mitotane
initiation; pulmonary metastasis radiologically disappeared. Over
the years, mitotane was less tolerated with side effects associating
asthenia, gastrointestinal complaints (nausea, diarrhoea), together
with many episodes of acute adrenal insufficiency despite high-
dose of hydrocortisone. The patient expressed a strong desire to
become pregnant and wished to stop treatment. After consider-
ing the risks and benefits, mitotane was discontinued in July 2019.
At the time, her testosterone and (DHEAS) levels remained low,
and there was no sign of disease recurrence on the last PET scan
performed in January 2020 (Fig. 1).
To the best of our knowledge, only six cases of ectopic ovarian
adrenocortical neoplasms have been reported [3–7]. Most patients
died between 2 days and 17 months after initial diagnosis. One
73-year-old woman developed recurrence and peritoneal metas-
tases 12 months after laparoscopic ovariectomy; she had not been
prescribed adjuvant treatment. Given the lack of valid data, the
treatment for ectopic adrenocortical carcinomas follows similar
recommendations to that for eutopic tumours. The European
Society of Endocrinology considers surgery mandatory, and
mitotane is approved both in an adjuvant setting and for patients
 Letter to the Editor / Annales d’Endocrinologie 81 (2020) 516–519
Fig. 1. A. PET-scanner in 2009 showing active pulmonary metastasis; B. PET-scanner in 2011 showing regression of pulmonary lesions.
with advanced disease [8]. A combination of etoposide, doxoru-
bicin, cisplatin, and mitotane is among the therapeutic options
for patients with residual, advanced adrenocortical carcinomas
after first-line therapy. One patient with an ectopic adrenocortical
carcinoma, who did not respond to conventional chemotherapy,
was unsuccessfully treated with ketoconazole [5] and received,
as the first case report, mitotane monotherapy. The follow-up
duration was 10 years and she exhibited a complete response.
Similarly, in our case, the medical history is suggestive of a slowly
progressive tumour. Despite the initial misdiagnosis, 17 years
passed between first symptoms and metastases occurrence.
When evaluating prognostic factors, current recommendations
by the European Network for the Study of Adrenal Tumors (ENSAT)
suggest considering classification of the tumour at initial diagno-
sis with evaluation of tumour stage, resection status, Ki67 index,
as well as autonomous cortisol secretion [8]. Ki67 index is miss-
ing in our case report. In addition, although, the patient lacked
the specific features of Cushing’s syndrome, weight loss during
mitotane treatment suggested that the drug might have controlled
a likely hypercortisolism. It has been reported that, mitotane con-
trols hypercortisolism over the long term but that the antisecretory
effects requires several months to appear; they are sustained there-
after because the drug is stored in adipose tissue. Mitotane has
demonstrated efficacy in adrenocortical carcinoma and it is used to
treat all forms of hypercortisolism [8]. Data on mitotane monother-
apy in advanced disease are scarce. Megerle et al. studied a large
cohort (127 patients) with advanced adrenocortical carcinomas, of
whom 20.5% showed an objective response (complete or partial
response, or stable disease). Only three patients (2.4%) exhibited
complete remission [9], emphasizing the exceptional clinical evo-
lution of our case. Interestingly, mitotane blood levels were found
to be correlated with the objective response rate, progression-
free survival and overall survival [9,10]. The European Society
of Endocrinology Guidelines recommend that the mitotane blood
level should remain over 14 mg/L [8]; this was always the case
for our patient. A complete response was achieved 5 years after
mitotane initiation, emphasising the long-lasting effects of the
drug. In the study of Megerle et al., long-term disease control (> 180
days) was achieved in 40.9% of patients; long-term benefits (> 12
months) were seen in 22% of patients. Mitotane treatment was
stopped 5 years after our patient achieved a complete response. The
517
future course of our patient is not predictable and indeed there are
no guidelines pertaining to the optimal time for mitotane discon-
tinuation. Hermsen et al. reported extremely long survival (12–28
years) after initial diagnosis of adrenal carcinoma in six patients,
despite recurrences and metastases; their patients had been off
mitotane for many years prior to the study [11]. Our patient wished
to become pregnant; thus, mitotane treatment was discontinued
because the drug is teratogenic. Pregnancy should be avoided until
mitotane becomes undetectable in the plasma, which usually takes
1 to 3 years. Long-term remission is advisable before attempt-
ing pregnancy, but the optimal duration of remission remains to
be determined. Pregnancy after monitoring the evolution of an
adrenocortical carcinoma must be monitored cautiously. Indeed,
adrenal carcinogenesis is affected by the hormonal milieu of preg-
nancy. An adrenocortical carcinoma diagnosed during pregnancy or
the immediate postpartum period may be associated with a poorer
prognosis than other such carcinomas, because the tumours are
more advanced and often associated with hypercortisolism with
a poor overall maternal survival [12]. However, in a study on 17
pregnancies patients previously treated for adrenocortical carcino-
mas, pregnancy did not seem to be associated with poorer clinical
outcomes, similarly to breast cancer, another hormone-dependent
cancer ([13]. Interestingly, according to the largest relevant study
to date, since adrenocortical carcinomahas been localized in all
patients,a “healthy mother effect” is suggested, women in good
health beeing more willing to attempt pregnancy [14]. There has
been no report of pregnancy after development of an ectopic ovar-
ian adrenocortical carcinoma. Regarding the foetal outcome, De
Corbière et al. reported no adverse effects when mitotane has been
discontinued for at least one year before the time of conception,
but data on mitotane plasma levels were missing [14].
In conclusion, the survival of our patient (who is now aged 33
years) was unexpected; this is an extraordinary outcome of an
exceptional disease and the first case of ectopic ovarian adreno-
cortical carcinoma with a survival exceeding 10 years.
Patient consent
Informed consent has been obtained from the patient for publi-
cation of the case report and accompanying images.
518 Letter to the Editor / Annales d’Endocrinologie 81 (2020) 516–519
Disclosure of interest
The authors declare that they have no competing interest.
This research did not receive any specific grant from any funding
agency in the public, commercial or not-for-profit sector.
Data sharing is not applicable to this article as no new data were
created or analyzed in this case study.
References
[1] Scully RE. Sex cord-stromal tumors. In: Blaustein A, editor. Pathology
of the female genital tract. New York, NY: Springer; 1982. p. 581–601,
http://dx.doi.org/10.1007/978-1-4757-1767-9 23.
[2] Hayes MC, Scully RE. Ovarian steroid cell tumors (not otherwise specified).
A clinicopathological analysis of 63 cases. Am J Surg Pathol 1987;11:835–45,
http://dx.doi.org/10.1097/00000478-198711000-00002.
[3] Marieb NJ, Spangler S, Kashgarian M, Heimann A, Schwartz ML,
Schwartz PE. Cushing’s syndrome secondary to ectopic cortisol produc-
tion by an ovarian carcinoma. J Clin Endocrinol Metab 1983;57:737–40,
http://dx.doi.org/10.1210/jcem-57-4-737.
[4] Young RH, Scully RE. Ovarian steroid cell tumors associated with cush-
ing’s syndrome: a report of three cases. Int J Gynecol Pathol 1987;6:40–8,
http://dx.doi.org/10.1097/00004347-198703000-00005.
[5] Donovan JT, Otis CN, Powell JL, Cathcart HK. Cushing’s syndrome secondary
to malignant lipoid cell tumor of the ovary. Gynecol Oncol 1993;50:249–53,
http://dx.doi.org/10.1006/gyno.1993.1202.
[6] Elhadd TA, Connolly V, Cruickshank D, Kelly WF. An ovarian lipid cell tumour
causing virilization and Cushing’s syndrome. Clin Endocrinol 1996;44:723–5,
http://dx.doi.org/10.1046/j.1365-2265.1996.693515.x.
[7] Chentli F, Terki N, Azzoug S. Ectopic adrenocortical carci-
noma located in the ovary. Eur J Endocrinol 2016;175:K17–23,
http://dx.doi.org/10.1530/EJE-16-0224.
[8] Fassnacht M, Dekkers O, Else T, Baudin E, Berruti A, de Krijger R, et al. European
Society of Endocrinology Clinical Practice Guidelines on the management of
adrenocortical carcinoma in adults, in collaboration with the European Net-
work for the Study of Adrenal Tumors. Eur J Endocrinol 2018;179:G1–46,
http://dx.doi.org/10.1530/EJE-18-0608.
[9] Megerle F, Herrmann W, Schloetelburg W, Ronchi CL, Pulzer A, Quin-
kler M, et al. Mitotane monotherapy in patients with advanced
adrenocortical carcinoma. J Clin Endocrinol Metab 2018;103:1686–95,
http://dx.doi.org/10.1210/jc.2017-02591.
[10] Gonzalez RJ, Tamm EP, Ng C, Phan AT, Vassilopoulou-Sellin R, Per-
rier ND, et al. Response to mitotane predicts outcome in patients
with recurrent adrenal cortical carcinoma. Surgery 2007;142:867–75,
http://dx.doi.org/10.1016/j.surg.2007.09.006 [discussion 867-875].
[11] Hermsen IGC, Gelderblom H, Kievit J, Romijn JA, Haak HR. Extremely long sur-
vival in six patients despite recurrent and metastatic adrenal carcinoma. Eur J
Endocrinol 2008;158:911–9, http://dx.doi.org/10.1530/EJE-07-0723.
[12] Abiven-Lepage G, Coste J, Tissier F, Groussin L, Billaud L, Dousset
B, et al. Adrenocortical carcinoma and pregnancy: clinical and bio-
logical features and prognosis. Eur J Endocrinol 2010;163:793–800,
http://dx.doi.org/10.1530/EJE-10-0412.
[13] de Simone V, Pagani O. Pregnancy after breast cancer:
hope after the storm. Minerva Ginecol 2017;69:597–607,
http://dx.doi.org/10.23736/S0026-4784.17.04113-2.
[14] de Corbière P, Ritzel K, Cazabat L, Ropers J, Schott M, Libé R, et al. Pregnancy in
women previously treated for an adrenocortical carcinoma. J Clin Endocrinol
Metab 2015;100:4604–11, http://dx.doi.org/10.1210/jc.2015-2341.
Laurence Salle ∗
Robin Mas
Marie-Pierre Teissier-Clément
Department of Endocrinology, Diabetology and
Nutrition, University Hospital of Limoges, 2, avenue
Martin-Luther-King, 87042 Limoges, France
∗ Corresponding
author.
E-mail address: laurence.teyssieres@orange.fr
(L. Salle)
https://doi.org/10.1016/j.ando.2020.07.1112
0003-4266/ © 2020 Elsevier Masson SAS. All rights reserved.
Recurrent hypothyroidism and thrombopenic throm-
botic purpura
Hypothyroïdie et purpura thrombotique thrombocytopénique
a r t i c l e i n f o
Keywords:
Hypothyroidism
Thrombotic microangiopathy
Mots clés :
Hypothyroïdie
Microangiopathie thrombotique
Thrombotic thrombocytopenic purpura (TTP) is a rare and
life-threatening thrombotic microangiopathy characterized by
microangiopathic hemolytic anemia, severe thrombocytopenia,
and organ ischemia linked to disseminated microvascular platelet
rich-thrombi [1]. TTP is caused by deficiency of a plasma metallo-
protease, ADAMTS13. When exposed to high shear stress in the
microcirculation, von Willebrand factor (VWF) and platelets are
prone to form aggregates. This propensity of VWF and platelet
to form microvascular thrombosis is mitigated by ADAMTS13,
which cleaves VWF before it is activated by shear stress to cause
platelet aggregation in the circulation. Deficiency of ADAMTS13,
due to autoimmune inhibitors in patients with acquired TTP and
mutations of the ADAMTS13 gene in hereditary cases, leads to
VWF – platelet aggregation and microvascular thrombosis. TTP
shows striking involvement of myocardial arteries and variable
degrees of vascular involvement in kidney, pancreas, brain, and
adrenal glands [2]. We have previously reported the case of a
patient in which reversible hypothyroidism was observed in the
course of TTP [3]. A second episode of hypothyroidism is reported in
the same patient during a relapse of thrombotic microangiopathy.
A 42-year-old man was admitted to the hospital in Novem-
ber 1997 because of headache, lethargy, fatigue and mild right
arm weakness. During the preceding 10 days, he had had five
transient neurologic ischemic attacks. Laboratory findings were as
following: thrombocytopenia (platelet count 27 giga/L); mechani-
cal hemolytic anemia (hemoglobin 9.7 g/dL, schizocytes 2%, lactate
dehydrogenase (LDH) 842 U/L), glycaemia 4,8 mmol/L, serum crea-
tinine 104 ␮mol/L. Head CT injected with iodine did not show any
acute or subacute infarction or hemorrhage. TTP was diagnosed and
after five plasma exchanges and 2 mg vincristine injection, remis-
sion was achieved and the patient was discharged from the hospital.
There was no evidence of an underlying disorder. Thyroid function
tests performed because of unusual asthenia at presentation before
plasma exchange and head CT disclosed mild hypothyroidism (TSH
17 mIU/mL, normal range < 5 mIU/mL). Thyroid gland was clinically
normal. Serum was negative for thyroid autoantibodies. Because
the clinical presentation of hypothyroidism was poor, we kept
watch without specific therapy. Thyroid function returned to nor-
mal value after 6 months.
The patient remained in remission for 20 years and then pre-
sented to our department with paresthesia in upper right limb
and dysarthria related to a limited ischemic stroke on magnetic
resonance imaging. Biological findings showed mild hemolytic ane-
mia (hemoglobin: 10.5 g/dL, thrombocytopenia (platelet count: 60
giga/L, and schistocytes (4%) at blood film examination. The LDH
level was 450 U/L (N 135–325 U/L), haptoglobin was < 0.1 g/L, cre-
atinine 84 ␮mom/L. On admission, TSH was slightly increased,
7.4 mIU/mL, (normal range 0.270–4.20), T4 was normal.
A diagnosis of relapsing autoimmune TTP was rapidly estab-
lished, based on the presence of microangiopathic hemolytic
518 Letter to the Editor / Annales d’Endocrinologie 81 (2020) 516–519
Disclosure of interest
The authors declare that they have no competing interest.
This research did not receive any specific grant from any funding
agency in the public, commercial or not-for-profit sector.
Data sharing is not applicable to this article as no new data were
created or analyzed in this case study.
References
[1] Scully RE. Sex cord-stromal tumors. In: Blaustein A, editor. Pathology
of the female genital tract. New York, NY: Springer; 1982. p. 581–601,
http://dx.doi.org/10.1007/978-1-4757-1767-9 23.
[2] Hayes MC, Scully RE. Ovarian steroid cell tumors (not otherwise specified).
A clinicopathological analysis of 63 cases. Am J Surg Pathol 1987;11:835–45,
http://dx.doi.org/10.1097/00000478-198711000-00002.
[3] Marieb NJ, Spangler S, Kashgarian M, Heimann A, Schwartz ML,
Schwartz PE. Cushing’s syndrome secondary to ectopic cortisol produc-
tion by an ovarian carcinoma. J Clin Endocrinol Metab 1983;57:737–40,
http://dx.doi.org/10.1210/jcem-57-4-737.
[4] Young RH, Scully RE. Ovarian steroid cell tumors associated with cush-
ing’s syndrome: a report of three cases. Int J Gynecol Pathol 1987;6:40–8,
http://dx.doi.org/10.1097/00004347-198703000-00005.
[5] Donovan JT, Otis CN, Powell JL, Cathcart HK. Cushing’s syndrome secondary
to malignant lipoid cell tumor of the ovary. Gynecol Oncol 1993;50:249–53,
http://dx.doi.org/10.1006/gyno.1993.1202.
[6] Elhadd TA, Connolly V, Cruickshank D, Kelly WF. An ovarian lipid cell tumour
causing virilization and Cushing’s syndrome. Clin Endocrinol 1996;44:723–5,
http://dx.doi.org/10.1046/j.1365-2265.1996.693515.x.
[7] Chentli F, Terki N, Azzoug S. Ectopic adrenocortical carci-
noma located in the ovary. Eur J Endocrinol 2016;175:K17–23,
http://dx.doi.org/10.1530/EJE-16-0224.
[8] Fassnacht M, Dekkers O, Else T, Baudin E, Berruti A, de Krijger R, et al. European
Society of Endocrinology Clinical Practice Guidelines on the management of
adrenocortical carcinoma in adults, in collaboration with the European Net-
work for the Study of Adrenal Tumors. Eur J Endocrinol 2018;179:G1–46,
http://dx.doi.org/10.1530/EJE-18-0608.
[9] Megerle F, Herrmann W, Schloetelburg W, Ronchi CL, Pulzer A, Quin-
kler M, et al. Mitotane monotherapy in patients with advanced
adrenocortical carcinoma. J Clin Endocrinol Metab 2018;103:1686–95,
http://dx.doi.org/10.1210/jc.2017-02591.
[10] Gonzalez RJ, Tamm EP, Ng C, Phan AT, Vassilopoulou-Sellin R, Per-
rier ND, et al. Response to mitotane predicts outcome in patients
with recurrent adrenal cortical carcinoma. Surgery 2007;142:867–75,
http://dx.doi.org/10.1016/j.surg.2007.09.006 [discussion 867-875].
[11] Hermsen IGC, Gelderblom H, Kievit J, Romijn JA, Haak HR. Extremely long sur-
vival in six patients despite recurrent and metastatic adrenal carcinoma. Eur J
Endocrinol 2008;158:911–9, http://dx.doi.org/10.1530/EJE-07-0723.
[12] Abiven-Lepage G, Coste J, Tissier F, Groussin L, Billaud L, Dousset
B, et al. Adrenocortical carcinoma and pregnancy: clinical and bio-
logical features and prognosis. Eur J Endocrinol 2010;163:793–800,
http://dx.doi.org/10.1530/EJE-10-0412.
[13] de Simone V, Pagani O. Pregnancy after breast cancer:
hope after the storm. Minerva Ginecol 2017;69:597–607,
http://dx.doi.org/10.23736/S0026-4784.17.04113-2.
[14] de Corbière P, Ritzel K, Cazabat L, Ropers J, Schott M, Libé R, et al. Pregnancy in
women previously treated for an adrenocortical carcinoma. J Clin Endocrinol
Metab 2015;100:4604–11, http://dx.doi.org/10.1210/jc.2015-2341.
Laurence Salle ∗
Robin Mas
Marie-Pierre Teissier-Clément
Department of Endocrinology, Diabetology and
Nutrition, University Hospital of Limoges, 2, avenue
Martin-Luther-King, 87042 Limoges, France
∗ Corresponding
author.
E-mail address: laurence.teyssieres@orange.fr
(L. Salle)
https://doi.org/10.1016/j.ando.2020.07.1112
0003-4266/ © 2020 Elsevier Masson SAS. All rights reserved.
Recurrent hypothyroidism and thrombopenic throm-
botic purpura
Hypothyroïdie et purpura thrombotique thrombocytopénique
a r t i c l e i n f o
Keywords:
Hypothyroidism
Thrombotic microangiopathy
Mots clés :
Hypothyroïdie
Microangiopathie thrombotique
Thrombotic thrombocytopenic purpura (TTP) is a rare and
life-threatening thrombotic microangiopathy characterized by
microangiopathic hemolytic anemia, severe thrombocytopenia,
and organ ischemia linked to disseminated microvascular platelet
rich-thrombi [1]. TTP is caused by deficiency of a plasma metallo-
protease, ADAMTS13. When exposed to high shear stress in the
microcirculation, von Willebrand factor (VWF) and platelets are
prone to form aggregates. This propensity of VWF and platelet
to form microvascular thrombosis is mitigated by ADAMTS13,
which cleaves VWF before it is activated by shear stress to cause
platelet aggregation in the circulation. Deficiency of ADAMTS13,
due to autoimmune inhibitors in patients with acquired TTP and
mutations of the ADAMTS13 gene in hereditary cases, leads to
VWF – platelet aggregation and microvascular thrombosis. TTP
shows striking involvement of myocardial arteries and variable
degrees of vascular involvement in kidney, pancreas, brain, and
adrenal glands [2]. We have previously reported the case of a
patient in which reversible hypothyroidism was observed in the
course of TTP [3]. A second episode of hypothyroidism is reported in
the same patient during a relapse of thrombotic microangiopathy.
A 42-year-old man was admitted to the hospital in Novem-
ber 1997 because of headache, lethargy, fatigue and mild right
arm weakness. During the preceding 10 days, he had had five
transient neurologic ischemic attacks. Laboratory findings were as
following: thrombocytopenia (platelet count 27 giga/L); mechani-
cal hemolytic anemia (hemoglobin 9.7 g/dL, schizocytes 2%, lactate
dehydrogenase (LDH) 842 U/L), glycaemia 4,8 mmol/L, serum crea-
tinine 104 ␮mol/L. Head CT injected with iodine did not show any
acute or subacute infarction or hemorrhage. TTP was diagnosed and
after five plasma exchanges and 2 mg vincristine injection, remis-
sion was achieved and the patient was discharged from the hospital.
There was no evidence of an underlying disorder. Thyroid function
tests performed because of unusual asthenia at presentation before
plasma exchange and head CT disclosed mild hypothyroidism (TSH
17 mIU/mL, normal range < 5 mIU/mL). Thyroid gland was clinically
normal. Serum was negative for thyroid autoantibodies. Because
the clinical presentation of hypothyroidism was poor, we kept
watch without specific therapy. Thyroid function returned to nor-
mal value after 6 months.
The patient remained in remission for 20 years and then pre-
sented to our department with paresthesia in upper right limb
and dysarthria related to a limited ischemic stroke on magnetic
resonance imaging. Biological findings showed mild hemolytic ane-
mia (hemoglobin: 10.5 g/dL, thrombocytopenia (platelet count: 60
giga/L, and schistocytes (4%) at blood film examination. The LDH
level was 450 U/L (N 135–325 U/L), haptoglobin was < 0.1 g/L, cre-
atinine 84 ␮mom/L. On admission, TSH was slightly increased,
7.4 mIU/mL, (normal range 0.270–4.20), T4 was normal.
A diagnosis of relapsing autoimmune TTP was rapidly estab-
lished, based on the presence of microangiopathic hemolytic
 Letter to the Editor / Annales d’Endocrinologie 81 (2020) 516–519
anemia, thrombocytopenia, and very low activity (5%) of ADAMTS-
13 in combination with the presence of an ADAMTS-13
autoantibody 62 IU/mL (n < 15). Complete remission was observed
after 6 plasma exchange, corticotherapy 1 mg/kg and 3 courses
of rituximab. Thyroid function return to normal range within 3
months. Thyroid antibodies were absent, thyroid ultra sound exam-
ination was normal. When last seen 30 month later, in June 2020,
the patient remains in remission, with normal Adamts 13 activity
and absence of Adamts-13 antibody.
TTP, a potentially fatal clinical syndrome, is primarily caused by
autoantibodies against the von Willebrand factor (VWF)-cleaving
metalloprotease ADAMTS13. In physiologic conditions, ultralarge
VWF multimers released from endothelial cells are cleaved by
ADAMTS13 in smaller VWF multimers, less adhesive to platelets.
In TTP, because of the absence of functional ADAMTS13 (either
absent by congenital defect or inhibited by specific autoantibod-
ies), ultralarge VWF multimers are released into the blood and bind
spontaneously to platelets to form aggregates within the arterial
and capillary microvessels. The VWF–platelet aggregates are large
enough to form microthrombi inducing tissue ischemia, platelet
consumption, and microangiopathic hemolytic anemia (schisto-
cytes on blood smear [1].
In TTP, the wide-spread vascular lesion spares practically no
organ. The most commonly affected organs seen in postmortem
examination include the kidney, brain, heart, spleen, and lung
[2]. However, infarction of various endocrine may occur in rare
instances. Diabetes mellitus has been reported, presumably the
result of islet cell infarction [2]. Diabetes insipidus from neurohy-
pophyseal lesions is also reported [4].
The most frequent clinical conditions associated with TTP are
bacterial infections and autoimmune diseases (mainly systemic
lupus erythematous [SLE], but also the antiphospholipid syndrome,
Gougerot–Sjögren syndrome), pregnancy, drugs (mitomycin C,
cyclosporine, quinine, clopidogrel, ticlopidine), HIV infection, pan-
creatitis, cancers, and organ transplantation, with some of them
being likely involved in the triggering mechanism of the TTP
episode [1]. In contrast, ∼50% of TTP have an idiopathic presen-
tation. Only a few cases of TTP co-occurring with Graves disease or
hypothyroidism have been described [5–10] 5-10. However, in all
cases, thyroid diseases were autoimmune and not secondary to the
microangiopathy.
As thrombotic microangiopathy has been observed previously in
the thyroid [2], and as the hypothyroidism was reversible after TTP
remission in our patient, we postulate that microvascular lesions
were involved in the pathogenesis of hypothyroidism; especially
autoimmune thyroiditis or hypothyroidism related to antithyroid
agents were excluded in our patient. Hypothyroidism was observed
before iodine injection at the first hospitalization and in the absence
of iodine overload during the second hospitalization. Lastly, plasma
exchange could be useful in the treatment of amiodarone-induced
hyperthyroidis because it decrease thyroid hormone level but the
patient presented with hypothyroidism before plasma exchange
[11] 11. Moreover, we have checked thyroid hormone levels in
patients after plasma exchange without significant change in hor-
mone status.
As the patient presented 2 episodes of hypothyroidism asso-
ciated with TTP flare, we conclude that hypothyroidism may
complete the commonly agreed upon manifestations of TTP and
we recommend screening of thyroid function in TTP.
Informed consent and patient details
Written informed consent for publication of their clinical details
was obtained from the patient.
519
Funding
The authors received no financial support for the research,
authorship, and/or publication of this article.
Disclosure of interest
The authors declare that they have no competing interest.
References
[1] Joly BS, Coppo P, Veyradier A. An update on pathogenesis and diagnosis of
thrombotic thrombocytopenic purpura. Expert Rev Hematol 2019;12:383–95,
http://dx.doi.org/10.1080/17474086.2019.1611423.
[2] Gregory A, Hosler, Ana M, Cusumano, Grover M, Hutchins. Throm-
botic thrombocytopenic purpura and hemolytic uremic syndrome
are distinct pathologic entities. Arch Pathol Lab Med 2003;127:834–9,
http://dx.doi.org/10.1043/1543-2165(2003)127<834:TTPAHU>2.0.CO;2.
[3] Durand JM, Gautier C, Salas S, Retornaz F, Lefevre P. Thrombotic throm-
bocytopenic purpura and hypothyroidism. Am J Hematol 1999;61:83–4,
http://dx.doi.org/10.1002/(sici)1096-8652(199905)61:1<83::aid-ajh16>3.0.
co;2-p.
[4] Qureshi M, Pathak N, Pinsker RW, Gintautas J, Santucci T. Unique case of throm-
botic thrombocytopenic purpura and diabetes insipidus. Proc West Pharmacol
Soc 2005;48:145–7.
[5] Chaughtai S, Khan I, Gupta V, Chaughtai Z, Ong R, Asif A, et al. Graves disease-
induced thrombotic thrombocytopenic purpura: a case report. J Med Case Rep
2019;13:377, http://dx.doi.org/10.1186/s13256-019-2307-1.
[6] Chaar BT, Kudva GC, Olsen TJ, Silverberg AB, Grossman BJ. Thrombotic throm-
bocytopenic purpura and Graves disease. Am J Med Sci 2007;334:133–5,
http://dx.doi.org/10.1097/MAJ.0b013e31812e9735.PMID:17700205.
[7] Coser P, Fabris P, Prinoth O, Quaini R, Gentilini I. Thrombotic thrombocy-
topenic purpura in hypothyroidism: an accidental association? Haematologica
1982;67:625–9.
[8] Okumura T, Hashimoto K, Aomura D, et al. Thrombocytopenic pur-
pura treated with rituximab associated with primary Sjögren’s
syndrome and primary hypothyroidism. Intern Med 2020;59:715–9,
http://dx.doi.org/10.2169/internalmedicine.3722-19 [Published online 2019
Nov 8].
[9] Hou L, Du Y. Atypical hemolytic uremic syndrome precipitated
by thyrotoxicosis: a case report. BMC Pediatrics 2020;20:169,
http://dx.doi.org/10.1186/s12887-020-02082-0.PMID: 32303208.
[10] Chhabra 1 S, Tenorio G. Thrombotic Thrombocytopenic Pur-
pura Precipitated by Thyrotoxicosis. J Clin Apher 2012;27:265–6,
http://dx.doi.org/10.1002/jca.21210 [Epub 2012 May 30].
[11] Zhu L, Zainudin SB, Kaushik M, Khor LY, Chang CL. Plasma exchange in
the treatment of thyroid storm secondary to type II amiodarone-induced
thyrotoxicosis. Endocrinol Diabetes Metab Case Rep 2016;2016:160039,
http://dx.doi.org/10.1530/EDM-16-0039.
Antoine Briantais a
Jean Baptiste Dalmas a
Laure Swiader a
Pascale Poullin b
Jean-Marc Durand a,∗
a Internal Medicine department, Timone Hospital,
Aix-Marseille University, 264, rue Saint-Pierre, 13385
Marseille cedex 05, France
b Hemapheresis department, La Conception Hospital,
Aix-Marseille University, 136, rue Saint-Pierre, 13005
Marseille cedex 05, France
∗ Corresponding author.
E-mail address: Jean-marc.durand@ap-hm.fr
(J.-M. Durand)
https://doi.org/10.1016/j.ando.2020.07.1111
0003-4266/ © 2020 Elsevier Masson SAS. All rights reserved.