Académique Documents
Professionnel Documents
Culture Documents
Gynecologic Oncology
H I G H L I G H T S
• ESR1 and ER expression are similarly high in HGSOC and in breast cancer.
• ER expression did not differ between platin-sensitive or –resistant HGSOC.
• ER expression did not differ between matched primary or recurrent HGSOC.
• Letrozole maintenance showed improved progression-free survival.
a r t i c l e i n f o a b s t r a c t
Article history: Objectives. Endocrine therapy is used as maintenance in estrogen receptor (ER) positive breast cancers and
Received 8 August 2017 has been proposed in low-grade serous ovarian cancers (LGSOC). Here we examine a rationale for its use as
Received in revised form 24 October 2017 maintenance in high-grade serous ovarian cancers (HGSOC).
Accepted 26 October 2017
Methods. We accessed the TCGA PANCAN dataset to evaluate the expression of ESR1. ESR1 expression data on
Available online 20 November 2017
all cancers (n = 8901) and HGSOC (n = 527) were followed by investigation of ER expression via immunohis-
Keywords:
tochemistry (IHC) (n = 4071). The same was performed in an independent cohort for matched primary and
Ovarian cancer recurrent HGSOC (n = 80). Finally, newly diagnosed ER+ HGSOC patients were offered a maintenance therapy
HGSOC with Letrozole.
Letrozol maintenance Results. ESR1 was strongly expressed in similar levels in HGSOC as in breast cancer. We found a strong ER
Aromatase inhibitor expression via IHC in both the primary and matched recurrent HGSOC, particularly in the Platinum-resistant
ER subgroup. The additional use of Letrozole as maintenance treatment was associated with a significantly
prolonged recurrence free interval (after 24 months 60% when taking Letrozole versus 38.5% in the control
group; p = 0.035; RFS: IC50 reached by one subject versus 13.2 months). This effect was also present in patients
treated additionally with Bevacizumab; 20.8% of patients had no recurrence after 12 months compared to 87.5%
when taking Letrozole in addition to Bevacizumab (p = 0.026).
Conclusions. Primary HGSOC have a slightly higher ESR1 than and a similar ER expression breast cancer where
aromatase inhibitor maintenance is routine for decades. Here we demonstrate evidence for the usefulness of
Letrozole in HGSOC, particularly in patients with chemotherapy resistance or residual disease.
© 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
https://doi.org/10.1016/j.ygyno.2017.10.036
0090-8258/© 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
80 V. Heinzelmann-Schwarz et al. / Gynecologic Oncology 148 (2018) 79–85
Table 1
Clinicopathological variables of patients with HGSOC in the various cohort.
Cohort A B C D
Clinicopathological characteristics TCGA PANCAN ER IHC Caris Matched primary/recurrent Letrozole maintenance No maintenance
ER results from Caris commercial biomarker database were obtained the interval between the end of adjuvant chemotherapy and TTNT in
using a data extraction tool. In accordance with Western Institutional correlation with/without the use of Letrozole using Kaplan-Meier
Review Board guidelines, patient identification was not possible analysis. Patients with residual disease after debulking surgery received
throughout the study, so the study was considered exempt, and institu- Bevacizumab maintenance treatment 7.5 mg/m2 3-weekly for
tional review board approval was waived. 15 months as to the Swiss guidelines. Hereby patients could receive
In the second independent Swiss matched primary/recurrent both Bevacizumab and Letrozole in parallel. Patients in the non-
HGSOC cohort (Cohort C; Table 1), protein detection of ER on TMA slides Letrozole group did not receive Letrozole maintenance after adjuvant
were stained using an HRP Linked- Antibody Conjugated automated chemotherapy. Survival analysis was done by Kaplan-Meier-analysis.
Staining System (Bond Leica Biosystems, Muttenz, Switzerland). For an- Relapse free survival was defined as the length of time after finishing
tigen retrieval, paraffin embedded tissue slides were incubated accord- primary adjuvant chemotherapy until relapse defined by progression
ing to standard procedures. Slides were incubated with the anti-human on a CT scan.
ER-alpha (Clone 6F11, Leica Biosystems Muttenz, Switzerland, dilution
1:25) for 30 min. Negative controls involved the omission of the prima- 2.5. Data analysis
ry antibodies. Counterstaining was performed with hematoxylin and 1%
acid alcohol. ER staining is regularly checked in international round In order to compare independent categorical or ordinal variables
robin tests. To assess ER expression levels, each core sample was scored across study groups, Fisher's exact tests or Mann-Whitney U tests
for overall percentage (between 0 and 100) and overall intensity (0 to were performed as appropriate. Paired categorical data were explored
3) by two independent pathologists (S.S. and G.S.) and discrepancies re- using McNemar-tests. Time to event data was analyzed using Kaplan-
solved by consensus. The scoring system IRS (Immunoreactive Score) Meier method or Cox-regression with corresponding hazard ratios and
and ATS (Allred Total score) were used and in addition total percentage p-values.
of positive nuclei was recorded. Positive ER by IRS score was defined as ESR1 expression was dichotomized using tree based model
at least 10% positive nuclei and at least a moderate or strong staining partitioning (“ctree” in R package “party”). In addition, we have also
intensity. Positive ER by ATS score was defined as at least N1% positive dichotomized the data set based on quartiles in order to create four
nuclei and moderate or strong staining intensity and positive ER by groups for ESR1 expression. A p-value of ≤0.05 was considered statically
percentage was defined as staining N0% (Table 2). significant. Statistical analysis was performed with the statistical soft-
ware R version 3.1.1 (https://www.R-project.org/). Ethical approval
2.4. Single-site prospective Letrozole maintenance for the retrospective analysis of clinico-pathological data was granted
by the Northwestern Swiss Ethical Board.
Since 2013 all newly, diagnosed HGSOC FIGO III/IV patients with ER
positive cancers equal or above 1% were offered Letrozole 2.5 mg daily 3. Results
maintenance treatment in an off-label fashion following debulking sur-
gery and adjuvant Platinum-based chemotherapy (Cohort D; Table 1). 3.1. Elevated ESR1 expression correlates with better disease outcome in the
All patients were informed and consented about the off-label use of TCGA PANCAN datasets
Letrozole and about the potential use of their clinical information pro-
vided to the hospital in line with the National Human Ethics Regulations We reviewed the TCGA PANCAN database and found ESR1 expres-
in Switzerland. Discontinuation of Letrozole treatment was possible sion to be a predictor in all cancers with an overall disease-free survival
whenever the patient wished to stop the treatment, had significant (n = 8901, HR 0.935, CI [0.9222 to 0.9474], p-value based on Cox
side effects or showed a symptomatic recurrence demanding further regression b 2e − 16) and relapse-free survival (n = 5664, HR 0.909,
chemotherapy treatment. The Letrozol group was compared to the CI [0.8905 to 0.9287], p-value based on Cox regression b 2e − 16)
group of patients, who declined maintenance treatment and who had (Fig. 1A and B). The outcome was comparable using quartile dichoto-
only observation (non-letrozole group). mized data (Fig. 1A and B). Next, we studied the expression of ESR1 in
To be included in the Letrozole analysis group, patients had to have different TCGA cancer types and found that ESR1 expression was strik-
taken Letrozole for at least three subsequent months. Maintenance ther- ingly upregulated in breast, endometrial and ovarian cancer (Fig. 1C),
apy did not have to start directly after the end of adjuvant treatment and indicating, that endometrial and ovarian cancer might be targetable
could be initiated at any time during this initial relapse free interval, with similar drugs as currently performed in ER positive breast cancer
resulting in a wide range when maintenance was initiated. We assessed patients. In order to study the ESR1 expression in serous ovarian cancer
patients in more detail, we examined the TCGA (n = 527) and Tothill
(n = 274) transcriptomic datasets and found that neither increased
Table 2 histopathological grading nor FIGO staging impacted on ESR1 ex-
Association of ER immunoexpression in matched primary and recurrent HGSOC. Percent- pression in those samples, pointing towards a promising treatment
ages in tables are related to subjects with expression values in primary and relapsed tu-
using aromatase inhibitors both in LGSOC and HGSOC (Fig. 1D).
mors; p-values of independent data are based on Fisher's exact tests, p-values of
dependent data are based on McNemar tests (Cohort C; Table 1). Three different scoring
systems were used: IRS (Immunoreactive Score), ATS (Allred Total score) and ER expres- 3.2. Almost half of HGSOC express ER
sion N0%.
Primary tumors Recurrent tumors p-Value ER positivity was found in 49.3% of HGSOC FIGO III/IV US-American
patients as measured by IHC (n = 4071; Cohort B, Table 1). Interesting-
+ER expression (IRS)
- Platinum-sensible (%) 23.9 19.6 0.75
ly, different sampling sites (n = 14) revealed a similar ER expression
- Platinum-resistant (%) 36.8 21.1 0.45 pattern and therefore had no effect on the overall result (Fig. 1E). This
- p-Value 0.39 1 cohort, however, examined ER at any time of disease presentation,
+ER expression (ATS) either at diagnosis or at recurrence.
- Platinum-sensible (%) 32.6 37.0 0.80
- Platinum-resistant (%) 42.1 52.6 0.72
p-Value 0.43 0.19 3.3. ER expression remains unaffected in primary and recurrent HGSOC
+ER expression (percentage N 0) independent of drug resistance
- Platinum-sensible (%) 32.6 39.1 0.63
- Platinum-resistant (%) 42.1 52.6 0.72 To our knowledge, there is no dataset yet available examining ER
p-Value 0.43 0.29
expression in primary and matched recurrent HGSOC FIGO III/IV
82 V. Heinzelmann-Schwarz et al. / Gynecologic Oncology 148 (2018) 79–85
A B
Relapse-free survival
Disease-free survival
ESR1 (0, 4.61) ESR1 (4.61, 6.39) ESR1 (6.39, 8.31) ESR1 (6.31, 16.1) ESR1 (0, 4.61) ESR1 (4.61, 6.39) ESR1 (6.39, 8.31) ESR1 (6.31, 16.1)
C D E
ESR1 expression
ESR1 expression
IHC ER
percentage
ESR1 expression
ESR1 expression
IHC PR
percentage
Fig. 1. ESR1 expression in TCGA dataset. Overall disease-free survival (A) and relapse-free survival (B) in the TCGA PANCAN dataset (Cohort A, Table 1) p-value (log rank test). (C) Box-
Whisker blot ESR1 expression for anatomical origin sorted by descending median ESR1 expression among all TCGA cancer types; (D) ESR1 expression for histopathological grading and
FIGO staging in the Tothill and TCGA data set for high-grade serous ovarian cancer samples (Cohort A, Table 1). (E) Barchart showing percentage of IHC results for ER and PR in HGSOC
broken down into various sampling sites (Cohort B, Table 1).
(Cohort C, Table 1). ER expression was equally high in primary and III and 25% FIGO IV Stage in the non-Letrozole group, respectively.
recurrent HGSOC using three different methods of assessment, Both groups were also comparable in regards to residual disease
namely IRS and ATS score and ER expression N 0% (Table 2). For ER, (R0 in 58%).
the expression rate was equally high in all subgroups, independent During a median observation time of 14 months, the recurrence free
of platinum-sensible or resistant disease (Table 2). interval ranged from 4 to 123 months. The use of Letrozole as mainte-
nance after adjuvant chemotherapy independent of the time when
3.4. Letrozole maintenance improves progression free survival in HGSOC this treatment was started was associated with a significantly
patients prolonged recurrence free interval (RFS, Table 3). After 24 months,
60% of patients were recurrence free in the Letrozole versus 38.5% in
In the single-site prospective maintenance group 50 ER positive the control group (p = 0.035, Fig. 2). Whilst the control group reached
HGSOC FIGO III/IV patients were treated with standard adjuvant IC50 after 13.2 months, no recurrence free survival could be calculated
Platinum-based chemotherapy. After this treatment had ceased, for the Letrozole maintenance group at this point in time as only one
23 patients received Letrozole 2.5 mg daily off-label as a mainte- subject reached the endpoint so far. The improvement of the RFS
nance therapy whilst 27 patients did not receive any such mainte- could also been seen in the high risk group of patients with residual
nance treatment (Cohort D, Table 1). The Letrozole group disease treated with Bevacizumab maintenance; 20.8% of patients had
consisted of 60% FIGO III and 40% FIGO IV Stage versus 75% FIGO no recurrence after 12 months against 87.5% when taking Letrozole in
Table 3
Median recurrence free survival in months for ER + patients in various subgroups. Median in the Letrozole group not reached for no Bevacizumab/no residual disease and +/−
Bevacizumab/residual disease as yet with only one event so far (*). Estimate 64 months.
Fig. 2. Relapse-free survival of HGSOC FIGO III/IV treated with/without Letrozole as a maintenance therapy. Letrozole treated (red line; n = 23) vs untreated group (black line; n = 27),
showing a significantly longer relapse-free survival when using Letrozole (p = 0.035). RFS is defined from the end of adjuvant chemotherapy till first symptomatic recurrence, confirmed
by RECIST. IC50 not reached in the Letrozole vs untreated group (13.2 months) (Cohort D; Table 1).
addition to Bevacizumab (p = 0.026, Fig. 3). Also in this situation, the 4. Discussion
start of Letrozole was not linked to the start of Bevacizumab and could
have been at a later point in time. Precaution and assessment of side 4.1. Endocrine therapy in ovarian cancer
effects during anti-hormonal therapy with Letrozole, especially in con-
text to the possible bone loss, was performed according to the standards At present, the role of endocrine therapy with aromatase inhibitors
used in breast cancer patients, namely by performing regular bone or Tamoxifen is only established for ovarian cancers in the recurrent set-
scans and supplementing Vitamin D, Calcium or Denosumab where ting [24]. Randomized controlled trials are scarce and these drugs are
necessary. During the therapy with Letrozole, no major side effects therefore mainly used in relapsed and heavily pretreated patients to
were observed. Only two patients (6.4%) did interrupt the treatment prolong TTNT [25,26]. However, even in the recurrent situation, results
due to minor side effects (hot flushes, fatigue, bone pain) after 5, respec- of these trials and retrospective analysis were rather disappointing due
tively 9 months. to the lack of reported ER and PR expression as predictive markers.
Fig. 3. Relapse-free survival of HGSOC FIGO III/IV treated with Bevacizumab as a maintenance therapy. Letrozole treated (red line; n = 9) vs untreated group (black line; n = 6), showing a
significantly longer relapse-free survival when using Letrozole in combination with Bevacizumab in patients with residual disease (p = 0.026). RFS is time from end of adjuvant
chemotherapy until first symptomatic relapse as confirmed by RECIST.
84 V. Heinzelmann-Schwarz et al. / Gynecologic Oncology 148 (2018) 79–85
[18] W. Sieh, M. Köbel, T. a Longacre, D.D. Bowtell, A. Defazio, M.T. Goodman, E. Høgdall, [27] E. Senkus, S. Kyriakides, F. Penault-Llorca, P. Poortmans, A. Thompson, S. Zackrisson,
S. Deen, N. Wentzensen, K.B. Moysich, J.D. Brenton, B. a Clarke, U. Menon, C.B. Gilks, F. Cardoso, ESMO Guidelines Working Group, Primary breast cancer: ESMO Clinical
A. Kim, J. Madore, S. Fereday, J. George, L. Galletta, G. Lurie, L.R. Wilkens, M.E. Carney, Practice Guidelines for diagnosis, treatment and follow-up, Ann. Oncol. (24 Suppl. 6)
P.J. Thompson, R.K. Matsuno, S.K. Kjær, A. Jensen, C. Høgdall, K.R. Kalli, B.L. Fridley, (2013) vi7–23, https://doi.org/10.1093/annonc/mdt284.
G.L. Keeney, R. a Vierkant, J.M. Cunningham, L. a Brinton, H.P. Yang, M.E. Sherman, [28] N.G. Version, N. Clinical, P. Guidelines, N. Guidelines, Breast Cancer2012.
M. García-Closas, J. Lissowska, K. Odunsi, C. Morrison, S. Lele, W. Bshara, L. [29] C. Falandry, B. Weber, A.-M. Savoye, F. Tinquaut, O. Tredan, E. Sevin, L. Stefani, F.
Sucheston, M. Jimenez-Linan, K. Driver, J. Alsop, M. Mack, V. McGuire, J.H. Savinelli, M. Atlassi, J. Salvat, E. Pujade-Lauraine, G. Freyer, Development of a geriat-
Rothstein, B.P. Rosen, M.Q. Bernardini, H. Mackay, A. Oza, E.L. Wozniak, E. Benjamin, ric vulnerability score in elderly patients with advanced ovarian cancer treated with
A. Gentry-Maharaj, S. a Gayther, A.V. Tinker, L.M. Prentice, C. Chow, M.S. Anglesio, first-line carboplatin: a GINECO prospective trial, Ann. Oncol. 24 (2013) 2808–2813,
S.E. Johnatty, G. Chenevix-Trench, A.S. Whittemore, P.D. Pharoah, E.L. Goode, D.G. https://doi.org/10.1093/annonc/mdt360.
Huntsman, S.J. Ramus, Hormone-receptor expression and ovarian cancer survival: [30] C. Davies, H. Pan, J. Godwin, R. Gray, R. Arriagada, V. Raina, M. Abraham, V.H.
an ovarian tumor tissue analysis consortium study, Lancet Oncol. 14 (2013) Medeiros Alencar, A. Badran, X. Bonfill, J. Bradbury, M. Clarke, R. Collins, S.R. Davis,
853–862, https://doi.org/10.1016/S1470-2045(13)70253-5. A. Delmestri, J.F. Forbes, P. Haddad, M.-F. Hou, M. Inbar, H. Khaled, J. Kielanowska,
[19] S. Stadlmann, U. Gueth, U. Reiser, P.-A. Diener, A.G. Zeimet, E. Wight, M. Mirlacher, G. W.-H. Kwan, B.S. Mathew, I. Mittra, B. Müller, A. Nicolucci, O. Peralta, F. Pernas, L.
Sauter, M.J. Mihatsch, G. Singer, Epithelial growth factor receptor status in primary Petruzelka, T. Pienkowski, R. Radhika, B. Rajan, M.T. Rubach, S. Tort, G. Urrútia, M.
and recurrent ovarian cancer, Mod. Pathol. 19 (2006) 607–610, https://doi.org/10. Valentini, Y. Wang, R. Peto, Adjuvant Tamoxifen: Longer Against Shorter (ATLAS)
1038/modpathol.3800575. Collaborative Group, Long-term effects of continuing adjuvant tamoxifen to
[20] M. Goldman, B. Craft, T. Swatloski, M. Cline, O. Morozova, M. Diekhans, D. Haussler, J. 10 years versus stopping at 5 years after diagnosis of oestrogen receptor-positive
Zhu, The UCSC cancer genomics browser: update 2015, Nucleic Acids Res. 43 (2015) breast cancer: ATLAS, a randomised trial, Lancet 381 (2013) 805–816, https://doi.
D812–7, https://doi.org/10.1093/nar/gku1073. org/10.1016/S0140-6736(12)61963-1.
[21] M.S. Cline, B. Craft, T. Swatloski, M. Goldman, S. Ma, D. Haussler, J. Zhu, Exploring [31] D. Zhao, F. Zhang, W. Zhang, J. He, Y. Zhao, J. Sun, Prognostic role of hormone recep-
TCGA pan-cancer data at the UCSC cancer genomics browser, Sci Rep 3 (2013), tors in ovarian cancer: a systematic review and meta-analysis, Int. J. Gynecol. Cancer
2652. https://doi.org/10.1038/srep02652. 23 (2013) 25–33, https://doi.org/10.1097/IGC.0b013e3182788466.
[22] caArray_EXP-485: TCGA Analysis of Gene Expression for Ovarian Serous [32] B.J. Monk, H. Dalton, J.H. Farley, D.M. Chase, I. Benjamin, Antiangiogenic agents as a
Cystadenocarcinoma Using Affymetrix HT_HG-U133A, (n.d.). maintenance strategy for advanced epithelial ovarian cancer, Crit. Rev. Oncol.
[23] R.W. Tothill, A.V. Tinker, J. George, R. Brown, S.B. Fox, S. Lade, D.S. Johnson, M.K. Hematol. 86 (2013) 161–175, https://doi.org/10.1016/j.critrevonc.2012.09.012.
Trivett, D. Etemadmoghadam, B. Locandro, N. Traficante, S. Fereday, J.A. Hung, Y.- [33] P.S.G. Linda, R. Mileshkin, Richard J. Edmondson, Rachel O'Connell, Katrin Marie
E. Chiew, I. Haviv, Australian Ovarian Cancer Study Group, D. Gertig, A. DeFazio, Sjoquist, David Cannan, Rema Jyothirmayi, Philip James Beale, Tony Bonaventura,
D.D.L. Bowtell, Novel molecular subtypes of serous and endometrioid ovarian cancer Jeffrey C. Goh, Marcia Hall, Andrew R. Clamp, John A. Green, Rosemary Lord,
linked to clinical outcome, Clin. Cancer Res. 14 (2008) 5198–5208, https://doi.org/ James Scurry, Michael Fried, Phase II study of anastrozole in recurrent estrogen
10.1158/1078-0432.CCR-08-0196. (ER)/progesterone (PR) positive endometrial cancer: the PARAGON trial—ANZGOG
[24] NCCN Guidelines, Ovarian Cancer Including Fallopian Tube Cancer and Primary Peri- 0903, J. Clin. Oncol. 34 (2016) (Suppl; Abstr 5520) (n.d.).
toneal Cancer Version 1.2016, NCCN.org 2016. [34] B. Thurlimann, A. Keshaviah, A.S. Coates, H. Mouridsen, L. Mauriac, J.F. Forbes, R.
[25] S.A. Weiner, D.S. Alberts, E.A. Surwit, J. Davis, D. Grosso, Tamoxifen therapy in recur- Paridaens, M. Castiglione-Gertsch, R.D. Gelber, M. Rabaglio, I. Smith, A. Wardley,
rent epithelial ovarian carcinoma, Gynecol. Oncol. 27 (1987) 208–213http://www. K.N. Price, A. Goldhirsch, A comparison of letrozole and tamoxifen in postmeno-
ncbi.nlm.nih.gov/pubmed/3570058. pausal women with early breast cancer, N. Engl. J. Med. 353 (2005) 2747–2757,
[26] C.A. Papadimitriou, S. Markaki, J. Siapkaras, G. Vlachos, E. Efstathiou, I. Grimani, G. https://doi.org/10.1056/NEJMoa052258.
Hamilos, M. Zorzou, M.-A. Dimopoulos, Hormonal therapy with letrozole for re-
lapsed epithelial ovarian cancer. Long-term results of a phase II study, Oncology
66 (2004) 112–117, https://doi.org/10.1159/000077436.