Vous êtes sur la page 1sur 8

Int. J. Oral Maxillofac. Surg.

2015; 44: 143–150


http://dx.doi.org/10.1016/j.ijom.2014.10.004, available online at http://www.sciencedirect.com

Clinical Paper
Head and Neck Oncology

A simple novel prognostic A. Almangusha,b, R. D. Colettac, I. O.


Belloa,d, C. Bitue, H. Keski-Sänttif, L. K.
Mäkinenf, J. H. Kauppilag, M. Pukkilah, J.
Hagströma,i, J. Larannej, S. Tommolak, Y.

model for early stage oral Soinil, V.-M. Kosmal, P. Koivunenm, L. P.


Kowalskin, P. Niemineno, R. Grénmanp, I.
Leivoa,s, T. Saloe,q,r

tongue cancer
a
Department of Pathology, Haartman Institute,
University of Helsinki, Helsinki, Finland;
b
Institute of Dentistry, University of Misurata,
Misurata, Libya; cDepartment of Oral
Diagnosis, School of Dentistry, State University
of Campinas, Piracicaba, São Paulo, Brazil;
A. Almangush, R.D. Coletta, I.O. Bello, C. Bitu, H. Keski-Säntti, L.K. Mäkinen, J.H. d
Department of Oral Medicine and Diagnostic
Kauppila, M. Pukkila, J. Hagström, J. Laranne, S. Tommola, Y. Soini, V.-M. Kosma, P. Sciences, King Saud University, Riyadh, Saudi
Koivunen, L.P. Kowalski, P. Nieminen, R. Grénman, I. Leivo, T. Salo: A simple novel Arabia; eDepartment of Diagnostics and Oral
Medicine, Institute of Dentistry, University of
prognostic model for early stage oral tongue cancer. Int. J. Oral Maxillofac. Surg. Oulu, Oulu, Finland; fDepartment of
2015; 44: 143–150. # 2014 International Association of Oral and Maxillofacial Otorhinolaryngology and Head and Neck
Surgeons. Published by Elsevier Ltd. All rights reserved. Surgery, Helsinki University Central Hospital,
Helsinki, Finland; gDepartments of Surgery and
Pathology, University of Oulu and Oulu
University Hospital, Oulu, Finland; hDepartment
of Otorhinolaryngology and Head and Neck
Surgery, Kuopio University Hospital, Kuopio,
Finland; iDepartment of Oral Pathology,
Institute of Dentistry, University of Helsinki,
Helsinki, Finland; jDepartment of
Otorhinolaryngology and Head and Neck
Surgery, Tampere University Hospital,
Tampere, Finland; kDepartment of Pathology,
Tampere University Hospital, Tampere, Finland;
l
Department of Pathology and Forensic
Medicine, University of Eastern Finland,
Kuopio, Finland; mDepartment of
Otorhinolaryngology and Head and Neck
Surgery, Oulu University Central Hospital, Oulu,
Finland; nDepartment of Head and Neck
Surgery and Otorhinolaryngology, A.C.
Camargo Cancer Centre, São Paulo SP, Brazil;
o
Department of Medical Informatics and
Abstract. The prognostication of patient outcome is one of the greatest challenges in Statistics, University of Oulu, Oulu, Finland;
p
Department of Otorhinolaryngology – Head
the management of early stage oral tongue squamous cell carcinoma (OTSCC). This and Neck Surgery, Turku University Hospital,
study introduces a simple histopathological model for the prognostication of University of Turku, Turku, Finland;
q
survival in patients with early OTSCC. A total of 311 cases (from Finland and Department of Diagnostics and Oral Medicine,
Brazil) with clinically evaluated early stage OTSCC (cT1–T2cN0cM0) were Oulu University Hospital and Medical Research
Centre, Oulu, Finland; rInstitute of Dentistry,
included in this multicentre retrospective study. Tumour budding (B) and depth of University of Helsinki, Helsinki, Finland;
invasion (D) were scored on haematoxylin–eosin-stained cancer slides. The cut-off s
Department of Pathology, University of Turku,
point for tumour budding was set at 5 buds (low <5; high 5) and for depth of Turku, Finland
invasion at 4 mm (low <4 mm; high 4 mm). The scores of B and D were
Key words: oral tongue squamous cell carci-
combined into one model: the BD predictive model. On multivariate analysis, a high
noma; prognosis; BD model; tumour budding;
risk score (BD score 2) correlated significantly with loco-regional recurrence depth of invasion.
(P = 0.033) and death due to OTSCC (P < 0.001) in early stage OTSCC. The new
BD model is a promising prognostic tool to identify those patients with aggressive Accepted for publication 6 October 2014
cases of early stage OTSCC who might benefit from multimodality treatment. Available online 11 November 2014

0901-5027/020143 + 08 # 2014 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
144 Almangush et al.

Oral tongue squamous cell carcinoma consequently be applied only to patients indicator representing the malignant poten-
(OTSCC) is the most aggressive type of at high risk. tial of early OSCC.14 Furthermore, it has
oral SCC and its management may com- Despite the role of epithelial mesenchy- been reported by many recent and well-
prise multimodality treatment even in the mal transition (EMT) in cancer cell mi- documented studies as an important prog-
early stages. Many histomorphological gration and metastasis,7 none of the nostic parameter in early OTSCC.15,16
grading models have been introduced previous prognostic models has included With this background, the aim of the
for the prognostication of OSCC.1–3 The histopathological parameters to consider present study was to introduce a practical
main role of histopathological prognosti- EMT. Thus, it would appear advantageous and simple model by combining the scores
cation is to complement the shortcomings for a prognostic model to include tumour for tumour budding and depth of invasion
of the TNM staging system for OSCC and budding, which correlates with EMT in (BD scoring from 0 to 2) to predict the
thus to provide the clinician with the tools many carcinomas,8,9 including tongue behaviour of early stage OTSCC. Using
to develop a precise treatment plan. In SCC.10 In addition, tumour budding is this model we aim to be able to select those
addition to the complexity of previous known to reflect the biological activity OTSCC patients who could be at high risk
models, a number of studies have reported of cancer11 and has been recognized in- of recurrence and/or distant dissemination
their inability to predict the outcome of creasingly as a significant, independent, of the disease and subsequently dying of the
patients with early stage OSCC4 or early and reproducible prognostic feature.12 disease.
OTSCC.5 This indicates the critical need The depth of tumour invasion or tumour
for a comprehensive predictive model in thickness (these two terms have been used
Patients and methods
order to make the appropriate choices in synonymously by some authors13,14), as
multimodality treatment of early OTSCC. measured from the surface of the tumour A total of 311 cases diagnosed with
Such choices may entail adjuvant therapy to the deepest point of invasion, has been OTSCC between 1979 and 2010 and clin-
with severe toxic effects,6 and should established as a valuable histopathological ically staged as cT1cN0cM0 or
cT2cN0cM0 were included in this study.
Table 1. Clinicopathological characteristics of the 311 study patients. Of these cases, 224 were from the Finnish
Characteristic Number of patients % university central hospitals (Helsinki,
Oulu, Turku, Tampere, Kuopio) and had
Age, years
Median: 63 comprised part of a previously studied
63 157 50.5 cohort of 233 Finnish patients.5 The nine
>63 154 49.5 patients whose data on recurrence were
Gender incomplete were excluded from this latter
Male 165 53.1 group. The remaining 87 patients were
Female 146 46.9 diagnosed and treated at A.C. Camargo
Clinical stage Hospital, São Paulo, Brazil.
T1N0M0 124 39.9 Clinical data and follow-up records
T2N0M0 187 60.1
including cause of death and dates of
Grade (differentiation)
I (well differentiated) 105 33.8
diagnosis, recurrence, and/or death were
II (moderately differentiated) 131 42.1 retrieved. Surgical resection of the prima-
III (poorly differentiated) 75 24.1 ry tumour was carried out for all patients.
Worst pattern of invasion With regard to clinical staging, 124 cases
Superficial 78 25.1 (39.9%) were classified as cT1cN0cM0
Deep 233 74.9 and 187 cases (60.1%) were cT2cN0cM0.
Lymphocytic host response One hundred and sixty five (53.1%) were
Superficial 169 54.3 males and 146 (46.9) were females. Their
Deep 142 45.7 median age at diagnosis was 63 years
Perineural invasion
(range 10–95 years) (Table 1).
Absent 269 86.5
Present 42 13.5 Tumour budding is defined as the pres-
Recurrence ence of a single cancer cell or small cluster
No 222 71.4 of <5 cancer cells at the invasive front.5
Yes 89 28.6 During the scoring of tumour budding, the
Status whole tumour area was scanned at low
Alive 153 49.2 magnification (4), then the highest num-
Died of OTSCC 63 20.3 ber of tumour budding was counted at a
Died of other causes 95 30.5 higher magnification (20) and used as
Tumour budding
the score for budding. Depth of invasion
Superficial (<5 buds) 215 69.1
Deep (5 buds) 96 30.9 (or tumour thickness) was measured from
Depth of invasion the tumour surface to the deepest point of
Superficial (<4 mm) 116 37.3 invasion. The cut-off point for tumour
Deep (4 mm) 195 62.7 budding was set at 5 buds (low <5; high
BD score 5), and the cut-off point for depth of
Low risk (score 0) 103 33.1 invasion was set at 4 mm (low <4 mm;
Intermediate risk (score 1) 125 40.2 high 4 mm). In addition, other previous-
High risk (score 2) 83 26.7 ly examined parameters,5 including worst
OTSCC, oral tongue squamous cell carcinoma; BD, tumour budding (B) and depth of invasion pattern of invasion, lymphocytic host re-
(D). sponse, and perineural invasion, were also
Prognostic model for early stage OTSCC 145

Table 2. Description of BD scores.


BD score Histological description
Score 0 Tumour with <4 mm depth of invasion and <5 buds at the invasive front
Score 1 This tumour should have only one of the following features:
Tumour with 4 mm depth of invasion and <5 buds at the invasive front, OR
Superficial tumour (<4 mm), but with high activity of tumour budding at the invasive front (5 buds)
Score 2 Tumour with 4 mm depth of invasion and with high activity of tumour budding at the invasive front (5 buds)
BD, tumour budding (B) and depth of invasion (D).

Table 3. Association of disease-free survival and disease-specific survival with the clinical and histopathological parameters (unadjusted
univariate analysis).
Disease-free survival (DFS) Disease-specific survival (DSS)
a
Variable HR (95% CI) P-value HR (95% CI) P-valuea
Age, years 0.003 0.014
63 1 1
>63 1.89 (1.23–2.89) 1.87 (1.13–3.11)
Gender 0.715 0.441
Male 1 1
Female 1.08 (0.71–1.64) 1.22 (0.74–1.99)
TNM stage 0.527 0.141
T1N0M0 1 1
T2N0M0 0.87 (0.57–1.33) 1.48 (0.87–2.54)
Grade (differentiation) 0.739 0.199
I (well differentiated) 1 1
II (moderately differentiated) 1.11 (0.68–1.82) 1.68 (0.92–3.07)
III (poorly differentiated) 1.25 (0.72–2.16) 1.58 (0.79–3.16)
Worst pattern of invasion 0.045 0.001
Superficial 1 1
Deep 1.71 (0.98–2.99) 3.29 (1.42–7.62)
Lymphocytic host response 0.512 0.889
Superficial 1 1
Deep 0.89 (0.57–1.33) 0.97 (0.59–1.59)
Perineural invasion 0.199 0.502
Absent 1 1
Present 1.46 (0.84–2.55) 1.27 (0.65–2.49)
Tumour budding 0.005 <0.001
Superficial 1 1
Deep 1.85 (1.21–2.82) 2.59 (1.58–4.26)
Depth of invasion 0.109 <0.001
Superficial 1 1
Deep 1.44 (0.91–2.28) 3.15 (1.60–6.18)
BD score 0.025 <0.001
Low risk 1 1
Intermediate risk 1.50 (0.87–2.58) 3.79 (1.57–9.15)
High risk 2.14 (1.22–3.74) 6.24 (2.59–15.04)
HR, hazard ratio; CI, confidence interval; BD, tumour budding (B) and depth of invasion (D).
a
The likelihood ratio statistic was used to test the significance of variables.

analyzed. The scoring of all parameters Score 2: both parameters (budding and estimated time to recurrence and time to
was carried out on haematoxylin–eosin- depth of invasion) are higher than the cut- disease-specific death. The statistical sig-
stained sections by an independent observ- off (Fig. 1G, H). nificance of the differences in Kaplan–
er (A.A.) and reviewed by an experienced In accordance with the BD model, each Meier curves between the BD groups
head and neck pathologist (I.L.). Both case was assigned a score between 0 and 2. was determined by log rank test. Univari-
observers were blinded to the patient data Cases with a score of 0 were classified as ate analysis was done using the Cox pro-
and clinical outcome. low risk, those with a score of 1 as inter- portional hazard regression model to
Tumour budding ‘B’ and depth of inva- mediate risk, and those with a score of 2 as evaluate the relationship between the
sion ‘D’ were combined in one model, the high risk. BD model scores and recurrence or death
BD model, as follows (Table 2): from OTSCC. Multivariate analysis (using
Score 0: neither budding nor depth of Cox regression) was performed to evalu-
Statistical analysis
invasion is higher than the cut-off ate the independent prognostic strength of
(Fig. 1A, B). The x2 test was used to assess possible BD scores when adjusted for other clini-
Score 1: only one of the parameters relationships between clinicopathological copathological covariates (age, gender,
(budding or depth of invasion) is higher variables and the BD score. The Kaplan– tumour stage, and tumour grade) and
than the cut-off (Fig. 1C–F). Meier method was used to obtain the worst pattern of invasion. The likelihood
146 Almangush et al.

ratio statistic was used to test the signifi-


cance of variables in the estimated models.
Statistical significance was set at
P < 0.05. Harrell’s C-statistic was used
to evaluate the predictive power of surviv-
al models.17,18 Since the BD score is a
linear derivative of both tumour budding
(B) and depth of invasion (D), it is not
possible to include all three variables in a
single model. Thus, two separate models
(BD included vs. B and D included) were
estimated for both disease-free survival
(DFS) and disease-specific survival
(DSS) to show their relationship with B,
D, and BD variables.
All statistical analyses were done using
IBM SPSS Statistics for Windows, version
20.0 (IBM Corp., Armonk, NY, USA) and
Stata 13 (StataCorp LP, College Station,
TX, USA).

Results
Clinicopathological characteristics of the
patients are summarized in Table 1. The
median follow-up time was 57 months
(range 1–278 months). Sixty-three
patients (20.3%) died of OTSCC, 95
patients (30.5%) died of causes other than
OTSCC, and 153 patients (49.2%) were
alive at the end of follow-up or at the last
visit on record.
Loco-regional recurrence (LRR) was
reported in 89 patients (28.6%), and 48
(53.9%) of them died from OTSCC. The
presence of LRR was significantly associ-
ated with poor DSS (P < 0.001).
Of all histopathological parameters
evaluated in this study, only the worst
pattern of invasion and tumour budding
were associated with both DFS and DSS
in the unadjusted univariate analysis
(Table 3). Depth of invasion was related
to DSS, but not to DFS. Fig. 1. Scores of the BD model. (A) Score 0: superficial tumour with only large islands at the IF
(20); (B) high magnification of the region inside the rectangle in A showing the absence of
tumour budding at the IF (100). (C) Score 1: presence of tumour budding at the IF of a
BD model and relationship with
superficial tumour (20); (D) high magnification of the region inside the rectangle in C
clinicopathological parameters demonstrating the presence of more than 5 buds at the IF (200). (E) Score 1: deep tumour
There were 103 cases (33.1%) with a low without any small clusters at the IF (4); (F) magnification of the rectangle region in E (10).
risk score (score 0), 125 (40.2%) with an (G) Score 2: presence of tumour budding at the IF of a deep tumour (40); (H) high
magnification of the rectangle in G showing tumour budding at the IF (200). A–F: haematox-
intermediate risk score (score 1), and 83
ylin–eosin-stained sections; G and H: pan-cytokeratin-stained sections. IF = invasive front.
(26.7%) with a high risk score (score 2).
There was no statically significant associ-
ation between BD score and age, gender, death from OTSCC (log rank P < 0.001) Multivariate analysis
or histological differentiation (tumour as the patients were previously classified
grade). into risk groups (Fig. 2A). On the other In the multivariate analysis, the BD score
hand, BD scores did not show any signifi- showed a predictive value in the high
cant association with death from other score group (score 2) compared to the
Univariate analysis
causes (log rank P = 0.808; Fig. 2B). low score group, with a hazard ratio
According to Kaplan–Meier analyses, There was a similar curve for BD scores (HR) of 2.19 (95% confidence interval
there was a stepwise gradation of BD in relation to LRR (log rank P = 0.023; (CI) 1.20–4.00) for recurrence, and of
scores ranging from favourable (score 0) Fig. 2C). In addition, worst pattern of 5.11 (95% CI 2.05–12.75) for cancer death
to intermediate (score 1) and unfavourable invasion was associated with patient sur- (Table 4). Based on this stepwise model,
prognosis (score 2) in relation to the risk of vival (Table 3). there was no statistically significant
Prognostic model for early stage OTSCC 147

association between worst pattern of inva-


sion and patient survival, although a high
worst pattern of invasion showed a ten-
dency towards an association with cancer
death in both models for DSS (Table 4).

Discussion
Early stage OTSCC (cT1–T2cN0cM0)
shows a higher mortality rate than early
stage tumours at other subsites of the oral
cavity (39% vs. 15.5%)19 indicating that,
for unknown reasons, OTSCC is highly
aggressive in nature. The optimal treat-
ment for patients with early OTSCC
remains a controversial issue,20 but in
many instances such cases are treated by
local surgical excision only. However,
early OTSCC is reported to have a risk
of loco-regional recurrence in 23%21 and/
or occult metastasis in 30% of cases.22
Thus, multimodality therapies should be
administered to those patients who are at
higher risk of metastasis or recurrence.
Unfortunately, these therapies have a po-
tential to cause complications, so it is of
great importance to avoid unnecessary
therapy. In general, there are no clear
indications as to when to perform an elec-
tive neck dissection (END) in early stage
head and neck cancer.23 This means that
patients may be subjected to the morbidity
associated with extensive neck dissections
even though they have no evidence of
loco-regional metastasis.13 Although sen-
tinel node navigation surgery (SNNS) has
been recommended for early stage
OTSCC in a recent study,24 further sup-
portive evidence is still required before
this can be advocated as a routine prog-
nostic procedure. In addition, the com-
plexity of lymphatic drainage in the
head and neck region usually make the
detection of all sentinel lymph nodes in
carcinomas of this area a difficult and
time-consuming procedure.25 Moreover,
in some instances, exposure, dissection,
and the time required to complete the
sentinel lymph node biopsy may approxi-
mate that which is sufficient to perform a
selective neck dissection.26
The use of single or combined histo-
pathological parameters for the prognosti-
cation of OSCC has been introduced in a
number of previous studies.1,2,14 In addi-
tion, traditional histopathological evalua-
tion of surgically removed tissues is still
Fig. 2. Low risk, intermediate risk, and high risk cases. (A) Deaths due to OTSCC. (B) Other the cornerstone of pathological grading in
causes of death. (C) Loco-regional recurrence. routine clinical practice.27 However, com-
bining histopathological parameters to
identify high risk patients could be highly
successful in reducing therapeutic compli-
cations. Such a model for prognostication
might also enhance survival by allowing
148 Almangush et al.

P-valueb
clinicians to determine the most appropri-

0.044

0.019

0.119
ate treatment for each risk group.28
In this study, we introduce a promising
model for the prognostication and treat-
Model 2a ment planning of patients with early

(1.01–3.06)

(1.09–4.65)

(0.80–4.80)
95% CI
OTSCC. In the multivariate analysis, a

0.7109
strong relationship was found with loco-
regional recurrence and with DSS (likeli-
hood ratio P = 0.033 and P < 0.001 re-
spectively). The significant association of
Death from OTSCC

the BD score with patient outcome may


1.76

2.26

1.96
HR

allow its use in treatment planning, which


1

1
could include the consideration of END
even in cases of early OTSCC when a high
P-valueb

<0.001

BD score is found. Unlike the previously

0.102
introduced histopathological prognostic
models, the BD model shows a strong
correlation with patient survival in early
a

stage OTSCC. However, the advanced


(2.05–12.75)
Model 1

(1.54–9.22)

(0.82–4.84)
95% CI

stages of OTSCC are serious diseases,


which require wide surgical resection,
0.7163

neck dissection, and adjuvant therapy


(mostly radiation therapy).
The current study evaluated six histo-
pathological parameters including tumour
3.77
5.11

2.00
HR

grade (differentiation), worst pattern of


1

invasion, lymphocytic host response, peri-


neural invasion, tumour budding, and
P-valueb

depth of invasion. In the multivariate sta-


0.020

0.486

0.357

tistical analysis, only tumour budding and


depth of invasion were strongly associated
with patient survival. However, our pro-
a

posed histopathological model combined


Model 2

HR, hazard ratio; CI, confidence interval; BD, tumour budding (B) and depth of invasion (D).
(1.10–2.93)

(0.72–1.98)

(0.72–2.43)
Table 4. Multivariate analysis for loco-regional recurrence and for death from early OTSCC.

95% CI

tumour budding to represent the invasion


0.6523

pattern and depth of invasion to represent


the depth of tumour penetration. Impor-
tantly, tumour budding is suggested to be a
sign of epithelial to mesenchymal transi-
The likelihood ratio statistic was used to test the significance of variables.
Adjusted for age, gender, grade (tumour differentiation), and TNM stage.

tion of carcinoma cells, reflecting the


1.80

1.20

1.32
HR

powerful invasive growth of tongue,10


1

colorectal,8 and breast9 cancers.


Fresh-frozen diagnosis of a surgical
P-valueb

specimen during surgery has been sug-


0.033

0.276

gested as a method for intraoperative path-


ological classification of OSCC29 that will
Loco-regional recurrence

influence the treatment protocol. During


a

surgical removal of an early OTSCC, we


Model 1

(0.89–2.72)
(1.20–4.00)

(0.76–2.51)
95% CI

recommend performing immunohisto-


0.6541

chemical cytokeratin staining on frozen


sections of the tumour specimen. Then,
BD scoring can be undertaken and cases of
high risk (score 2) might benefit from
1.59
2.19

1.38
HR

END. A definitive operation can then be


1

carried out without interrupting the anaes-


thesia.
In addition to the significant prognostic
Worst pattern of invasion

value of the BD model, a combination of


these two parameters will increase the
Depth of invasion

accuracy and reliability of prognostication


Tumour budding

Intermediate

compared to the use of each parameter


Superficial

Superficial

Superficial

Superficial

separately. The main clinical significance


BD scores

C-statistic

of the BD model is that it allows the


Deep

Deep

Deep

Deep

decision on multimodality treatment for


b
a

patients with a high risk score (score 2) to


Prognostic model for early stage OTSCC 149

be made. In the present study, 46% of Patient consent 14. Jerjes W, Upile T, Petrie A, Riskalla A,
patients who died from OTSCC had a high Hamdoon Z, Vourvachis M, et al. Clinico-
Not required. pathological parameters, recurrence, locor-
risk score (score 2), indicating a high
mortality rate in this group of patients. egional and distant metastasis in 115 T1-T2
From a biological point of view, the BD oral squamous cell carcinoma patients. Head
References Neck Oncol 2010;2:9.
model deals with depth of tumour inva-
sion, which indicates the efficacy of tissue 1. Anneroth G, Batsakis J, Luna M. Review of 15. Lin MJ, Guiney A, Iseli CE, Buchanan M,
penetration, and it also measures tumour the literature and a recommended system of Iseli TA. Prophylactic neck dissection in
malignancy grading in oral squamous cell early oral tongue squamous cell carcinoma
budding, which reflects EMT and tumour
carcinomas. Scand J Dent Res 1987;95: 2.1 to 4.0 mm depth. Otolaryngol Head Neck
dissociation at the invasive front. Tumour
229–49. Surg 2011;144:542–8.
budding has also been reported recently to 16. Ganly I, Goldstein D, Carlson DL, Patel SG,
have a strong association with the density 2. Bryne M, Koppang HS, Lilleng R, Stene T,
Bang G, Dabelsteen E. New malignancy O’Sullivan B, Lee N, et al. Long-term re-
of stromal myofibroblasts in oral SCC.30 gional control and survival in patients with
grading is a better prognostic indicator than
Briefly, the BD model gives a snapshot of ‘low-risk’ early stage oral tongue cancer
Broders’ grading in oral squamous cell car-
tumour invasiveness and its ability to cinomas. J Oral Pathol Med 1989;18:432–7. managed by partial glossectomy and neck
progress or metastasize. 3. Brandwein-Gensler M, Teixeira MS, Lewis dissection without postoperative radiation:
The current study has the advantage of CM, Lee B, Rolnitzky L, Hille JJ, et al. Oral the importance of tumor thickness. Cancer
homogeneity of tumour size and location, squamous cell carcinoma: histologic risk 2013;119:1168–76.
as all cases were T1 or T2 OTSCC. At the assessment, but not margin status, is strongly 17. Harrell Jr FE, Lee KL, Mark DB. Multivari-
same time, it was a multi-institutional predictive of local disease-free and overall able prognostic models: issues in developing
study involving two different geographic survival. Am J Surg Pathol 2005;29:167–78. models, evaluating assumptions and adequa-
regions (Finland and Brazil) and a large 4. Bundgaard T, Rossen K, Henriksen SD, cy, and measuring and reducing errors. Stat
patient cohort. In addition, the predictive Charabi S, Sogaard H, Grau C. Histopatho- Med 1996;15:361–87.
model introduced in this study is based on logic parameters in the evaluation of T1 18. Uno H, Cai T, Pencina MJ, D’Agostino RB,
simple and well-defined parameters; this is squamous cell carcinomas of the oral cavity. Wei LJ. On the C-statistics for evaluating
in contrast to previous models that have Head Neck 2002;24:656–60. overall adequacy of risk prediction proce-
included parameters such as lymphocytic 5. Almangush A, Bello IO, Keski-Säntti H, dures with censored survival data. Stat Med
host response, tumour keratinization, and Mäkinen LK, Kauppila JH, Pukkila M, 2011;30:1105–17.
et al. Depth of invasion, tumor budding, 19. Rusthoven K, Ballonoff A, Raben D, Chen
degree of nuclear pleomorphism, which
and worst pattern of invasion: prognostic C. Poor prognosis in patients with stage I and
have low reproducibility and are subject to
indicators in early-stage oral tongue cancer. II oral tongue squamous cell carcinoma.
high inter-observer variation. Cancer 2008;112:345–51.
In conclusion, the BD model is a simple Head Neck 2013;36:811–8.
6. Gil Z, Carlson DL, Boyle JO, Kraus DH, 20. O-charoenrat P, Pillai G, Patel S, Fisher C,
predictive model that can be applied easily Archer D, Eccles S, et al. Tumour thickness
Shah JP, Shaha AR, et al. Lymph node
on routine haematoxylin–eosin- or cyto- predicts cervical nodal metastases and sur-
density is a significant predictor of outcome
keratin-stained slides. Moreover, the BD in patients with oral cancer. Cancer vival in early oral tongue cancer. Oral Oncol
model is an important addition to existing 2009;115:5700–10. 2003;39:386–90.
prognosticators of OTSCC, and may even 7. Liang X. EMT: new signals from the inva- 21. Brennan S, Corry J, Kleid S, Porceddu S,
be useful for other oral subsites. We sug- sive front. Oral Oncol 2011;47:686–7. Yuen K, Rischin D, et al. Prospective trial to
gest that it could be used as a prognostic 8. Prall F. Tumour budding in colorectal carci- evaluate staged neck dissection or elective
index to help plan an individualized treat- noma. Histopathology 2007;50:151–62. neck radiotherapy in patients with CT-staged
ment protocol. Further validation of this 9. Liang F, Cao W, Wang Y, Li L, Zhang G, T1-2 N0 squamous cell carcinoma of the oral
model in other patient cohorts is recom- Wang Z. The prognostic value of tumor tongue. Head Neck 2010;32:191–8.
mended. budding in invasive breast cancer. Pathol 22. Yuasa-Nakagawa K, Shibuya H, Yoshimura
Res Pract 2013;209:269–75. R, Miura M, Watanabe H, Kishimoto S, et al.
10. Wang C, Huang H, Huang Z, Wang A, Chen Cervical lymph node metastasis from early-
Funding X, Huang L, et al. Tumor budding correlates stage squamous cell carcinoma of the
This study was funded by The Finnish with poor prognosis and epithelial–mesen- oral tongue. Acta Otolaryngol 2013;133:
Cancer Society, Sigrid Juselius Founda- chymal transition in tongue squamous cell 544–51.
carcinoma. J Oral Pathol Med 2011;40: 23. Grabenbauer GG, Rödel C, Ernst-Stecken A,
tion, and Libyan Ministry of Higher Edu-
545–51. Brunner T, Hornung J, Kittel K, et al. Neck
cation and Scientific Research.
11. Teramoto H, Koike M, Tanaka C, Yamada S, dissection following radiochemotherapy of
Nakayama G, Fujii T, et al. Tumor budding advanced head and neck cancer—for select-
Competing interests as a useful prognostic marker in T1-stage ed cases only? Radiother Oncol 2003;66:
squamous cell carcinoma of the esophagus. J 57–63.
None. 24. Yamauchi K, Fujioka Y, Kohno N. Sentinel
Surg Oncol 2013;108:42–6.
12. Lugli A, Karamitopoulou E, Zlobec I. Tu- node navigation surgery versus observation
Ethical approval mour budding: a promising parameter in as a management strategy for early tongue
colorectal cancer. Br J Cancer 2012;106: carcinoma. Head Neck 2012;34:568–72.
This study was approved by the institu- 1713–7. 25. Liu XC, Yang WJ, Zhang CP, Qu XZ, Wang
tional review boards of all hospitals in- 13. Huang SH, Hwang D, Lockwood G, Gold- T, Qiu LH, et al. An exploration of the
cluded. In addition, the approval of the stein DP, O’Sullivan B. Predictive value of surgical modality of sentinel lymph node
Brazilian Human Research Ethics Com- tumor thickness for cervical lymph-node biopsy in patients with cN0 tongue carcino-
mittee and the Finnish National Supervi- involvement in squamous cell carcinoma ma: an animal study. Oral Surg Oral Med
sory Authority for Welfare and Health of the oral cavity: a meta-analysis of reported Oral Pathol Oral Radiol Endod
(VALVIRA) were obtained. studies. Cancer 2009;115:1489–97. 2011;112:439–45.
150 Almangush et al.

26. Nason RW, Torchia MG, Morales CM, Thli- 28. Kim KY, Cha IH. Risk stratification of oral with intensity of tumor budding and density
veris J. Dynamic MR lymphangiography and cancer patients using a combined prognostic of stromal myofibroblasts in oral squamous
carbon dye for sentinel lymph node detec- factor including lymph node density and cell carcinoma. J Oral Pathol Med
tion: a solution for sentinel lymph node biomarker. J Cancer Res Clin Oncol 2013;43:199–204.
biopsy in mucosal head and neck cancer. 2012;138:483–90.
Head Neck 2005;27:333–8. 29. Chen CH, Hsu MY, Jiang RS, Wu SH, Chen Address:
27. Woolgar JA, Triantafyllou A. Pitfalls and FJ, Liu SA. Shrinkage of head and neck Alhadi Almangush
procedures in the histopathological diagno- cancer specimens after formalin fixation. J Department of Pathology
sis of oral and oropharyngeal squamous cell Chin Med Assoc 2012;75:109–13. Haartman Institute University of Helsinki
carcinoma and a review of the role of pa- 30. Marangon Junior H, Rocha VN, Leite CF, de FIN-00014
thology in prognosis. Oral Oncol Aguiar MC, Souza PE, Horta MC. Laminin- Finland
2009;45:361–85. 5 gamma 2 chain expression is associated Tel: +358 45 2044668; Fax: +358 9 19126675
E-mail: alhadi.almangush@helsinki.fi