Although several histopathological parameters and grading systems have been described as predictive of the treatment response and outcome of oral squamous cell carcinoma (OSCC), none is universally accepted. A new scoring system, the histological risk model, was recently described to be a powerful predictive tool for recurrence and overall survival in OSCC. The aim of this study was to verify the predictive role of the histological risk model in a cohort of 202 patients at all stages of oral/mobile tongue squamous cell carcinoma (OTSCC). Demographic and clinical data were collected from the medical records and the tumours were evaluated using the histological risk model. Statistical analyses were performed using the χ 2 test, the Kaplan–Meier method, and the Cox regression model. The histological risk model showed no statistical correlation with demographic or clinical parameters and did not Predict the outcome of the OTSCC patients. However, multivariate regression analysis revealed a significant correlation of the clinical disease stage with the disease outcome. Despite major efforts to identify new predictive parameters and histological systems, clinical features are still the most reliable prognostic factors for patients with OTSCC.
Squamous cell carcinoma (SCC) is the most common malignancy of the oral cavity. The incidence varies worldwide, with India and South and Southeast Asia showing the highest rates. The global incidence is estimated to be 275,000 new cases per year, with approximately 14,000 new cases and 5000 deaths due to the disease in Brazil.
The tongue is the most common and deadliest site for oral SCC, particularly the oral (mobile) portion (OTSCC), where the tumour tends to show a more aggressive behaviour due to the high frequency of regional lymph node metastasis. The presence of lymph node metastasis is one of the most important prognostic factors for the survival of patients with OTSCC. Despite modern surgical techniques and new therapeutic strategies, particularly new chemotherapy drugs, the mortality rates for OTSCC continue to be high in most countries, with an overall 5-year survival rate usually below 50%. Furthermore, a study of 23,667 oral cancer cases treated between 1973 and 1999 in the USA (data from the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute) showed no statistical improvement in the survival of patients with oral cancer at localized and distant stages, whereas patients at regional stage showed a significant decrease in survival rates from 49.2% to 43.8%.
The poor prognosis of oral cancers has encouraged studies to search for new parameters to improve the management and predict the prognosis. Over the last century, different morphological features have been suggested as predictive of the prognosis of oral squamous cell carcinoma (OSCC), and different histological grading systems of malignancy have been developed on the basis of those features. In line with this, Brandwein-Gensler et al. proposed an updated and modified multiparameter system based on the evaluation of surgical specimens using three histological parameters: worst pattern of invasion (WPOI), lymphocytic host response (LHR), and perineural invasion (PNI); these were shown to have significant predictive power for recurrence and overall survival in OSCC. More recently, the same group validated this predictive model for patients with early-stage OSCC and showed that the model is able to predict patients at high risk of loco-regional recurrence and a worse prognosis.
The present study was conducted to evaluate the correlation of the histological risk model proposed by Brandwein-Gensler et al. with demographic and clinical features and the outcome of 202 patients with OTSCC. Herein, we provide evidence that the clinical stage of disease is more important with regard to the prognosis of patients with OTSCC than the histological risk model.
Patients and methods
Two hundred and two patients with primary OTSCC who were diagnosed and treated between 1980 and 2007 were included in this study. The inclusion criteria included complete demographic and clinical data, treatment based on radical surgery with or without postoperative radiotherapy, availability of paraffin-embedded blocks, and follow-up information of at least 5 years for survivors.
Demographic, clinical, and histopathological parameters
Demographic and clinical data, including gender, age, ethnicity, habits such as smoking and alcohol consumption, TNM stage, status of the surgical margins, recurrence, presence of a second primary tumour, and survival, were obtained from the patient records. The surgical margin, identified as the closest distance between the tumour and the surgical resection edge (both in deep muscle and on the lateral mucosa), was categorized into two groups based on a cut-off value of 5 mm. Margins of less than 5 mm were considered involved, and margins of 5 mm or more were classified as free. The outcomes investigated were: (1) overall survival, the time from treatment initiation until death or last follow-up information; (2) disease-specific survival, the time from treatment initiation until death due to cancer or last known date alive; and (3) disease-free survival, the time from treatment initiation until diagnosis of the first recurrence (local, regional, or distant) or last follow-up information for those without recurrence.
Paraffin-embedded blocks of all cases were retrieved and new sections were stained with haematoxylin and eosin. Slides were assessed independently by three investigators, who were blinded to the demographic and clinical data and outcomes; any disagreement was resolved by discussion. Tumours were scored regarding WPOI, LHR, and PNI, and classified according to the histological risk model proposed by Brandwein-Gensler et al. Before tumour grading, slides with WPOI of each case were selected, and observers calibrated. To determine the intra-observer degree of agreement, 50 random samples were selected and examined twice with a 2-week interval by each observer. The intra-observer Cohen’s Kappa coefficient ( k ) ranged from 0.77 to 0.89. The inter-observer k -value, representing the degree of agreement among readers, was 0.84 for all samples.
Correlations between the histopathological risk model and demographic and clinical parameters of the tumours were assessed by cross-tabulation and χ 2 test. For the statistical analyses, low and intermediate risk were grouped together and compared with high risk. Survival curves were constructed based on the Kaplan–Meier method and compared with the log-rank test. For the multivariate survival analysis, the Cox proportional hazard model with a stepwise method including all parameters was employed. The level of significance was set at 5% ( P ≤ 0.05).
Of the 202 OTSCC patients included in the study, 110 (54.5%) were men. The patients ranged in age from 21 to 95 years, with a median of 58 years. Most of the patients were Caucasian ( n = 180, 89.1%) and reported smoking (152 patients, 75.2%) and drinking alcohol (141 patients, 69.8%). Regarding the clinical stage, 115 (61.5%) patients were classified as early-stage (stages I and II) and 72 (38.5%) as advanced-stage (stages III and IV). Surgery as monotherapy was performed in 124 (61.4%) of the patients, whereas 78 (38.6%) were treated with a combination of surgery and postoperative radiotherapy. Radiotherapy was mainly advocated for those with advanced stages of the disease, positive neck disease, or surgical margins less than 5 mm. Twenty-one (10.4%) cases had surgical margins of less than 5 mm. During follow-up, 52 (25.7%) patients developed local recurrence (34 local only, 10 local and regional, five local and distant, and three local, regional, and distant), 39 (19.3%) developed regional recurrence (21 regional only, 10 local and regional, five regional and distant, and three local, regional, and distant), and 18 (8.9%) patients developed distant recurrence (five distant only, five local and distant, five regional and distant, and three local, regional, and distant). The overall survival ranged from 1 to 262 months, with a median of 59 months (mean 80.2 months). Fifty-eight patients (28.7%) died due to the tumour.
The histological risk model based on the Brandwein-Gensler score system classified 26 (12.9%) tumours as low risk, 100 (49.5%) as intermediate risk, and 76 (37.6%) as high risk. With respect to the three parameters individually, WPOI type 1, 2, or 3 was identified in 37 (18.3%) samples, type 4 in 150 (74.3%) samples, and type 5 in 15 (7.4%) samples ( Fig. 1 ). For LHR, 91 (45.0%) cases were classified as type 1 (strong), 45 (22.3%) as type 2 (intermediate), and 66 (32.7%) as type 3 (weak) ( Fig. 2 ). Perineural invasion was detected in only 29 (14.4%) samples – in small nerves (diameter <1 mm) in 28 (13.9%), and in large nerves (diameter ≥1 mm) in one (0.5%) ( Fig. 3 ).
The correlation between the histological risk model and demographic and clinical parameters is shown in Table 1 . No significant correlations were observed between the epidemiological and clinical parameters of the OTSCCs and the histological risk model. Since the histological risk model was found to be significantly associated with early stage tumours (clinical stage I – T1N0M0 and stage II – T2N0M0), we verified its association with subgroups of clinical stage. No significant associations were observed comparing stage I with stage II ( P = 0.08) or stage III with stage IV ( P = 0.76).
|Parameter||Histological risk model||P -value|
|Low/intermediate n (%)||High n (%)|
|<58 years||65 (51.6)||44 (57.9)|
|≥58 years||61 (48.4)||32 (42.1)||0.38|
|Male||71 (56.3)||39 (51.3)|
|Female||55 (43.7)||37 (48.7)||0.48|
|Caucasian||112 (88.9)||68 (89.5)|
|Non-Caucasian||14 (11.1)||8 (10.5)||0.89|
|No||29 (23.8)||13 (18.1)|
|Yes||93 (76.2)||59 (81.9)||0.35|
|No||39 (31.7)||15 (20.8)|
|Yes||84 (68.3)||57 (79.2)||0.10|
|Early (I + II)||73 (62.9)||42 (59.2)|
|Advanced (III + IV)||43 (37.1)||29 (40.8)||0.60|
|> 5 mm||112 (88.9)||69 (90.8)|
|< 5 mm||14 (11.1)||7 (9.2)||0.67|
|No||97 (77.0)||53 (69.7)|
|Yes||29 (23.0)||23 (30.3)||0.25|
|No||104 (82.5)||59 (77.6)|
|Yes||22 (17.5)||17 (22.4)||0.39|
|No||113 (89.7)||71 (93.4)|
|Yes||13 (10.3)||5 (6.6)||0.36|
|No||98 (79.0)||60 (82.2)|
|Yes||26 (21.0)||13 (17.8)||0.59|
The 5-year overall survival, disease-specific survival, and disease-free survival rates were 76.8%, 66.7%, and 71.8%, respectively. Smoking habit ( P = 0.02), clinical stage ( P = 0.0002), and treatment ( P = 0.01) were observed to have a significant impact on overall survival ( Table 2 ). For disease-specific survival, statistically significant differences were demonstrated for ethnicity ( P = 0.03), clinical stage ( P < 0.0001), treatment ( P = 0.02), and histological risk model ( P = 0.03) ( Table 2 ). The incidence of recurrence (disease-free survival) was significantly influenced by smoking habit ( P = 0.02), clinical stage ( P = 0.0001), and therapeutic modality ( P = 0.01) ( Table 2 ).
|Parameter||Overall 5-year survival||P -value||Disease-specific 5-year survival||P -value||Disease-free 5-year survival||P -value|
|Early (I + II)||83.6%||80.0%||80.8%|
|Advanced (III + IV)||62.8%||0.0002||48.8%||<0.0001||58.7%||0.0001|
|Surgery + radiotherapy||67.5%||0.01||56.9%||0.02||63.0%||0.01|
|> 5 mm||77.5%||68.3%||76.7%|
|< 5 mm||69.9%||0.19||61.5%||0.27||57.6%||0.13|
|Histological risk model|