Peeping Beyond TNM Stage: Prognostic Factors Affecting Oncological Outcomes in Surgically Treated Early Oral Cancer

Abstract

Surgery is the primary modality for the treatment of early oral cancer. The present study aims to evaluate the oncological outcomes, patterns of failure, and prognostic predictors of recurrence in patients of early oral cancer (Stage I and II) treated by surgery alone. It is a single institutional, observational retrospective cohort study conducted from 2012 to 2017.The study was approved by institutional ethics committee. All consecutive patients who underwent upfront curative surgery alone (wide excision of tumour + neck dissection) for pathologically proven early oral cavity SCC(Squamous Cell Carcinoma) pT1-2N0 were included in the study. 113 patients were included in the study after a median follow-up of 58.2 months. The median age was 58.5 years. 25 patients were stage I, and 88 patients were stage II. The most common subsite was buccal mucosa. There were 31 recurrences and 24 deaths. Using Kaplan Meier method, 3 and 5 year overall survival was 92% and 71.8%, respectively, while 3 year and 5 year recurrence-free survival was 77.9% and 69.4%, respectively. Perineural invasion and poor differentiation affected recurrence-free survival significantly (p value < 0.05). More than one-fourth of surgically treated early oral SCC patients developed recurrence. Presence of poorly differentiated histology and the perineural invasion were the high risk factors which hampered the recurrence free survival. High consumption of betel quid consumption in this part of the world leads to differences in the involved subsite from the Western literature. Adding adjuvant treatment in the presence of these adverse histopathological features may improve prognosis. Randomised studies are warranted to answer this dilemma.

Level of Evidence IV.

Introduction

According to GLOBOCON 2020, the global incidence of lip and oral cavity cancer is 3,77,713 cases per year, amidst which India accounts for 1,35,929 new cases per year [1]. Surgery is the primary treatment modality for most early oral cancer [2]. Elective neck dissection in early oral cancer lowers the risk of death/recurrence [3]. Survival over years has shown similar trends.The 5-year reported survival for stage I and II oral cancer is around 70%. Locoregional recurrences can occur in up to 30–35% of patients [4]. Bernier et al. [5] and Cooper et al. [6] identified precisely the high-risk factors which warrant adjuvant treatment in locally advanced head and neck cancer. Data pertaining to adjuvant treatment in early oral cancer is sparse and inconclusive. There is heterogeneity in behaviour highlighting different patterns of prognostic outcomes despite similar-stage cancers. This forces us to ponder that there might be interplay of minor adverse factors which enable these cancers to act differently. Several histopathological factors like tumour grade, lymph-vascular invasion(LVI), and perineurial invasion (PNI), grade of differentiation, close margins independently or in combination, may influence recurrence. In past several studies have been conducted catering to this grey area but most of the studies were on tongue cancer. Due to high consumption of smokeless tobacco, the incidence pattern of India differs from the West. The inconsistencies in tumor thickness and depth of invasion reporting also might have hampered the results.

The present retrospective cohort study aims to evaluate the long term oncological outcomes, patterns of failure, and prognostic predictors of recurrence in patients with early oral cancer (Stage I and II) treated solely by surgery.

Methods

The present study is a single institutional, observational retrospective cohort study conducted from 2010–2017. The study was approved by institutional ethics. All consecutive patients who underwent upfront curative surgery (wide excision of tumour + neck dissection) for pathologically proven oral SCC (Squamous Cell Carcinoma) < 4 cm in greatest dimension and less than 10 mm depth of invasion, attending head and neck surgery department in a tertiary care centre in North East India were included in the study. Patients who received neoadjuvant treatment, postoperative adjuvant treatment, history of previous head and neck cancer, distant metastasis and inadequate data were excluded from the study.

A total of 113 patients were enrolled in the study. A consistent treatment protocol was followed during the entire study period. Evaluation of patients included detailed history, physical examination and standard blood investigations. Pre-operative biopsy was done for all patients. Contrast-enhanced computed tomography(CECT) or magnetic resonance imaging(MRI) was done depending on the subsite involved. Pulmonary metastasis was assessed with a Chest X-ray or non-contrast CT (Computed tomography) Scan. Detailed data regarding patient demographics, clinical history, histopathology slides, report, details of treatment undertaken and outcome were collected retrospectively from the hospital electronic database and hospital records. All the patients underwent wide excision of the primary tumour and simultaneous neck dissection.

The available histopathology hematoxylin and eosin (H&E)–stained glass slides were retrieved and re reviewed by the pathology department for uniformity and was re staged according to AJCC (American Journal Committee on Cancer) Cancer Staging Manual 8th edition guidelines in size in greatest dimension and DOI (depth of invasion). PNI (peri neural invasion), LVI (lympho vascular invasion), grade of differentiation and margin status. No new parameter was added in reporting.

Size was measured in greatest dimension which was calculated by gross examination with histopathological correlation, reported in centimetre (cm). Depth of invasion was calculated by plump line drawn perpendicular from horizon of uninvolved basement membrane to deepest point of invasion. It was reported in millimetres (mm). Peri neural invasion was defined as tumor wrapping around the nerve or involving perineurium. Lympho vascular invasion included invasion in lymphatic vessel lumen with outside nearby tumor bulk.

Grade of differentiation was graded according to percentage of differentiated cells. Well differentiation = 75–100% differentiated cells, moderate differentiation = 50–75% differentiated cells, poor differentiation < 50% differentiated cells.

Margin distance was calculated from distance of tumor cells from edge of tissue.

Post operative histopathological details were noted.

The entire subject population’s OS (Overall survival) and loco regional recurrence-free survival (LRRFS) were analysed. OS was defined as the and time between the day of surgery and the date of death or last contact, whichever is earlier. LRFFS was defined as the time between the day of surgery to the date of local, regional or distant recurrence or the date of the last clinical follow-up. Pattern of failure (local, regional, distant) was assessed for the cases.

All statistical analyses were performed with SPSS version 21. The OS and LRFFS were calculated using the Kaplan–Meier method. Pattern of failure, and prognostic factors influencing survival were investigated through univariate and multivariate analyses using the Cox proportional-hazards model. Mean, and percentages were used for descriptive statistics. Association between categorical variables was assessed using the chi-square test, and risk was calculated using binary logistic regression. T-test was used to find the mean difference for continuous variables subject to the following normality. A p-value less than 0.05 was considered statistically significant.

Results

One hundred thirteen patients met criteria and were included in the study for analysis. The median follow-up was 60 months. The median age was 58.5 years. Males comprised 73.5% (83/113) of the study population. Hundred and one patients consumed arecanut with or without tobacco. Twenty seven patients had a family history of cancer, probably due to similar habits. Fort two patients had associated submucosal fibrosis due to high consumption of betel quid.

Twenty-five patients were stage I, and 88 patients were stage II. Elective neck dissection was not done in six superficial T1 lesions. The most common subsites were buccal mucosa (42 patients), followed by the tongue (28 patients) followed by lower gingivo buccal sulcus (20 patients).Table 1 shows detailed demographic, clinicopathological data of the study population.

Table 1.

Demographic and clinico pathological details of the study participants

Median age (years)	58.5 ± 11.8
Male:Female ratio	2.8:1 (83:30)
Tobacco consumption	101
Associated Submucosal fibrosis	42
Family history of cancer	27
Stage of tumor
Stage I	25
Stage II	88
Subsites
Buccal mucosa	42
Tongue	28
Lower gingivo- buccal sulcus	20
Lip	10
Upper gingivo- buccal sulcus	9
Retro molar trigone	4
Tumor size
< 2 cm	25
2–4 cm	88
Depth of invasion
< 5 mm	57
> 5 mm	56
Grade of differentiation
Well	68
Moderate	31
Poor	14
Lympho-vascular invasion
Present	29
Absent	84
Peri-neural invasion
Present	31
Absent	82
Margins
< 5 mm	14
> 5 mm	99
Recurrence
Yes	31
No	82
Surviving
Yes	89
No	24

There are 31 recurrences and 24 deaths. 3 and 5-year overall survival was 92% and 71.8% respectively, and recurrence-free survival was 77.9% and 69.4%, respectively (as shown in [Fig. 1]). Out of 31 recurrences, 19 patients had local recurrence in oral cavity, 6 patients had second primary tumors in esophagus (3), hypopharynx (2) and oropharynx (1). 4 tongue cancer and one buccal muccosa and one lower alveolus cancer developed second primary tumors. Regional recurrence was reported in 4 patients and distant metastasis in 2 patients. 19 deaths were due to cancer causes while rest were due to non cancer causes majorly due to associated co morbidities. Most of the patients included in the study had more than or equal to 5 mm margin. There were 4 cases with 3 mm margin with 10 cases having 4 mm margin. Adjuvant radiation was not considered after discussing with institutional tumor board as they were superficial tumors (mostly T1) and no other risk factor was present besides the close margin. No cases with positive margin cases were included as those patients eventually received adjuvant treatment.

Fig. 1.

Kaplan–Meier survival analysis: overall survival (A) and recurrence free survival (B)

None of the adverse features like the grade of differentiation, tumour stage, lympho-vascular invasion, peri neural invasion, close margin status, or depth of invasion showed statistically significant association with overall survival. [Table 2] depicts that out of all the adverse features, the perineurial invasion had the closest association with overall survival (p value-0.163). However, grade of differentiation and peri neural invasion had a statistically significant association with recurrence-free survival (p-value of 0.005 and < 0.001, respectively) as shown in [Table 3].Correlation of perineural invasion with recurrence free survival and overall survival is depicted in [Fig. 2].

Table 2.

Showing relation of histopathological factors with 3 year and 5 year overall survival

	3 Year (%)	5 Year (%)	p Value
OS	92.0	71.8
DOI < 5 mm	92.7	69.7	0.7340
DOI > 5 mm	90.0	75.9
WDSCC	95.6	78.6	0.8230
MDSCC	90.2	78.6
PDSCC	78.6	60.7
LVI -NO	92.8	73.3	0.6680
LVI -YES	89.7	67.0
PNI-NO	95.1	75.8	0.1630
PNI -YES	83.9	60.2
Margins-Close	92.9	67.5	0.81
Margins-Free	91.9	72.3

Table 3.

Showing relation of histopathological factors with 3 year and 5 year recurrence free survival (RFS)

	3 Year (%)	5 Year (%)	p Value
RFS	77.9	69.4
DOI < 5 mm	78.3	65.8	0.5350
DOI- > 5 mm	76.7	76.7
WDSCC	85.3	77.5	0.0050
MDSCC	78.6	78.6
PDSCC	61.3	48.0
T1	80.0	76.0	0.6160
T2	77.3	67.1
LVI -NO	82.1	71.9	0.1250
LVI -YES	65.5	61.9
PNI-NO	85.4	80.7	< 0.001
PNI -YES	58.1	40.7
Margins- Close	85.7	68.6	0.66
Margins- Free	76.8	69.0

Statistically significant p value < 0.05 is shown in bold

Fig. 2.

Depicting correlation of peri neural invasion with (a) overall survival and (b) recurrence free survival

Discussion

Oral cavity cancer is a leading cause of mortality and morbidity worldwide. The most common histopathological subtype is squamous cell cancer [7]. Surgery is the primary treatment for patients with early-stage squamous cell cancer of the oral cavity. Ellis MA et al. studied National Cancer Database (NCDB) and concluded that primary radiation for early-stage oral cancer was associated with increased mortality compared to surgery with a hazard ratio of 1.97 [8]. Many studies have revealed outcomes of early oral cancer. Luryi et al. [9] used the National Cancer Database (NCDB) data to study survival in 6830 patients with exclusively early-stage OSCC. Overall survival at five years was 69.7% in this study. Shim et al. reported that the 5-year OS and DFS in T1-2N0-1 tongue SCC were 80.8% and 80.2%, respectively. [10] In early tongue cancer, Ganly et al. observed 5-year DSS, OS and RFS rates of 86% and 79% and 70%, respectively. [6] According to Surveillance, Epidemiology, and End Results (SEER) database enrolling 6,791 patients with stage I and II SCC of the oral cavity, reported five-year disease-specific survival rates of stage I and II oral tongue cancers to be 83.5% [11]. In the present study, 3 and 5-year overall survival was 92% and 71.8% respectively, and recurrence-free survival was 77.9% and 69.4%, respectively. There was decline between three year and five year overall survival probably because of higher median age and deaths due to associated co morbid conditions, development of second primary tumors in esophagus and pharynx and loss to follow up.

In our study, grade of differentiation and peri neural invasion [graph 2] tended to influence recurrence significantly but not overall survival. However, peri neural invasion (PNI) was the most important factor amongst all the other adverse pathological features to affect overall survival, though it was not statistically significant (p value = 0.1630).

The perineural invasion has been the most investigated factor in early oral cancer. In a Taiwanese [12] study, PNI showed adverse impacts on both OS and LRFFS. Radiotherapy has proved to benefit loco regional recurrence by reducing paracrine integration between the tumour cells. Nair et al. [13] concluded that patients with perineural invasion have worse survival and are prone to develop recurrences. In a systematic review of thirteen retrospective studies, however evidence was lacking to recommend adjuvant RT for all patients with HNSCC with PNI [14]. There is not consensus till date and hence search must go on.

The present study showed that poorly differentiated tumours affect recurrence significantly. Arduino et al.showed that histological grading is an independent factor in predicting survival in patients with oral squamous cell carcinoma [15]. Shim et al. [10] and Liao et al [16] reported that the tumor grade affected OS and DFS rates.

Presence of poor differentiation on pre operative biopsy and clinico radiological nerve involvement can also aid in counselling patients regarding increased risk of recurrence. So, they can serve as biomarkers for prognostication.

Depth of invasion did not affect survival or recurrence in the present study. After the prospective randomized trial of D’Cruz et al. [3], elective neck dissection has become the standard of care in oral cancer. D’Cruz et al. also commented that elective neck dissection negates the impact of depth of invasion in early oral cancer. Elective neck dissection was done in all cases except six patients in present study.

3–5 mm margin didn’t impact recurrence or survival in the present study. So the question of discussion is what is the minimum safe distance where surgeon’s knife meets the patient in early oral cancer. A recent multicenter study study concluded for early tongue cancer, horizontal margin of 3.3 mm was a cut off for close margin which affected recurrence. [17]

Regarding adjuvant treatment in early oral cancer, there is no level I evidence to date. The beneficial role of PORT in preventing recurrence or increasing survival in pT1-T2N0-N1 is of equipoise. A recent review by Ivaldi et al [18] concluded that there is no evidence to guide the precise indications of PORT (Post operative radiotherapy) in early-stage OSCC. In the present study, one-fourth of the patients recurred, so definitely there is more to unveil beyond just TNM stage. It is the need of the hour to expose these obscured factors.

The limitation of the present study is that it is retrospective in nature. Due to the lack of the availability of complete data, the sample size was not substantial. The worst pattern of invasion or tumor budding was not included as a factor for analysis as it was not reported in all the patients because of the study's time frame. The study has not addressed different types of PNI (> 1 mm invasion, multifocal, extra tumoral, PNI density). To address these shortcomings, another study is being planned further.

The pros of the study were that it had a long median follow-up period of 60 months. It included solely patients with squamous cell cancer pT1-T2N0 oral cavity. As it is known, the presence of cervical metastasis affects survival outcomes. So, to avoid bias, N + necks were excluded from the study. Also, single modality treatment, i.e. surgery, was taken into consideration. There was standardisation in the reporting patterns. The strict inclusion and exclusion criteria allowed us to evaluate the effect of adverse prognostic variables on early oral cancer oncologic outcomes.

Regarding adjuvant radiotherapy in early oral cancer, multicenter randomized controlled study (AREST trial) is being conducted which may solve this perplexity of radiotherapy indications in early oral cancer.

Conclusion

Early oral cavity cancers generally have favourable outcome but are not without recurrences. However, more than one-fourth of early oral SCC patients treated by surgery alone developed recurrence. A poorly differentiated histology and perineural invasion are the risk factors that significantly affected locoregional recurrence-free survival. This accentuates the emerging need for consideration of adjuvant treatment in the presence of these adverse factors in early oral cancers. Further prospective randomized studies can unveil the specific indications for postoperative adjuvant treatment in the scenario described above.

Funding

No funding sources.

Declarations

Conflict of interest

There is no conflict of interest amongst the authors. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Ethical approval

The study is conducted according to ethical standards after obtaining ethical consent.

Informed consent

Consent was waived off due to retrospective nature of the study.