Oncological outcomes in patients undergoing major glossectomy for advanced carcinoma of the oral tongue

Abstract

Introduction

Major glossectomy is the treatment of choice in locally advanced tongue cancer. It remains the only option in the presence of recurrent or residual disease. The long-term outcomes for patients undergoing major glossectomy have traditionally been poor, with significant morbidity and poor oncological outcomes. The aim of this study was to report on oncological outcomes in patients undergoing major glossectomy.

Methods

All patients undergoing major glossectomy between 2014 and 2018 were included in the study. The data of 85 patients with advanced carcinoma of the oral tongue were evaluated. All were under the care of a single surgical and reconstructive team at two hospitals in Mumbai.

Results

The median patient age was 45 years. At the most recent follow-up, 55 patients (65%) were alive, 47 of whom were disease free. Twenty-nine patients (34%) had locoregional recurrence and twenty-five (29%) had distant metastasis. At a median follow-up of 19 months, rates for 2-year locoregional control, disease free survival (DFS) and overall survival (OS) were 69%, 61% and 62% respectively. Perinodal extension demonstrated a trend towards poor DFS (p=0.060), as did perineural invasion (p=0.055). Node positivity was a significant factor for poor OS, DFS and locoregional control. Multiple node involvement was significantly associated with poor OS on multivariate analysis (p=0.002).

Conclusions

Node positivity and multiple node involvement were associated with poor outcomes. Major glossectomy may be offered as a curative option for selected patients with advanced carcinoma of the oral tongue with node negative or limited neck nodal disease (N1).

Introduction

Head and neck cancer is a common cause of mortality in India.¹ Cancers of the oral cavity are significant public health problem in India.^2,3 Tongue is the most common subsite for oral cavity cancers in Western world² while gingivobuccal complex cancer is predominant in India. The majority of publications from India have focused on gingivobuccal complex cancer, with a paucity of literature on tongue cancer. Advanced cancer of the tongue is associated with high morbidity and poses a therapeutic challenge for oncologists. These patients have significant pain, difficulty in swallowing (which results in fluid/electrolyte imbalance), malnutrition, difficulty in speaking and halitosis. They also experience associated high post-treatment morbidity and poor oncological outcomes. Despite advances in treatment options and improved rehabilitation methods, outcomes have not improved proportionately.^4,5

Major glossectomy is a radical form of treatment with significant associated morbidity. Usually, it is considered a palliative procedure because of poor survival in this group of patients.^5,6 Advances in microvascular reconstructive surgery have shown good results in restoring mucosal integrity and functional rehabilitation. The present analysis was undertaken with the aim of reporting oncological outcomes and prognostic factors for oral tongue carcinoma patients undergoing major glossectomy.

Methods

Data from a prospectively maintained database of patients undergoing major glossectomy were analysed retrospectively. All patients with locally advanced carcinoma of the oral tongue (cT3/T4) who underwent major glossectomy under a single surgical and reconstructive team at two hospitals in Mumbai between March 2014 and March 2018 were included in the study. All staging was done using the seventh edition of the American Joint Committee on Cancer staging system. Patients with inoperable nodes, primary disease/nodal disease encasing the carotid artery by >180°, disease involving the prevertebral fascia, disease extending into the skull base or hyoid bone and metastatic disease were excluded.

All patients had histologically proven, locally advanced squamous cell carcinoma of the tongue and underwent major glossectomy with laryngeal preservation and neck dissection as well as appropriate reconstruction of surgical defects. For the purposes of this study, major glossectomy comprised anterior two-thirds glossectomy, near total glossectomy (anterior two-thirds plus part of the base of tongue), total glossectomy and composite resection of the tongue (glossectomy involving more than half of the tongue with en bloc mandible resection). All patients were treated with adjuvant radiotherapy/concurrent chemoradiotherapy depending on histopathology results.

For the first two years after completion of adjuvant radiotherapy/chemoradiotherapy, all patients were followed up at three-monthly intervals, then every six months until five years and annually thereafter. Clinical examination of the oral cavity, larynx and neck was performed during follow-up visits. Radiological investigations were undertaken when necessary.

Locoregional control (LRC) was estimated from the date of surgery and date of local and/or regional recurrence. Disease free survival (DFS) was estimated from the date of surgery and date of any one of following: local recurrence, regional recurrence, distant metastasis, second primary tumour or death due to non-cancer causes. All patients had a minimum follow-up period of six months. Analysis was performed using SPSS^® Statistics version 20 (IBM, New York, US). Kaplan–Meier survival curves were produced. Prognostic factors affecting the various disease endpoints were analysed with the logrank test for univariate analysis and the Cox proportional hazards model for multivariate analysis. Age, sex, node positivity, multiple node involvement and perinodal extension were included in multivariate analysis.

Results

The study cohort comprised 85 patients. The median patient age was 45 years (range: 17–71 years) and the majority were male with a male-to-female ratio of 3:1. Eighty patients (94%) had a history of tobacco/smokeless tobacco consumption. The median tumour size was 4cm and the median tumour thickness was 2cm. Sixty-seven patients had cT4 disease at presentation.

The surgical procedures were divided into two groups for subanalysis: near total/total glossectomy and tongue composite resection, where bone was also resected with tongue. Seventy-one patients (84%) had near total/total glossectomy defects and the rest had composite resections (16%). None of the patients underwent laryngectomy/laryngopharyngectomy along with major glossectomy.

Margins were free (>5mm) in 80 patients (95%) while the remaining 5 patients (5%) had a close margin (<5mm). None had positive margins. Sixty-two patients (73%) had neck nodes positive for metastasis, with forty-seven (55%) having multiple node involvement. Fifty patients (59%) had perinodal extension, perineural invasion was present in 55 patients (65%) and 21 patients (25%) had lymphovascular emboli. All demographic and histopathological parameters are shown in Table 1.

Table 1.

Basic patient demographics and baseline characteristics

Variable	n
Median age	45 years (mean: 48 years, range: 17–71 years)
Male-to-female ratio	65:20 (3:1)
Smoking status  Smoker  Non-smoker	80 (95%) 5 (5%)
T stage  T3  T4	25 (30%) 67 (70%)
N stage  N0  N1  N2	23 (27%) 15 (17%) 47 (56%)
Stage  IV	85 (100%)
Reconstruction  Primary closure  Pedicle flap  Microvascular (ALT/FOCF)	1 (1%) 22 (26%) 62 (73%)

ALT = anterolateral thigh; FOCF = fibular osteocutaneous flap

Sixty-two patients (73%) had microvascular reconstruction for the defect following surgery while pedicle flap reconstruction was performed in 22 patients (26%). One patient underwent primary closure for an anterior two-thirds glossectomy defect because of associated comorbidities.

The surgical and medical complications were documented prospectively in treatment charts. Five patients had major complications; these included four flap failures and one patient died following surgery owing to sudden cardiac arrest. Eight patients had minor complications such as partial flap necrosis, bleeding and gaping of sutures. The median period for tracheostomy was 10 days (range: 6–17 days). Decannulation occurred earlier in the free flap group than for pedicle flap patients (median day 7 vs day 10). Nasogastric tube feeding lasted a median of 21 days after surgery in both groups.

All patients were referred for adjuvant radiotherapy/chemoradiotherapy depending on histopathology results. Thirty-five patients received adjuvant radiotherapy only; four patients had undergone radiotherapy prior to surgery and two patients defaulted on their adjuvant treatment. Concurrent chemoradiotherapy was given in 51 patients. The median dose of radiation was 60Gy cisplatin, given weekly, and the median number of concurrent chemotherapy cycles was 5 (range: 3–6 cycles). None of the patients died during adjuvant treatment.

At the time of analysis, 55 patients were alive and 50 patients had no evidence of disease. The median follow-up duration of alive patients was 19 months (range: 6–51 months). Overall, 29 patients (31%) had locoregional recurrence while 25 (26%) had distant metastasis with the lung being the most common site of involvement. The overall survival (OS) rate for the cohort was 62% at two years and 56% at three years. The LRC rate was 69% and 60%, and the DFS rate was 61% and 48% respectively (Fig 1).

Figure 1.

Kaplan–Meier curves for locoregional control (LRC), disease free survival (DFS) and overall survival (OS)

On univariate analysis, there was no significant difference in OS (p=0.121), LRC (p=0.587) or DFS (p=0.291) between the near total glossectomy/total glossectomy and composite resection groups. Node positive disease had a significant effect on DFS (p=0.011) and OS (p=0.024) (Table 2). In the node positive group, multiple node involvement (N2b, N2c and N3) was significantly associated with lower LRC (p=0.004), DFS (p=0.001) and OS (p=0.002). Although perinodal extension did not affect LRC or OS, there was a trend for association with poor DFS but this did not reach statistical significance (p=0.060). Similarly, there was a trend towards lower DFS in patients with perineural invasion; however, this also narrowly missed being statistically significant (p=0.055). On multivariate analysis, only multiple node involvement was significantly associated with poor OS (p=0.028).

Table 2.

Univariate analysis of factors affecting locoregional control (LRC), disease free survival (DFS) and overall survival (OS) rates

		n	2-year LRC	p-value*	2-year DFS	p-value*	2-year OS	p-value*
Sex	Male	65	68%	0.753	63%	0.312	63%	0.512
	Female	20	72%		51%		57%
Age	>40 years	58	68%	0.584	58%	0.891	61%	0.945
	<40 years	27	69%		63%		65%
Nodal status	Negative	23	77%	0.176	77%	0.011	79%	0.024
	Positive	62	66%		55%		55%
N stage	N0, N1	38	83%	0.004	78%	0.001	77%	0.002
	N2, N3	47	58%		47%		51%
Pathological tumour size	<4cm	42	72%	0.772	66%	0.705	67%	0.324
	>4cm	43	66%		55%		61%
Tumour thickness	<2cm	41	77%	0.361	72%	0.249	69%	0.369
	>2cm	44	59%		49%		55%
Lymphovascular invasion	Absent	64	69%	0.924	62%	0.314	66%	0.510
	Present	21	67%		56%		59%
Perineural invasion	Absent	30	78%	0.222	75%	0.055	70%	0.296
	Present	55	64%		54%		58%
Perinodal extension	Absent	35	76%	0.125	68%	0.060	69%	0.175
	Present	50	64%		56%		57%

^*Logrank test

Two patients developed second primary tumours. After disease recurrence, median survival was 4 months (range: 1–10 months).

Discussion

Tongue cancer is associated with a worse prognosis than that for other similarly staged tumours of the oral cavity. Advanced squamous cell carcinoma of the tongue is a devastating disease, causing pain, difficulty in speaking and swallowing, and halitosis. Many surgeons are reluctant to suggest highly morbid surgery and prefer major glossectomy as a salvage procedure after radiotherapy or chemoradiotherapy as a primary treatment modality. Even after best efforts, OS remains poor.^7,8.Survival is dismal in locally advanced tongue cancer, especially in node positive disease.⁹

The majority of the literature on major glossectomy comes from previous decades when microvascular reconstruction was not the gold standard and radiation techniques were primitive. Reconstruction of these surgical defects is of paramount importance, not only for early rehabilitation but also to commence adjuvant treatment on time.^10–14

In our current study of major glossectomy defects, 73% underwent reconstruction using a microvascular free flap and 27% with a pedicle flap. The factors favouring a pedicle flap over a free flap were comorbidities and cost effectiveness (especially important in the Indian population). Patients with microvascular reconstruction had minimal postoperative morbidity (eg flap failure [4.7%] and other minor complications) compared with historical cohorts, enabling early commencement of adjuvant treatment. Bone resection along with partial glossectomy was carried out in 16% of patients. Higher morbidity in terms of swallowing and speech is to be expected in these patients because of associated loss of bony support. A series published by Lin et al in 2015 showed that 75% of patients undergoing major glossectomy had acceptable swallowing function after reconstruction.¹⁵ However, these functional outcomes were not evaluated by the authors and this is a limitation of their study.

Our study has demonstrated better LRC, DFS and OS than previously reported outcomes.^16–19 In our cohort, the two-year OS rate was 62% and the LRC rate was 69%. In a small series reported in 2019 by Quinsan et al, the five-year OS and cancer specific survival rates were 19% and 30.8% respectively.¹⁷ Five-year outcomes were not within the scope of our current study. The composite defects where bone was removed along with partial glossectomy were analysed separately for clinical outcomes. There was no significant difference in OS, LRC or DFS in this subgroup compared with near total glossectomy/total glossectomy patients. Table 3 summarises survival outcomes published in the literature.

Table 3.

Comparison of survival outcomes in the literature

Paper	Number of patients	Overall survival	Disease free survival	Cancer specific survival
Sinclair, 201119	30		40% (1-year)
Shockley, 201318	22	68% (1-year)
Pillai, 201716	60		86% (1-year)
Quinsan, 201917	22	19% (5-year)		30.8% (5-year)
Present study	85	56% (3-year)	48% (3-year)

Perineural invasion has been quoted as an independent prognostic factor as recently as 2018²⁰ but in our study, it did not reach statistical significance. On univariate analysis, the two factors that were associated with survival were presence of positive nodes and number of positive nodes but on multivariate analysis, only multiple node involvement remained statistically significant for poor OS. In future studies, it might be interesting to look at the lymph node ratio.

In our analysis, perinodal extension showed a trend towards poor DFS. This has already been well documented in the literature as a significant independent prognostic factor.⁹ The incidence of distant metastasis was 29%, with lung and bone being the most common sites. This is higher than that reported in the literature (20% distant metastasis in advanced head and neck cancer).²¹ This high incidence could be because of improved LRC owing to well evolved radiotherapy and chemotherapy protocols in adjuvant settings. This highlights the need for a possible role of systemic maintenance chemotherapy to reduce the incidence of distant metastasis.

Strengths of the present study include its uniform treatment protocol, the consecutive patient series and the reasonably long follow-up period. Our study is one of the largest in the last decade reporting clinical outcomes of major glossectomy. Its limitations include lack of formal assessment of quality of life, swallowing and speech evaluation, and comparison with similar groups treated with non-surgical protocols (radiotherapy/chemoradiotherapy). Our future work will focus on further follow-up of this group of patients along with a formal assessment of functional outcomes and quality of life.

Conclusions

Major glossectomy for advanced carcinoma of the oral tongue is feasible for a selected group of patients. The oncological outcomes are encouraging in terms of LRC and OS. Node positivity and multiple node involvement were associated with poor oncological outcomes. Major glossectomy may be offered as a curative option for selected patients with advanced carcinoma of the oral tongue with node negative or limited neck nodal disease (N1).