33.1 Introduction

Throughout the head and neck, either surgery or radiotherapy can be utilized as the primary modality for curative intent ablation of locoregional disease. Although organ-preserving definitive radiotherapy is an acceptable, and in many cases preferred, treatment for squamous cell carcinomas (SCCs) of the oropharynx, larynx, hypopharynx, and nasopharynx, the vast majority of oral cavity cancers are best managed with ablative surgery with or without adjuvant therapy. This is because (1) the oral cavity is particularly sensitive to adverse sequelae from radiotherapy at the doses required for definitive treatment, and (2) oral cavity squamous cell carcinoma (OCSCC) is considered radioresistant in comparison to SCCs of other head and neck subsites. Nevertheless, definitive radiation-based approaches can play an important role in OCSCC for a variety of reasons. In this chapter, we will focus on external beam radiotherapy (EBRT)-based approaches. Brachytherapy is discussed in Chapter 34.

33.2 Principles of Definitive External Beam Radiotherapy for Oral Cavity Cancer

Definitive radiation is a viable curative approach in select patients with early stage OCSCC (T1-2N0) to avoid functional and cosmetic defects, and is supported by National Comprehensive Cancer Network guidelines as an alternative to surgery. ¹ However, several important caveats are notable. Importantly, the majority of the data showing excellent results with definitive radiation in this setting utilized either intraoral cones, a type of EBRT where a radiation source emitting either orthovoltage photons or electrons is placed inside the mouth abutting the mucosa, or brachytherapy, where radioactive sources are temporarily or permanently implanted directly into the tumor. These can be utilized either alone or in combination with standard linear accelerator-based EBRT. Unfortunately, few modern radiation oncology departments have intraoral cones available, and brachytherapy for OCSCC is also not widely used due to a relatively limited number of radiation oncologists with expertise in the delivery of brachytherapy for head and neck cancer. The most common radiation modality utilized in modern practice for head and neck cancer is, by far, linear accelerator-based EBRT. However, this modality is typically avoided for early stage OCSCC unless combined with a boost delivered via intraoral cone or brachytherapy, given that linear accelerator-based EBRT has relatively high toxicity and, potentially, relatively poor local control in comparison to other modalities for OCSCC. For example, multiple studies have suggested that EBRT without brachytherapy is associated with significantly higher local recurrence rates in early stage oral tongue SCC. ^{2 – 4} Additionally, in patients with tumors thicker than 4 mm, some type of elective neck treatment, such as neck dissection or elective neck irradiation, is typically indicated given the high risk of micrometastatic nodal spread.

For locally advanced (T3-T4 or N +) OCSCC, upfront surgical resection followed by adjuvant radiotherapy is the standard of care. Definitive radiation is typically reserved for patients with unresectable disease, those who are medically inoperable, or who refuse surgery. Multiple studies have demonstrated superior outcomes with surgery combined with pathology-directed adjuvant therapy, in comparison to definitive EBRT alone, with or without chemotherapy. For example, some of the only prospective randomized data come from a small randomized trial conducted in Singapore, where 119 patients with locally advanced head and cancer were randomized to either surgery and adjuvant radiation or definitive cisplatin and 5-FU-based concomitant chemoradiation. ⁵ Although in the entire cohort there was no difference in survival or locoregional control with either approach, in the subgroup of 32 patients with oral cavity primaries, there was dramatically better disease-specific survival (DSS) in patients undergoing surgery and postoperative radiation versus those receiving definitive chemoradiation (▶ Fig. 32.2, 5-year DSS: 68 vs. 12%, P = 0.038). There were trends toward lower locoregional relapse (approximately 30 vs. 70% at 5 years, P = 0.355) and distant relapse (5-year distant recurrence: 8 vs. 50%, P = 0.05) in the surgery arm. Only 4 out of 19 patients with oral cavity tumors undergoing definitive chemoradiation avoided surgery after completing treatment. Although this trial had multiple limitations, including limited a number of patients, early termination due to poor accrual, lack of intent-to-treat analysis, antiquated radiation techniques, somewhat lower radiation doses than used in modern practice (66 Gy in 33 fractions), and reliance on unplanned subgroup analyses, the absolute magnitude of the survival and relapse differences in the two arms are striking and suggest that definitive chemoradiation may be suboptimal for locally advanced OCSCC.

Multiple retrospective studies have also compared surgical and radiation-based approaches for oral cavity cancer (▶ Table 33.1). For example, a bi-institutional retrospective study of 104 OCSCC patients undergoing either surgery and adjuvant radiotherapy or definitive chemoradiation showed markedly worse locoregional control, DSS, and overall survival for patients undergoing definitive chemoradiation in multivariable models. ⁶ In addition, patients undergoing chemoradiation were more likely to require feeding tubes for at least 30 days or more after treatment (30 vs. 11%). Similarly, an analysis of 6,900 stage III-IVA OCSCC patients from the National Cancer Data Base showed vastly better survival in propensity score-matched cohorts undergoing surgery and postoperative radiation versus those undergoing chemoradiation (3-year OS: 53.9 vs. 37.8%). ⁷ This difference appeared to be driven by clinical stage T3-T4a tumors, whereas survival was not significantly different for those with T1-T2 disease. Similar results were observed in an analysis of the Surveillance, Epidemiology, and End Results (SEER) database ⁸ and a variety of smaller studies. ^{4 , 9 – 11}

All of these studies are limited by retrospective design and near universal imbalances in clinical and demographic features favoring the surgical cohort. Nevertheless, these studies and other noncomparative studies reporting results with definitive radiation for OCSCC (▶ Table 33.2) strongly suggest that OCSCC is relatively radioresistant. Despite this, long-term cure can be achieved in a minority of cases without surgery. Thus, some OCSCCs must have greater inherent sensitivity to radiation and chemotherapy than others. One attractive approach would be to only attempt organ preserving chemoradiation in the subset of patients with tumors that are inherently sensitive to DNA-damaging therapies. The utility of induction chemotherapy to chemo-select patients for definitive radiation has shown utility in larynx cancer, ¹⁴ but has been less well studied in oral cavity cancers. The chemo-selection paradigm hypothesizes that patients responding to chemotherapy have increased biologic sensitivity to DNA-damaging therapies, such as radiation, and likely will be better candidates for organ preservation. This was tested in a prospective single arm phase II trial at the University of Michigan. ¹³ Nineteen patients with stage III-IV OCSCC without bone invasion underwent 1 cycle of induction chemotherapy with cisplatin/carboplatin and 5-FU. Three weeks later, patients with at least a 50% clinical response underwent definitive platinum-based chemoradiation, whereas nonresponding patients underwent surgery and postoperative radiation. The trial was stopped after meeting early stopping rules for futility. In total, 10 patients responded to induction chemotherapy and underwent definitive chemoradiation, with 30% of them having disease control at 5 years. Of the 9 nonresponding patients, 4 died before surgery could be performed. Compared to a matched group of patients treated with surgery and postoperative radiation at the same institution, patients enrolled on this trial had worse locoregional control (5-year locoregional control (LRC): 72 vs. 26%, P < 0.001), DSS (5-year DSS: 75 vs. 46%, P = 0.001), and overall survival (5-year overall survival [OS]: 65 vs. 32%, P = 0.01). Moreover, gastrostomy tube (9 vs. 42%), osteoradionecrosis (2 vs. 16%), and tracheostomy tube rates (0?vs. 5%) were all higher in the cohort enrolled on this trial in comparison to the historical matched surgery cohort.

Some of the most favorable results for treating locally advanced OCSCC definitively with EBRT came from the University of Chicago. For example, in a retrospective analysis of 39 patients with T4 OCSCC enrolled on 4 prospective phase II protocols, including 16 with bone invasion, 5-year local control was 75% with definitive chemoradiation with or without induction chemotherapy, with 5-year cause-specific and overall survival of 67 and 56%, respectively. ¹⁸ Osteoradionecrosis occurred in 18% of patients, and 29% experienced severe long-term complications. A larger study of the same institution of patients with locally advanced oral cavity cancers enrolled on one of 15 clinical trials, comparing 111 patients undergoing definitive chemoradiation to 27 patients undergoing primary surgery and postoperative chemoradiation, showed no difference in overall survival or progression-free survival. ¹² The rate of osteoradionecrosis was identical to what was observed in the T4 subset, and deaths from treatment complications were seen in 7%. However, it is notable that the results from the University of Chicago are difficult to extrapolate to other institutions, given that they utilize both a unique radiation regimen (75 Gy given in 1.5 Gy fraction twice daily with one week on, one week off of radiation) and a unique multi-agent radiosensitizing chemotherapy regimen (5-FU and hydroxyurea combined with either cisplatin or paclitaxel) that are rarely utilized at other centers. It is unclear if these regimens are central to the excellent results reported, as few, if any, other institutions employ them. On the other hand, an alternative explanation is that the University of Chicago preferentially treated locally advanced OCSCC with chemoradiation rather than surgery, so there may have been less selection bias against the chemoradiation patients. Another study that demonstrated fairly good results with definitive chemoradiation for stage III-IVB OCSCC was reported by Fuwa and colleagues, achieving local control in 69% of cases at 3 years. ¹⁹ Local control was 78% for T2-T3 tumors and 51% for T4 tumors. However, they used intra-arterial platinum chemotherapy, sometimes combined with systemic chemotherapy, which is rarely used at other institutions. In addition, 31% of patients received brachytherapy. Thus, these results are also difficult to extrapolate to other settings.

For those undergoing definitive EBRT with locally advanced disease, concomitant chemotherapy should be delivered in patients healthy enough to receive it. In the MACH-NC Collaborative Group meta-analysis, ²⁵ which demonstrated improved locoregional control and overall survival with the addition of concomitant chemotherapy to radiation for head and neck cancers, 21% of patients had OCSCC. In patients unable to receive chemotherapy due to medical contraindications or refusal, hyperfractionated or accelerated radiation is an alternative treatment intensification approach that can improve both locoregional control and overall survival in head and neck cancer. ²⁶ OCSCC made up 12.6% of patients in the MARCH meta-analysis. There is no prospective randomized data to support the use of cetuximab with definitive radiation for oral cavity cancer, given that these patients were excluded from the only trial demonstrating a benefit from this therapy with radiation. ²⁷

In summary, radiation is an acceptable alternative to surgery in select cases of early stage T1-2N0 OCSCC. If radiation is used in this setting, strong consideration should be given to using intraoral cones or brachytherapy either for the entire treatment or as a boost, given that the majority of historical series used these modalities. Nevertheless, surgery may be preferred in many patients with early stage OCSCC due to excellent control, favorable toxicity profile, and quicker treatment times with this modality. For those with more advanced disease, the vast majority of the data suggest that primary external beam radiation with or without chemotherapy does not result in equivalent oncologic outcomes to surgery followed by postoperative radiation. However, EBRT can still play an important role in those with unresectable disease, medical inoperability, or who refuse surgery due to morbidity or functional concerns. It is important to note that although rates of locoregional control with definitive chemoradiation are suboptimal, chemoradiation can still be curative in a subset of these patients that are either unable or unwilling to undergo surgery, and should not be withheld from patients amenable to curative-intent nonoperative therapy.