35 Adjuvant Radiotherapy for Oral Cavity Cancer
Advanced cancers of the oral cavity often necessitate radiotherapy in addition to surgery. This chapter provides an evidence-based assessment of indications for adjuvant radiotherapy and considerations for treatment.
Adjuvant (or postoperative) radiotherapy (RT) is frequently employed for patients with surgically resected oral cavity cancers who remain at high risk of locoregional failure. The goal of adjuvant treatment is to eradicate residual microscopic foci of cancer cells prior to progression to gross disease. For tumors with certain clinicopathologic features and those for which negative surgical margins cannot be attained without significant morbidity, adjuvant RT with or without concurrent systemic therapy is often indicated to decrease the probability of locoregional recurrence (LRR). This chapter provides a comprehensive overview of the indications for adjuvant RT, current clinical data underlying these recommendations, and essential considerations for delivery of adjuvant oral cavity RT including dose, targeting, techniques, and patient set-up.
35.2 Principles of Adjuvant Radiotherapy
The use of adjuvant external beam RT to improve outcomes for cancers of the head and neck has a long history. MacComb and Fletcher first described the paradigm of definitive surgery followed by RT for locally advanced head and neck cancers in 1957. 1 Fletcher and Evers later reported the results of “prophylactic irradiation” using Co-60 in 1970 in a series of 19 patients with resected head and neck cancers for whom surgical specimens did not show adequate margins. 2 On follow-up, no clinical evidence of disease was seen in 42% of patients who received adjuvant RT, versus only 11% of 147 patients who underwent salvage RT for a gross postoperative recurrence. The authors presciently concluded that, “One should not wait for a recurrence to appear in patients who are at high risk of failure after the surgical procedure. Postoperative radiotherapy should be given as soon as there is wound healing.”
This fundamental tenet of adjuvant head and neck RT remained essentially unchanged over the subsequent half-century. Progress continues to be made, however, in our understanding of specific pathologic features of oral cavity squamous cell carcinomas (SCCs) that portend a high risk of recurrence, anatomic regions that warrant inclusion within the radiation field, appropriate doses for adjuvant RT, and radiosensitization with concurrent systemic therapies. Treatment of the postoperative oral cavity also presents special considerations in reproducible radiation set-up, localizing sites of potential residual disease, and minimizing risk of surgical complications and radiation toxicities in order to achieve optimal patient outcomes.
35.3 Diagnosis and Evaluation
The same principles of careful historical and physical examination detailed in Chapter 33 for definitive RT to oral cavity cancers apply in the adjuvant setting. For patients with high-risk features who are expected to undergo adjuvant RT with or without chemotherapy, the evaluation should ideally be performed by a multidisciplinary team prior to any surgical intervention. This is also an appropriate setting when nonsurgical management of an oral cavity cancer is being considered due to unresectable disease or for patients who are poor surgical candidates or refuse surgery.
A thorough postoperative examination including inspection and palpation of the oral cavity and neck, delineation of surgical scars and boundaries, assessment of the health of any reconstructed tissue, and dental evaluation should also be performed before initiating adjuvant RT. Care should be taken to identify and address any persistent facial edema or postoperative fluid collections so as to avoid anatomic changes during RT that may necessitate replanning. As with definitive RT, it is preferable for any necessary tooth extractions or invasive dental work to be performed prior to irradiation given the risk of osteoradionecrosis, provided that these procedures do not disrupt the surgical site or cause significant treatment delay.
35.4 Indications for Adjuvant Radiotherapy
Criteria for adjuvant RT for resected oral cavity cancers have evolved over time. Currently, the 2018 update of the National Comprehensive Cancer Network (NCCN) Clinical Practice Guidelines recommends adjuvant RT for oral cavity cancers with positive surgical margins, extranodal extension (ENE) of lymph node metastases, pT3 or pT4 primary tumor, N2 or N3 nodal disease, nodal involvement of neck levels IV and V, perineural invasion (PNI), or lymphovascular invasion (LVI). 3 For positive margins not amenable to re-resection, the NCCN provides a category 1 recommendation (reflecting uniform consensus based on high-level evidence) for the addition of concurrent systemic therapy to adjuvant RT in pT3-4 tumors and advises consideration of concurrent systemic therapy in pT1-2 tumors. Adjuvant chemoradiotherapy is a category 1 recommendation for any oral cavity cancer with ENE irrespective of T stage or margin status. The prognostic value of histopathologic grade and tumor human papillomavirus (HPV) status for postoperative recurrence in oral cavity cancers is less well established. 4 – 6 The above pathologic characteristics are frequently subtle. Whenever possible, it is advisable that surgical specimens be reviewed by a qualified pathologist with fellowship training in head and neck pathology or equivalent, given the implications for adjuvant treatment recommendations and prognosis.
The NCCN guidelines for adjuvant treatment of oral cavity cancers primarily reflect the best available data for LRR risk, with the clinical rationale that surgery alone is insufficient when this risk is high. There unfortunately remains a paucity of clinical trials directly comparing surgery to surgery and adjuvant RT for oral cavity cancers. Kokal and colleagues conducted an early prospective randomized trial of these modalities in stages III-IV SCCs of the oral cavity, larynx, and pharynx in the 1980s and found no benefit to locoregional control or overall survival (OS) from the addition of RT. 7 This study included only 15 oral cavity patients and utilized a relatively low RT dose of 50 Gy, however, limiting applicability to current practice. A later prospective study of 140 Indian patients with resected stages III-IV buccal mucosa SCC randomized to adjuvant RT (at doses of 58-65 Gy) or observation showed significantly improved disease-free survival (DFS) in the RT arm (68 versus 38%) as well as a trend toward improved OS (94 versus 84%). 8
In contrast to these limited randomized trial data, nonrandomized analyses have strongly favored the use of adjuvant RT. In one retrospective matched-pair analysis by Lundahl et al of 95 patients with nodal metastases from head and neck SCC (29% arising in the oral cavity), those who underwent neck dissection alone had significantly increased relative risks of nodal recurrence in the dissected and undissected neck (relative risk [RR], 5.82 and 4.72, respectively), decreased cancer-specific survival (RR, 2.21), and decreased OS (RR, 1.67) compared to patients who received adjuvant RT. 9 The addition of radiation following surgery has therefore remained a standard of care practice for high-risk cancers of the oral cavity. In the following sections, the specific data underlying each clinicopathologic feature identified by the NCCN guidelines are discussed in detail. The presence of any of these findings differentiates oral cavity SCCs requiring adjuvant therapy from those which may be observed.
35.4.1 Positive Surgical Margins
The inability to achieve clear surgical margins after resection of SCCs of the oral cavity is a common indication for adjuvant RT. The current NCCN guidelines define a positive surgical margin as the presence of invasive carcinoma or in situ carcinoma at the specimen edge, and a close surgical margin as distance less than 5 mm from the invasive tumor front to the specimen edge. In situ carcinoma is generally not an indication for adjuvant RT, although it may be an indication for resection of additional margins. Multiple studies have shown an increased risk of local recurrence in patients for whom close or positive margins were observed. 10 – 13 It is important to note that these definitions have varied significantly across different clinical trials and publications. Unsurprisingly, advanced T stage is a significant predictor of positive margin status in the oral cavity, 12 , 14 as are high and intermediate histopathologic grades. 15 Among oral cavity subsites, retrospective analysis of greater than 20,000 patients in the National Cancer Database (NCDB) showed increased rates of positive margins for SCCs arising in the retromolar trigone (odds ratio 2.40) and buccal mucosa (odds ratio 2.06), as compared to the reference subsite of oral tongue. 15 A recent NCDB analysis of greater than 120,000 patients with oral cavity cancer demonstrated an overall positive margin rate of 12.75% across all subsites, with minimal change in rates over the study period of 1998 to 2012. 14
It is well recognized that complete excision of gross and microscopic tumor in oral cavity cancers leads to optimal oncologic outcomes. The seminal work of Loree and Strong demonstrated a twofold increase in risk of local recurrence (36 versus 18%) in patients with positive or close margins compared to those with clear margins. 12 Further, the presence of positive surgical margins has been demonstrated to decrease overall and disease-specific survival across multiple subsites of the oral cavity. 13 , 16 , 17 The combination of adjuvant radiation and systemic therapy is often indicated in head and neck tumors with positive margins given the results of the Radiation Therapy Oncology Group (RTOG) 9501 and European Organisation for Research and Treatment of Cancer (EORTC) 22931 randomized phase III clinical trials, 18 , 19 in which approximately one-quarter of enrolled patients had oral cavity cancers. The addition of concurrent cisplatin chemotherapy to adjuvant RT significantly improved rates of LRR (in both trials), DFS (in both trials), and OS (in EORTC 22931 only) on initial analysis as compared to adjuvant RT alone. 10-year follow-up showed continued benefit of chemoradiotherapy for LRR (21 versus 33.1%) and DFS (18.4 versus 12.3%), but not OS, in the subset of tumors with positive margins or ENE. 20 It is important to note that the definition of positive margins differed slightly in each trial: tumor within 5 mm of the specimen edge for EORTC 22931 and microscopically involved resection margins for RTOG 9501.
Several recent studies have attempted to scrutinize the 5 mm cutoff as designating “close margins” in predicting LRR of resected oral cavity cancers. A retrospective analysis of 382 patients at Memorial Sloan Kettering Cancer Center (MSKCC) with oral tongue SCC demonstrated that margin cutoffs of > 2.2 mm (invasive carcinoma to inked edge) were not predictive of increased LRR. 21 Similarly, Tasche et al retrospectively stratified 432 patients with oral cavity SCC by margin distance and found no significant increase in LRR for any distance > 1 mm. 22 Earlier, Dutch and Australian studies had also found no difference in LRR or disease-specific survival for patients with oral cavity SCCs with close margins (< 3 or < 5 mm, respectively) who underwent observation, compared to those with negative margins. 23 , 24 Together, these data provide a provocative argument for observation in resected oral cavity cancers with close margins but no other risk factors for LRR. However, it should be noted that the studies were not designed or powered to specifically test the impact of RT in these patients. Further prospective analysis will be necessary to fully elucidate the long-term oncologic outcomes of this approach.
35.4.2 Extranodal Extension
Metastasis to cervical lymph nodes is seen in approximately half of all oral cavity SCCs. 25 While the size, number, and nodal station of lymph nodes all have prognostic significance in oral cavity cancers, the presence or absence of ENE has perhaps the greatest impact on recurrence risk and treatment outcomes. ENE (alternately extracapsular extension, extracapsular spread, extranodal spread) refers to perforation of the lymph node capsule and invasion of perinodal tissue by tumor. This comprises microscopic extension detected on pathologic examination of resected lymph nodes as well as macroscopic extension which may be suspected on clinical examination or preoperative radiology studies. The American Joint Committee on Cancer (AJCC) now includes ENE as a component of N stage for the oral cavity in the 8th Edition staging guidelines, a change from the previous edition. Clinically, overt ENE is designated cN3b while ENE in surgical specimens is designated pN2a for a single positive lymph node of = 3 cm in dimension and pN3b for all other cases. 26
Clinically, the poor prognosis of ENE-positive oral cavity SCCs is a consequence of increased rates of both LRR and distant metastases. Analysis of a UK cohort of 400 patients who underwent neck dissection for oral cavity SCC demonstrated ENE incidence of 25%, and these patients showed a two- to threefold increase in risk of local, regional, and distant failure relative to node-negative patients or node-positive patients without ENE. 27 Five-year OS was 23% for patients with ENE versus 61.6% for patients without ENE, with a similar disparity in cancer-specific survival. A recent MSKCC retrospective study also reported a significant difference in rates of 5-year cancer-specific survival (25% for patients with ENE versus 57.8% for patients without ENE). 28 Rates of ENE did not appear to vary significantly by oral cavity subsite in either study. There is again a lack of clinical trials directly comparing surgery alone to surgery with adjuvant RT in the setting of ENE. Among the most compelling data are those from a series of 441 patients with head and neck cancer treated at the Medical College of Virginia by different surgical groups with variable referral patterns for adjuvant RT. 29 Three-year local control for patients with ENE who underwent surgery alone was significantly inferior to those who received adjuvant RT, at 31 versus 66%. In patients with both ENE and positive surgical margins, the difference in 3-year local control with and without adjuvant RT was even more striking, at 68 versus 0%.
There exists an ongoing debate regarding the relative prognostic implication of microscopic versus macroscopic ENE in oral cavity SCCs. While macroscopic ENE is almost universally acknowledged to be an adverse clinicopathologic finding, there is no such consensus for microscopic extension. As detailed above, the NCCN considers the presence of ENE (whether macroscopic or microscopic) to be a category 1 indication for adjuvant RT with concurrent systemic therapy regardless of other favorable tumor characteristics, given the short- and long-term outcomes of the RTOG 9501 and EORTC 22931 trials. 18 – 20 A 2002 analysis showed a significant long-term survival decrement associated with both microscopic and macroscopic ENE in a cohort of 173 patients who underwent resection and neck dissection for oral cavity (83%) or oropharyngeal (17%) SCC, as compared to patients with no ENE. 30 Similarly, Greenberg and colleagues identified equally poor long-term survival outcomes in patients with oral tongue cancer who had = 2 or > 2 mm of ENE. 31
Conversely, more recent studies in varying patient populations have demonstrated similar DFS outcomes in patients with microscopic ENE compared to patients with no ENE, 28 , 32 calling into question the necessity of adjuvant chemoradiation in this setting. Broader interpretation of these data is difficult: the MSKCC analysis compared no ENE to any ENE < 1.7 mm based on receiver-operator curve analysis, 28 while Prabhu et al compared no ENE to any degree of “microscopic” ENE without overt obliteration of nodal tissue architecture. 32 Given this uncertainty and the current absence of prospective evidence, it has remained our practice to recommend adjuvant RT with concurrent chemotherapy to all patients with resected oral cavity SCCs where operative pathology shows any degree of ENE, as long as they have no contraindications and are expected to tolerate treatment.
35.4.3 Pathologic T Stage
Clinical and pathologic evaluation of the T stage of oral cavity cancers is an essential consideration in selecting appropriate treatment paradigms. The primary extent of tumors may not be fully apparent prior to resection, necessitating pathologic upstaging and adjuvant therapy. Currently, the AJCC 8th Edition staging criteria for pT3 and pT4 tumors of the oral cavity are greatest tumor dimension of > 4 cm or depth of invasion (DOI) of > 10 mm (pT3), or locally advanced disease (pT4). 26 The pT4 classification is divided into pT4a and pT4b classifications based on extent of invasion, with the latter generally being considered unresectable. pT4a comprises tumors invading through the cortical bone, the skin of the face, or into the maxillary sinus; or, for lip SCCs, invasion of the inferior alveolar nerve or floor of mouth. pT4b represents very advanced disease with invasion into the masticator space, pterygoid plates, or skull base; or tumors encasing the internal carotid artery. DOI is measured from the basement membrane to the deepest point of invasive tumor and should not be confused with tumor thickness, which measures from the tumor surface to the deepest point of invasion. Large, exophytic lesions may therefore still have a shallow DOI. Though pT3-4 status increases the risk of additional adverse pathologic findings such as inability to obtain clear surgical margins, as described previously, advanced T stage is also an independent negative prognostic factor in SCCs of the oral cavity.
Increasing primary tumor size and involvement of nearby critical structures often reflect delayed diagnosis or aggressive tumor biology and are associated with poorer outcomes in essentially all malignancies and anatomic sites. The significance of DOI, however, has been increasingly appreciated for tumors arising in the oral cavity. The AJCC 8th Edition head and neck staging guidelines are the first iteration of the manual to incorporate DOI as a modifier of the T stage in oral cavity cancers. 33 This important change reflects the extensive body of work that has emerged over the last several decades demonstrating that increasing pathologic DOI within the oral cavity is associated with increased rates of locoregional failure, 34 , 35 increased rates of cervical lymph node metastases, 36 – 38 and worsened survival outcomes. 39 – 42 With this revision to the prior T stage criteria, any oral cavity tumor with DOI > 10 mm is an indication for adjuvant RT by current NCCN clinical practice guidelines, independent of its maximal dimension. Indeed, many institutions, including ours, had already incorporated DOI into clinical practice as a potential indication for adjuvant RT, and several studies suggest benefit of RT for patients with this risk factor. 38 , 43