7 Role of Radiotherapy in the Treatment of Skin Malignancies
Non-melanomatous skin cancers (NMSC), specifically basal cell carcinoma (BCC) and squamous cell carcinoma (SCC), are the most common malignancies in the United States. They are primarily managed surgically and carry an excellent prognosis, with a 1 to 5% rate of disease recurrence after complete excision and only exceedingly rare instances of distant metastasis. Historically, radiotherapy played a prominent role in the definitive management of these cancers as an alternative to surgical resection, especially in cosmetically sensitive areas. With the improvement in surgical techniques in recent decades, especially the widespread use of Mohs micrographic surgery (MMS), the use of curative radiation for skin cancers has declined. However, radiation is still commonly used in the definitive setting for patients with NMSC who are poor surgical candidates or have larger lesions in cosmetically sensitive regions of the face, and in the postoperative setting for more advanced tumors that have high-risk pathologic features that are associated with excessive risks of locoregional recurrence after surgery alone. Radiotherapy plays a similar role in the postoperative management of cutaneous melanoma. Radiation also offers excellent palliation for noncurable patients. This chapter reviews common indications, dose and fractionation schedules, techniques, and oncologic and cosmetic outcomes for radiotherapy for NMSC and melanoma ( Fig. 7.1).
7.2 Indications for Radiotherapy for NMSC
7.2.1 Indications for Definitive Radiotherapy
Both surgery and RT can offer excellent cure rates for the management of early-stage NMSC. Surgery is typically the preferred method of treatment, given it can be performed in a single session, and has been associated with superior oncologic and cosmetic outcomes. One prospective randomized study compared the outcomes of MMS and definitive RT in 347 patients with less than 4 cm BCC of the face. The local failure rate was 0.7% for patients treated with MMS and 7.5% for patients treated with RT. Additionally, the cosmetic outcome was rated “good” or better more often in those patients who underwent surgery (87 vs. 69%). 1 Though the technique of RT was not controlled (55% received interstitial brachytherapy and 45% received orthovoltage therapy), which could have affected the quality of the comparison, as the only randomized study, this trial remains pivotal in guiding medical decision making.
A variety of factors can influence the decision to favor RT rather than surgery for NMSC. Typical contraindications to definitive RT include very large tumors and those with bone and/or cartilage invasion, given that control rates are typically inferior, as well as recurrent tumors that have been previously radiated or arise in an area of previous radiotherapy. Radiation should also be avoided in patients with genetic radiosensitivity syndromes, such as xeroderma pigmentosum and basal cell nevus syndrome, and in those with active connective tissue diseases, such as scleroderma and systemic lupus erythematosus. 2 Patients who are candidates for definitive RT include those with unresectable disease, older patients with comorbidities limiting their surgical options, and patients with lesions involving the eyelid, canthi of the eye (especially those who carry surgical risk to the lacrimal duct), external ear ( Fig. 7.2 and Fig. 7.3), nose ( Fig. 7.4 and Fig. 7.5), brow ( Fig. 7.6), temple ( Fig. 7.7), or lip, which may incur significant cosmetic or functional defects from surgical resection.
A robust literature, much of it older when RT was more commonly used, supports the safety and efficacy of definitive RT for many of these aforementioned types of patients. A review of 986 BCC and SCC of the skin overlying the eyelid treated with definitive RT demonstrated a 5-year cure rate of 96.4%. 3 A review of 334 BCC and SCC of the external ear at the Princess Margaret Hospital treated with definitive RT demonstrated a 2-year local control rate of 87%, with severe late toxicity noted in only 7% of patients. 4 A review of 100 patients with SCC of the nasal skin at the Princess Margaret Hospital treated with definitive RT demonstrated a 2-year local control rate of 90%, with no severe toxicities observed. 5
While definitive RT monotherapy can provide effective tumor control for T1–3N0 NMSC, T4 tumors and those with nodal metastases demonstrate inferior outcomes. Initial surgical management is recommended in these patients, often followed by adjuvant radiotherapy. T4 disease is defined as tumor invasion into the axial or appendicular skeleton or perineural invasion (PNI) of the skull base. The University of Florida reported a local control rate of just 53% at 5 years in patients with T4 BCC and SCC. 6 , 7 In a review of 68 T4 BCC and SCC of the skin of the head and neck treated with definitive RT, local control rates were lower in patients with recurrent disease, compared to primary lesions (41 vs. 67%; p = 0.07) and the three most important prognostic indicators for inferior local control and cause-specific survival rates were bone involvement (p < 0.01), recurrent lesions (p < 0.01), and nerve involvement (p < 0.02). 6 Even the more favorable lesions had local control rates of less than 70%. Similarly, a large retrospective review which included 531 BCC and SCC treated with definitive RT demonstrated local control rates of 94 and 89% for primary BCC and SCC tumors, respectively, and 86 and 68% for BCC and SCC recurrent tumors, respectively. 8 In patients with nodal metastases, locoregional recurrence rates with definitive RT alone range from 30 to 50% and cancer-related mortality is as high as 30%. 9 Such suboptimal outcomes support the use of intensification of treatment and multimodal therapy for these patients with advanced disease. In patients who are not eligible for initial surgical resection, concurrent cisplatin-based chemotherapy can be considered to enhance the effectiveness of radiotherapy. While there is no high-quality evidence demonstrating clear benefit to such an approach in cutaneous malignancies, many head and neck oncologists extrapolate from the mucosal SCC of the head and neck, where concurrent chemotherapy significantly improves disease control and survival in the setting of locally advanced disease treated nonoperatively. 10
Indications for Postoperative Radiotherapy
Postoperative RT is rarely used for BCC. Patients with BCC have an exceedingly low risk of recurrence after surgery alone, and even patients with a positive margin, focal cartilage invasion, or PNI are often still candidates for close observation and salvage re-resection if needed. 11 , 12 However, postoperative RT is often considered in cases where there are persistently positive margins after multiple resections, T4 disease that extensively invades bone or soft tissue, lymph node metastasis, or clinically apparent PNI ( Fig. 7.8). 13
Postoperative RT is much more commonly used in resected SCC ( Fig. 7.9). Postoperative RT is indicated for patients with high-risk resected disease, such as T4 disease, in transit or satellite metastases, nodal metastases, persistently positive margins, and recurrent disease ( Fig. 7.10 and Fig. 7.11). In addition, patients with two or more intermediate-risk factors, including tumor size greater than 2 cm, poorly differentiated histology, ear and hair-bearing lip location, PNI, and immunosuppressed status, may also benefit from adjuvant RT.
Retrospective data have demonstrated high rates of local recurrence in patients with T4 tumors managed with surgery alone. 7 In addition, the rate of occult lymph node metastasis is high, ranging from 29 to 50%, in patients with advanced T-stage disease. In patients with deeply infiltrative (≥ 8 mm) tumors, or tumors that extensively invade deep subcutaneous fat, occult lymph node metastasis can be as high as 30%. 9 Patients with recurrent primary tumors, PNI, lymphovascular space invasion, and those that are immunosuppressed are at significantly higher risk for having lymph node metastases and should undergo careful evaluation for regional disease with a contrast-enhanced diagnostic neck computed tomography (CT) and/or a positron emission tomography scan. 14 , 15 , 16 In these patients, sentinel lymph node biopsy and/or neck dissection is recommended in conjunction with resection of the primary tumor. Postoperative RT can also be useful as an elective treatment to the undissected neck.
For patients with clinically involved lymph nodes, a therapeutic neck dissection followed by postoperative RT is the current standard of care. Although lymph node metastases are rare in SCC of the skin overall (~ 5%), they are clearly associated with a poor prognosis. After neck dissection alone, locoregional recurrence rates are 11 to 38% and even after multimodality therapy, the 5-year disease-free survival rates are 60 to 70%. Independent predictors of worse survival in this population include increased nodal size ≥ 3 cm, multiple lymph node involvement, extracapsular extension, incompletely excised nodes, and surgery monotherapy. 15 , 16 , 17 A review of 167 patients in Australia with SCC metastatic to the parotid or cervical nodes compared outcomes for surgery versus combination of surgery and postoperative RT at a median dose of 60 Gy in 30 fractions. The use of postoperative RT was associated with significantly lower rates of locoregional recurrence (20 vs. 43%), and higher 5-year disease free (73 vs. 54%; p = 0.004), and 5-year overall survival (66 vs. 27%; p = 0.003) compared to patients who received surgery alone. 16 Similar results supporting the benefit of postoperative RT for locoregional control and 5-year disease-free survival in patients with cutaneous SCC metastatic to lymph nodes and/or periparotid lymph nodes have been reproduced in a number of retrospective reviews. 18 , 19 , 20 , 21 For patients who received lymph node dissection for primary SCC located on the trunk or extremities, postoperative RT is typically recommended when multiple nodes are involved or extracapsular extension is present. Similar to mucosal SCC of the head and neck, RT can be avoided in immunocompetent patients with a single involved lymph node, smaller than 3 cm, without extracapsular extension on parotidectomy or cervical lymph node dissection, as rate of regional recurrence is less than 5%. 22
PNI, while not common (5–10% of SCC), is another important risk factor for recurrence that should be weighed in the decision to administer adjuvant RT. PNI is typically divided into clinical and microscopic PNI. Clinical PNI is defined by neurologic manifestations, most commonly involving the trigeminal or facial nerves, leading to pain, paresthesias, paralysis, formication (the sensation of bugs crawling on the skin), or radiographic evidence of nerve enhancement. 23 , 24 Microscopic PNI is appreciated only histologically and is identified after surgery in a patient who was asymptomatic preoperatively. PNI is important given it is associated with increased risks of local recurrence as well as regional and distant metastases. Factors associated with increased risk for SCC with PNI include male sex, tumor size greater than 2 cm, midfacial tumor location, recurrent tumor, and poorly differentiated subtypes. 25 The degree of PNI is predictive of risk of recurrence; therefore, a careful history as well as physical and imaging examination is critical. Magnetic resonance imaging (MRI) has the advantage of identifying the extent of macroscopic disease through nerve enlargement or enhancement or obliteration of the normal fat plane surrounding a nerve ( Fig. 7.12). CT offers the advantage of detecting whether bone erosion is present, particularly tumor invasion into foramina associated with cranial nerves.
The difference between clinical and microscopic PNI was highlighted in a series from the University of Florida which compared the outcomes of patients with PNI treated aggressively with surgery and postoperative RT. Those with clinical PNI had significantly lower 5-year rates of local control (57 vs. 90%; p = < 0.001) and overall survival (57 vs. 69%; p = 0.03) compared to those found to have microscopic PNI. 13 , 23 RT is therefore always recommended in cases of clinical PNI. The role of postoperative RT in patients with pathologic PNI is less clear. Lin et al found that focal versus extensive microscopic PNI carries different prognoses, with relapse-free survival better in the former group (86 vs. 74%; p =0.1). 24 Unfortunately, the distinction between focal and extensive was not well defined. As such, adjuvant RT is not recommended in immunocompetent patients with nonrecurrent disease, in whom one or two isolated areas of PNI are found in small unnamed nerves, with a diameter of less than 0.1 mm, given the outcome is expected to be fairly good with surgery alone. However, patients with microscopic PNI, that is multifocal, involves larger nerves (> 0.1 mm in diameter) and named nerves, or occurs in immunosuppressed patients, are at higher risk and thus adjuvant RT is recommended. 2 , 25
For irradiation of clinical PNI, the clinical target volume should include areas at high risk of failure, specifically the involved nerve, the portion of the nerve proximally at the skull base, the distal skin innervated by the nerve, major communicating branches, and the compartment in which the nerve is embedded. 26 For example, for patients with CN VII involvement, the area of treatment should be tracked back to the nerve′s exit through the stylomastoid foramen, with care taken to administer adequate coverage of the geniculate ganglion by avoiding excessive restriction of the radiation dose delivered to the ipsilateral cochlea. Similarly, with V1/V2 nerve involvement, the gasserian ganglion in Meckel cave and the cavernous sinus should be targeted ( Fig. 7.13). Targeting the nerve root as it exits the brainstem should be considered for nerves that are radiographically involved at the skull base. For patients with unresectable disease invading the skull base, high-dose radiation (70 Gy), with or without concurrent chemotherapy, is required ( Fig. 7.14).
Another consideration highly relevant to radiation field design is the observation that PNI may also be associated with increased nodal failure in addition to recurrence in the tumor bed and along the nerve pathway. Lin et al demonstrated that patients who developed recurrent disease with pathologic PNI had a significantly higher risk of local recurrence (40 vs. 19%; p < 0.01) as well as regional recurrence (29 vs. 5%; p = 0.02). 24 It is important to consider the location of the lesion in such scenarios, given that tumors on the scalp, for example, may be less likely to metastasize to lymph nodes, compared to cheek, ear, or nasal skin lesions. Understanding the drainage patterns of these tumors is crucial (e.g., parotid nodes for head and neck sites, axilla for trunk and extremity lesions) and should be incorporated into radiation targeting.
7.2.2 Indications for Radiotherapy for Cutaneous Melanoma
Indications for Definitive Radiotherapy
Melanoma is believed to be a relatively radioresistant malignancy. Therefore, maximal surgical resection is paramount in the management of these tumors. RT has limited benefit in achieving long-term control in the presence of gross disease. RT is delivered to gross melanoma in patients deemed to have unresectable disease, or those with rapid postoperative recurrence who have failed surgical management. In these cases, higher doses per fraction are generally preferable and the intent of treatment is largely palliative.
Indications for Postoperative Radiotherapy
Adjuvant radiotherapy to the primary tumor bed, nodal basin, or both, has been studied as a means of reducing locoregional recurrence. If adequate margins are obtained, local recurrence for primary melanoma is infrequent, with rates consistently reported to be less than 5%. 27 , 28 However, there is a smaller subset of patients with higher rates of failure in the tumor bed, including deeply invasive T4 disease, the presence of satellitosis, the presence of desmoplastic subtype, or persistently positive margins despite multiple attempts at excision. In these cases, adjuvant RT to the primary tumor bed should be considered ( Fig. 7.15 and Fig. 7.16).
O′Brien and colleagues analyzed 629 patients with head and neck melanoma and found that increased local recurrence rates were associated with increased tumor thickness: < 0.76 mm, 2%; 0.76 to 1.49 mm, 5%; 1.5 to 3.99 mm, 15%; and ≥ 4 mm, 20%. 29 Typically, adjuvant RT is considered for T4 tumors, although typically reserved for those with additional high-risk features. One of which is satellitosis, which represents discontiguous sites of disease within 2 cm of the primary tumor and is known to be a marker of increased recurrence risk. 30 Desmoplastic melanoma is a rare histologic subtype (1% of all melanomas) and is characterized by spindle-shaped cells with associated collagen production. This subtype is often associated with perineural spread and local recurrence rates have been reported as high as 20 to 50%. 31 , 32 In these cases, based on the relatively higher risk for local recurrence, adjuvant RT should be considered.
More extensive literature guides the use of adjuvant RT to nodal basins, both in the elective setting and following lymph node dissection. 33 , 34 Data supporting the benefit of RT in reducing regional recurrence come from the randomized phase III study by Burmeister et al in which higher risk patients were randomized to surgery and lymph node dissection with or without adjuvant radiation to the nodal basin. 35 Eligible patients had one of the following features associated with increased failure rates:
Serum lactate dehydrogenase < 1.5 times the upper limit of normal and the presence of extracapsular extension (ECE).
≥ 1 involved parotid node of any size.
≥ 2 involved cervical nodes and/or ≥ 3 cm of tumor within a node.
≥ 2 involved axillary nodes and/or ≥ 4 cm of tumor within a node.
Importantly, no systemic therapy was delivered in this trial. In the 217 patients included on the study, those who received RT were significantly less likely to experience regional recurrence compared to those in the observation arm (18 vs. 33%; p = 0.041). As distant metastasis is the predominant mode of melanoma disease failure, it was not surprising that an improved rate of regional control did not translate into an overall survival average (54 vs. 44%; p = 0.12). The most common Grade 3 or 4 adverse events were seroma and wound infections, which were equivalent in both arms. 35 The lack of systemic therapy in this study as well as the lack of a survival difference with the use of RT has raised the question of the true benefit of adjuvant RT for all patients eligible for this study. That said, it reinforces the guiding principle where locoregional control is a priority, and in a case with extensive nodal disease, especially with extranodal extension and significant soft-tissue involvement, adjuvant RT should be considered as a proven way to reduce rates of locoregional recurrence ( Fig. 7.17, Fig. 7.18, and Fig. 7.19).