Skyler B. Johnson and Roy H. Decker
New and emerging data have recently challenged the belief that malignant melanoma is a radioresistant tumor (1). Early experiments of melanoma in cell culture models demonstrated survival curves with a particularly broad shoulder (2,3) and robust DNA repair capability (4), which suggested insensitivity to radiation damage and to conventional dose fractionation. But clinical data only sometimes support a better response when radiation is delivered with higher dose per fraction, when compared with more conventionally fractionated schedules (5–8). Moderate hypofractionation (5–10 fractions) has been used adjuvantly after resection or node dissection, with control rates in excess of 50% (9–12), and the use of stereotactic body radiation therapy (SBRT) and stereotactic radiosurgery (SRS) for melanoma metastases to the body or brain is associated with excellent local control. But an effort to demonstrate superior outcomes with high doses per fraction in a randomized study was not successful. The Radiation Therapy Oncology Group (RTOG) 8305 trial included 137 patients with melanoma randomized to either 32 Gray (Gy) in 4 fractions or 50 Gy in 20 fractions and showed no difference in response rate (13).
What has become evident is that melanoma responsiveness to radiation is much more complex and heterogeneous—there is significant variability in melanoma responsiveness in cell culture models, for example, with several cell lines exhibiting a narrower cell survival shoulder similar to nonmelanoma malignancies (14–16) and a better response to standard fractionation (17,18). Overall, the laboratory and clinical experience supports the possibility that there is a wide range in radiation response in melanoma, dependent on the individual cell line or patient, as well as the fraction size, total dose, and tumor volume (8,19). Thus, the optimal dose and fractionation schedule for melanoma remains an area of active investigation.
ROLE OF RADIATION IN MELANOMA
There is a limited role for the use of radiotherapy for localized melanomas. Patients with localized melanoma are most commonly and appropriately treated with wide excision; radiotherapy is given adjuvantly in limited circumstances for patients with certain high-risk features. However, definitive radiotherapy can be considered for early cutaneous melanoma in patients who are inoperable due to comorbid medical conditions, or as local palliation in patients with advanced disease where potentially curative therapy is not possible. Definitive radiation is also appropriate in select patients with noncutaneous melanoma—for example, brachytherapy or external beam radiation in medium-sized ocular melanomas, or primary or adjuvant radiation for mucosal or visceral melanomas. An example of the results of treatment of melanoma with definitive radiation is summarized in Table 10.1.
Approximately 91% of all melanomas diagnosed are cutaneous (23). Local excision with wide margins is the standard of care for cutaneous melanoma, and local recurrences are generally low even for those with high-risk features (thickness >4 mm, ulceration, satellitosis, and head and neck site) (24–26). Definitive radiotherapy has been shown to be an effective treatment for select presentations of cutaneous melanomas (27), and in this context is relatively underutilized (28).
Most reported retrospective data using radiotherapy as the primary treatment modality in cutaneous melanoma occur in the setting of lentigo maligna because it is often found in elderly patients with large lesions, which often include the face, making radiotherapy a natural nonoperative choice to improve cosmetic outcomes. In these studies, local recurrences vary greatly (29–35). However, pooled data suggest local recurrences between 5% and 9% (20,36,37), which is in the same range as the recurrence rate expected following surgical excision (27).
Radiation therapy technique, dose, and fractionation for cutaneous melanomas vary greatly by histologic variant, location, and size, and likely account for the vast differences in recurrence rates seen institution to institution. In general, patients should be treated with superficial radiotherapy fields, which include the visible lesion plus a 1 to 2 cm margin. Commonly used doses include the following: 30 to 35 Gy in 5 fractions, 42 Gy in 7 fractions, or 50 to 60 Gy in 2 to 4 Gy fractions, depending on the location and size of the lesion.
Mucosal melanomas are rare, making up approximately 1% of all melanomas. Of these, the majority occur in the head and neck region (55%), followed by the gastrointestinal (24%), vulvovaginal (18%), and urinary tract (3%) (23).
Head and neck
Melanomas of the head and neck typically occur in the oral and nasal cavity, as well as the paranasal sinuses, and are treated surgically with or without postoperative radiation. The effectiveness of definitive radiation therapy in the management of head and neck melanomas is unclear. Limited series, which report on the use of definitive radiation therapy in patients not eligible for surgery, have reported local control rates as high as 80% (38–42) and excellent overall survival. In general, multimodal treatment with surgery followed by postoperative radiation likely provides the best current approach to improve local control (43–45).
Similar to mucosal melanomas of the head and neck, the mainstay of treatment for anorectal melanomas is primarily surgical with either abdominoperineal resection or wide local excision. There is little evidence to evaluate the efficacy of radiation therapy as a definitive, primary local therapy. Case reports of 12 patients treated with radiation alone had a range of survival times between 4 and 48 months (39,46–49).
Melanoma can arise in other areas of the gastrointestinal tract including the esophagus or small bowel. These are extremely rare. For example, less than 0.3% of all esophageal cancers are primary melanoma (50). Primary melanomas of the small bowel are rare; the majority of reported melanoma involving the small bowel are most commonly associated with metastasis and second only to the lung in frequency of metastatic spread (51). These have an extremely poor prognosis, and given their rarity evidence of the efficacy of radiation is extremely limited. Small series suggest that radiation may provide durable control. Of 10 patients treated with radiation alone, in a series evaluating different treatment modalities for primary esophageal melanoma, the mean survival was found to be approximately 15 months. This was better than surgery alone but slightly worse than surgery followed by radiation, which was approximately 17 months. The latter is currently the preferred approach (52).
The second most common malignancy of the vulva are vulvar melanomas, which are located in a distribution that includes the mucosa of either labia minora, labia majora, or the clitoris. Vaginal melanomas are much rarer than vulvar melanomas and are primarily located in the lower vagina. Both sites have poor 5-year survival with typical rates below 50%, despite the majority of patients (70%) who present with localized disease (53,54).
Radiation for unresectable vaginal melanomas has been recommended as it has been shown to be an effective alternative to surgery, which can achieve acceptable local control (55,56). In a series of 35 women treated at Memorial Sloan Kettering, radiotherapy was the primary therapy in 11 (35%). Of these, six of the patients had locally advanced disease that was not surgically resectable and the remaining five had early stage disease. When comparing median survival, primary surgery was associated with longer survival than radiation therapy (25 months vs. 13 months) (57). The difference in survival remained even when excluding those with locally advanced disease. A study using the Surveillance, Epidemiology, and End Results (SEER) database demonstrated that of 644 vulvar melanoma patients, only 33 (5.1%) were treated with definitive radiotherapy (58). In this cohort, the survival of those with vulvar melanoma treated with surgery or radiation was similar (58), although many patient and treatment details were not available.
Urethral melanomas often result in delayed diagnosis and treatment because they are extremely rare, making up less than 4% of urethral carcinomas and 1% of melanomas. In part because of this, they convey a poor prognosis with approximately 50% of all cases presenting with nodal or distant metastases at the time of diagnosis (59,60). Current treatment standards generally include surgery with either wide tumor excision or urethrectomy followed by adjuvant therapy, which includes radiotherapy, chemotherapy, or immunotherapy. There is currently no available evidence in the literature which supports definitive radiotherapy, though palliative treatment may be appropriate in patients who cannot undergo resection.
Early in the history of the treatment of ocular melanomas, enucleation was the standard of care. Later, in an effort to decrease morbidity and preserve the eye and possibly visual function, the Collaborative Ocular Melanoma Study (COMS) trials established I-125 plaque brachytherapy as a reasonable alternative to enucleation for medium-sized tumors (2.5–10.0 mm height and ≤16 mm diameter). One thousand three hundred and seventeen patients with medium-sized choroidal melanoma were randomized to enucleation or I-125 brachytherapy. After 5 and 12 years, there was no significant survival difference between groups (5 years: enucleation 81% vs. I-125 82% and 12 years: enucleation 59% vs. I-125 57%) (61,62). Follow-up studies demonstrated that there was no increase in risk of second cancers in those receiving brachytherapy, and those receiving this treatment modality had better visual function at 2 years (63,64). The known complications of plaque radiotherapy include the development of cataracts, retinopathy, optic neuropathy, macular edema, and pain (22,65,66).
Typically, plaques should include a 2-mm margin around the largest basal diameter and I-125 dose of 85 Gy to at least 5 mm or to the apex of the tumor using a dose rate of 0.5 to 1.25 Gy/hour. However, there is some retrospective data, which demonstrated that lower doses may have similar rates of survival and disease control (67), and presumably a lower risk of long-term side effects. Therefore, for tumors close to critical structures (macula, lens, optic disc, and sclera), a dose reduction may be considered.
For large ocular melanomas (>10.0 mm height and >16 mm diameter), COMS effectively showed that neoadjuvant external-beam radiotherapy did not provide any additional benefit when compared with enucleation alone (68).
Proton beam radiotherapy may also be considered for the treatment of ocular melanoma and has shown to provide excellent local control rates and ocular preservation, similar to that seen with brachytherapy (69,70). Interestingly, a recent meta-analysis demonstrated that protons may result in decreased odds of local recurrence when compared with brachytherapy (71), but further investigation is likely needed to compare treatment methods.
Following adequate wide excision, local recurrence rates are generally low. However, local recurrence rates can be high in patients with adverse pathologic features; for these patients, adjuvant radiation therapy should be considered. Important pathologic features that confer increased risk of recurrence are listed in Table 10.2. Adjuvant radiation therapy has been effectively used in patients following an initial surgery where there were positive or close margins, and in those with adverse pathologic features.
Recurrence Rates (%)
Tumors >4 mm (25)
Desmoplastic neurotropic subtype (72)
Head and neck (24)
As stated previously, local recurrence rates following adequate wide local excision are low, <10%. However, certain pathologic features or combinations of features result in increased recurrence rates; adjuvant radiotherapy should be considered to improve local control. In cutaneous melanoma, the key adverse pathologic characteristics that confer higher risk of local recurrence are desmoplastic subtype, extensive neurotropism, and locally recurrent disease. Desmoplastic subtype was originally felt to be associated with local failure rates of up to 50% (74). A recent large series revealed that the rate may be as low as 11%, although this was still much higher than that seen for all forms of cutaneous melanoma, which was approximately 3% (75). Interestingly, in this same series the local failure rates were significantly higher in the presence of neurotropism and surgical margins less than 1 cm (75). Retrospective evidence comparing adjuvant radiation versus surgery alone for desmoplastic lesions demonstrated that adjuvant radiation was an independent predictor of improved local control and improved local recurrence rates from 24% to 7% after a median follow-up of more than 6 years (72). The majority of these patients were treated with opposed electrons, photons, or both, with a 3- to 4-cm margin around the surgical site to 30 Gy in five fractions delivered biweekly.
Prospective data have demonstrated that patients with cutaneous melanoma at high risk for local recurrence benefit from the addition of adjuvant radiation following surgery. One hundred and seventy-four patients received adjuvant radiation therapy to a dose of 24 to 30 Gy in four to five fractions if they had a tumor, which was 1.5 mm thick or greater or Clark level of four or greater, or high risk of nodal recurrence. They showed a 5-year actuarial local recurrence rate of 88%, which was much better than historical controls of approximately 50% (76). These promising results resulted in a phase III RTOG/Eastern Cooperative Oncology Group (ECOG) trial, which was unfortunately closed due to poor accrual.
Despite what appears to be considerable local control benefit with adjuvant radiation, rates of its use are low, as suggested by a recent SEER analysis that reported only approximately 7% of patients with desmoplastic subtypes receive adjuvant radiation (77). Fortunately, a phase II trial is ongoing and is comparing the efficacy of adjuvant radiation in patients with desmoplastic melanoma (NCT00060333). Additionally, a phase III trial comparing surgery alone versus adjuvant radiation therapy for cutaneous melanomas with neurotropism of the head and neck is ongoing (NCT00975520). These trials will provide high-quality evidence for further evaluation of the appropriate treatment modalities for patients with cutaneous melanoma, particularly those with desmoplastic and neurotropic features.
Head and neck
Mucosal head and neck tumors have higher rates of local recurrences following surgery than melanomas of other locations (78) and adjuvant radiotherapy has been shown to improve local control, but has not shown to have a significant impact on overall survival (45,79–81). This has been replicated in many studies and suggests that these patients have high rates of distant failure and improvement in systemic therapies is needed. Retrospective data have shown that when comparing adjuvant radiation to surgery alone, local control rates were improved (55%–62% vs. 26%–30%), even in the presence of more advanced disease in the adjuvant radiation group (79,81). The ideal dose and fractionation schedule is currently unknown. However, there are data to suggest that a dose of at least 54 Gy over 5 to 6 weeks is required (80).
Current trends in the treatment of gastrointestinal melanoma, depending on location, favor conservative wide excision. Similar to other mucosal sites, adjuvant radiation following surgery improves local control but does not appear to improve overall survival. Retrospective data have demonstrated that in addition to negative margin sphincter-sparing surgery, adjuvant radiation in 5 Gy fractions results in effective local control of 82% at 5 years. However, similar to other mucosal melanoma disease sites, there continued to be extremely high rates of metastasis and death (82). Additionally, it was demonstrated that extended field radiotherapy, which included the inguinal lymph nodes, resulted in higher rates of toxicity and did not improve outcome (82).
Adjuvant radiotherapy for patients with vulvovaginal melanomas may become more common as there has been a recent trend toward less aggressive surgical resection as primary management. Currently, the role of adjuvant radiotherapy is limited to palliation of those with locally advanced disease (60). For patients requiring groin and pelvic radiotherapy, photons to a total dose of 45 to 55 Gy in 1.8 to 2.0 Gy fractions are recommended and will be covered in more detail later. In general, for melanomas of the vulva, radiation can be delivered to the primary site using electrons to a dose of 45 to 50 Gy in 1.8 to 2.0 Gy fractions. Higher dose per fraction regimens are not generally recommended in an attempt to limit toxicity to the sensitive skin of the vulva and the surrounding mucous membrane. In a series of women with vaginal melanoma, it was demonstrated that in those women who had surgery alone, there were 100% recurrence rates, 45% of which were local recurrences, suggesting that there may be a role for radiation. However, when evaluating women who underwent surgery followed by radiation, although local recurrences only occurred in 7% of patients, overall recurrence rates were still 94% because of the high rates of distant recurrence. Interestingly, there was a 13-month improvement in overall survival for those who received adjuvant radiation compared with surgery alone. However, this difference was not statistically significant, as it was limited by small sample size. The authors concluded that wide excision and adjuvant radiation improved local control and is a reasonable treatment option, given the overall poor prognosis (83).
As stated earlier, the current treatment for urethral melanoma includes surgery with either wide tumor excision or urethrectomy followed by adjuvant therapy, which includes radiotherapy, chemotherapy, or immunotherapy. A recent systematic review of the literature demonstrated that there were recurrences in over 70% of reported cases, the majority of which were locoregional recurrences with 55% occurring at the primary site and 28% in inguinal lymph nodes. Interestingly, of the 118 cases with treatment info, only 43 (36%) underwent any adjuvant therapy, 24 (56%) of whom underwent radiotherapy (59). Evidence that demonstrates the benefit of adjuvant radiation therapy is limited, but given the high rates of local and locoregional recurrences, consideration of radiation as an adjuvant to surgery to improve local control should be made.
Role of radiation in locally advanced melanoma
Radiation in the setting of locally advanced disease following surgery improves local control rates in patients with high risk of nodal failure (9–12,84–90). Patient selection is important and includes location, size, number of involved nodes, and/or extranodal extension (Table 10.3). A phase III trial of 250 patients with multiple positive or large nodes located in the head and neck, axilla, and groin were randomized to observation versus radiotherapy to 48 Gy in 20 fractions and showed that those who received radiation had significantly improved local control at 5 years when compared with observation (82% vs. 67%). There was no difference in overall survival (91,92). High-risk patients with any of the risk features listed in Table 10.3 or those with clinically positive nodes who have had wide excision and complete therapeutic lymph node dissection should be considered for radiotherapy to the nodal basin. In summary, there is high-quality evidence that adjuvant radiotherapy can prevent nodal recurrence (Table 10.4), but without an improvement in overall survival. The lack of a survival benefit suggests that the same risk factors that predict local recurrence also predict an increased risk of occult distant metastatic disease. The recommendation for adjuvant radiation should therefore balance the decrease in recurrence risk with the incidence of late radiotherapy-related effects.
Recurrence Rates (%)
Extranodal tumor extension (93)
Cervix location (93)
Nodal size ≥3 cm (93)
Number of nodes:
≥1 node (94)
≥4 nodes (95)
*Recurrence rates based on the presence of axilla or groin nodes.