Cutaneous melanoma is a disease that often has an aggressive and unpredictable course. It was historically thought to be a radioresistant neoplasm; however, substantial radiobiologic and clinical evidence has emerged to refute this notion. Improved local control has been demonstrated with the use of adjuvant radiation therapy delivered to the primary site or regional lymphatics in patients with high-risk clinical or pathologic features. Despite improved local control, high-risk cutaneous melanoma often spreads systemically, leading to poor survival. In the setting of systemic progression, radiation therapy can frequently palliate symptomatic sites of metastatic disease.
An estimated 62,480 new cases of cutaneous melanoma were diagnosed in the United States in 2008, approximately 4% of all newly diagnosed cancers. Cutaneous melanoma was also responsible for an estimated 8420 deaths. While incidence is rising, mortality has decreased with a 5-year survival rate of 92% from 1996 to 2003 compared with 82% from 1975 to 1977. This pattern is felt to be due to increased awareness and screening programs leading to earlier detection. Solar ultraviolet (UV) radiation has been linked to the development of cutaneous melanoma, which results from the malignant transformation of epidermal melanocytes. A higher incidence of melanoma has been found in areas of the body with high sun exposure, among people with other sun-related skin conditions, within populations dwelling in areas of high ambient sunlight, and among populations with increased sun sensitivity. The presence of multiple nevi is also an accepted risk factor.
Cutaneous melanoma remains a surgically treated disease for the majority of patients. Most cutaneous melanomas clinically present as visible, pigmented lesions. Approximately 80% of patients present with localized disease, and most are cured by simple excision. However, as lesions grow in size, the risk of locoregional recurrence and distant spread increases dramatically. Lesions may recur at the primary site, as in-transit lesions, or in the neighboring lymphatics. Distant sites such as the lung, liver, brain, and bone are also at risk. Although surgery retains a role even in patients with advanced disease, adjuvant therapy becomes increasingly important in these patients as well as those at high risk for locoregional or distant spread. Multiple strategies using systemic modalities including cytotoxic agents and immunomodulators have been evaluated, with limited success.
Radiotherapy represents an important yet underused modality in the treatment of cutaneous melanoma. The incorporation of radiotherapy into treatment algorithms for patients has been reluctant and controversial. The first use of radiotherapy in the treatment of melanoma may have been by Simpson in 1913, when he successfully treated a large black nevus with radiotherapy. Although there was no histologic confirmation of melanoma, he was able to remove the lesion with radiotherapy without damage to the surrounding normal skin. Throughout the early 1900s, cutaneous melanoma was labeled as a categorically radioresistant tumor despite a paucity of data to support this claim. More modern studies have produced evidence to refute this hypothesis, and radiotherapy has gained acceptance in the definitive, adjuvant, and palliative treatment of cutaneous melanoma either alone or combined with other modalities. Treatment planning and techniques must be customized to this unique disease. Acute and late toxicities may manifest with treatment and may be augmented with a multimodality treatment approach. Future directions of the use of radiotherapy in cutaneous melanoma include improved treatment conformality resulting from advancements in the technology of treatment delivery and targeting. This review discusses these pertinent issues surrounding the role of radiotherapy in cutaneous melanoma.
Melanoma and radiobiologic considerations
Radioresponse
The putative radioresistance of melanoma was in part based on observations from an early clinical review published in 1936 from Memorial Sloan-Kettering, demonstrating a response rate of 2.5% with radiotherapy in approximately 400 patients with advanced disease. Equipment was poor compared with modern standards, and details describing the type of radiotherapy delivered were not provided. During this era Paterson classified melanoma as a radioresistant disease, with little justification. Despite evidence from clinical reviews demonstrating efficacy and tolerability with treatment with radiotherapy, this misconception was established and propagated in the literature, with data rarely given to support this notion.
Renewing interest in this field, in vitro studies in the 1970s using rodent and human melanoma lines suggested that the so-called radioresistance associated with melanoma may instead reflect a broad shoulder in the low-dose portion of the cell survival curve, with enhanced capacity of the melanoma cells for repair of sublethal radiation-induced damage. It has been suggested that high radiotherapy doses per fraction may be needed to overcome the cellular repair process. Small clinical trials have provided support for this approach, although results have been mixed. This belief has been called into question in view of the improved understanding of the mathematical models to determine radiation response, as well as the need to consider both the tumor and surrounding healthy tissues accounting for the therapeutic ratio. Further confounding this issue is the wide range of radiosensitivities observed with in vitro melanoma cell lines, which remains unexplained.
Fraction Size
Overgaard and colleagues have suggested that response rate for melanoma is dependent on fraction size, with complete response rates of 57% with fractions more than 4 Gy compared with 24% with fractions less than 4 Gy ( P <.001) from the results of an examination of 204 lesions of melanoma in 114 patients. Bentzen and colleagues found that a fractionation schedule of approximately 6 Gy per fraction was superior for disease control. In this study, tumor size and total dose (after accounting for tumor size) were found to significantly impact tumor control whereas total treatment time was not significant. Patients with multiple lesions had highly correlated response rates as compared with interpatient responses, implying that additional unknown parameters such as patient inherent radiosensitivity or immunologic status may be important.
Clinical response rates with different hypofractionation regimens have been variable. A Danish prospective trial including 35 recurrent or metastatic melanoma lesions in 14 patients randomized treatment to high dose per fraction radiotherapy at either 9 Gy in 3 fractions or 5 Gy in 8 fractions. Complete durable regression was found in 69% of lesions treated in both arms, with an overall response rate of 97%. No difference was found in either arm. The investigators stated that acute and late toxicity in normal tissue was acceptable and similar in both arms. A European multicenter prospective trial showed lower response rates. This trial randomized treatment of 134 metastatic or recurrent lesions of malignant melanoma in 70 patients to radiotherapy (8 or 9 Gy in 3 fractions in 8 days) alone or followed by hyperthermia. Two-year actuarial local control rate was 28% for radiotherapy alone versus 46% with the addition of hyperthermia ( P = .008). The higher dose and dose per fraction regimen also resulted in superior local control, with 25% for the 24 Gy patients compared with 56% for the 27 Gy patients on univariate analysis ( P = .02). Tumor size was not found to significantly influence outcome.
At the M.D. Anderson Cancer Center (MDACC), Ang and colleagues performed a phase 2 study to assess the efficacy and toxicity of elective adjuvant radiation given in 5 fractions of 6 Gy to patients with cutaneous melanoma of the head at neck who were considered to have high-risk features for locoregional relapse. There were 174 patients enrolled, and 79 of these patients received elective irradiation after wide local excision of lesions 1.5 mm or more thick, or Clark’s level IV to V. Thirty-two patients received adjuvant radiation after excision of primary lesion plus a limited neck dissection, and 63 patients received adjuvant radiation after neck dissection for nodal relapse. With a median follow-up of 35 months, the actuarial 5-year locoregional control rate was 88% and the 5-year survival rate was 47% for the entire group. The acute tolerance to the adjuvant hypofractionated regimen was considered to be excellent, with the most frequently observed acute reaction being transient parotid swelling. Moist skin desquamation and confluent mucositis of short duration was noted in less than 5% of patients. Three patients had late radiation complications including moderate neck fibrosis, mild ipsilateral hearing impairment, and transient exposure of external auditory canal cartilage.
Stevens and colleagues from the Royal Prince Alfred Hospital in Sydney, Australia published a retrospective review of 174 patients with local or locoregional melanoma. All patients were treated postoperatively to 30 to 36 Gy in 5 to 7 fractions over 2.5 weeks. Indications for radiation in those with disease limited to the primary site included positive margins, neurotropic desmoplastic histopathology, close margins, tumor satellites, and multiple recurrences. Patients with regional nodal metastases were irradiated for positive surgical margins, extracapsular spread, multiple involved lymph nodes, large lymph nodes, perineural or vascular involvement, and parotid lymph node involvement. The median time to recurrence was 6 months, with recurrence in the radiation field developing in 11% of patients. Median disease-specific survival was 25 months for the entire cohort, 54 months for those with disease limited to primary site, and 23 months in those with locoregional nodal involvement. Few patients developed severe acute side effects. In terms of late complications, arm lymphedema was reported in 58% of patients treated postoperatively to the axilla.
Investigators at the University of Florida published a retrospective review of 56 patients with a median follow-up of 1.7 years who were treated with wide local excision and sentinel node biopsy, or with dissection of regional lymph nodes in those with high-risk features. Patients with residual gross disease, close or positive margins, disease recurrence, satellitosis, and regional node metastases were treated with postoperative radiation. Seventy-three percent of patients were treated with 5 fractions of 6 Gy for a total dose of 30 Gy. Fractions were delivered twice a week over 2.5 weeks. Twenty-five percent of patients were treated to 60 Gy at a median of 2 Gy per fraction. Twelve percent of patients had in-field recurrences and 43% distant relapses. The 5-year in-field locoregional control rate was 87%. There was no difference noted in in-field locoregional control between the hypofractionated and conventional fractionation treatments. Two patients that received hypofractionated treatment had severe late complication of osteoradionecrosis of temporal bone and radiation plexopathy.
When compared with conventional radiotherapy regimens, hypofractionation was not found to improve outcome in the only prospective randomized trial designed to address this question, Radiation Therapy Oncology Group (RTOG) 83-05. In this trial, 137 patients were randomized to treatment with 4 fractions of 8 Gy given in 21 days weekly, and 20 fractions of 2.5 Gy given over 26 to 28 days, 5 days per week. The complete response rate was similar for both arms, 24.2% in the hypofractionated arm and 23.4% in the conventional arm. Partial response rates were also equivalent, with 35.5% in the hypofractionated arm and 34.4% in the conventional arm. There were no reported differences in normal tissue complications. Nevertheless, hypofractionated regimens for cutaneous melanoma remain under investigation. Dose fractionation for the treatment of cutaneous melanoma is further discussed later in this article.
In conclusion, enough evidence has emerged to define melanoma as a more radioresponsive tumor than historically indicated. This disease displays a wide range of radiosensitivities and may have an enhanced capacity for repair of radiation-induced damage, as shown in preclinical studies. Hypofractionated delivery of radiotherapy may help to overcome this finding and these regimens have been found to be feasible, with variable response rates. However, hypofractionation has not been shown to result in significantly improved outcomes compared with conventional radiotherapy in a single, prospective randomized trial. The ideal schedule for radiotherapy in cutaneous melanoma is still unknown. The understanding of the radiobiology has improved greatly since the early reports of the 1930s, but further studies are needed to better understand this disease and to define the ideal treatment schedule for radiotherapy in these patients.
Adjuvant radiation therapy to primary site
Surgery has remained the cornerstone of management in most patients with cutaneous melanoma. Primary tumors measuring less than 1 mm in thickness have 5-year survival rates that exceed 90% with surgical resection alone. However, retrospective reviews have revealed prognostic indicators for the likelihood of locoregional recurrence, indicating a role for adjuvant radiation to the primary lesion. High-risk clinicopathologic features include increasing tumor thickness, desmoplastic subtype, ulceration, tumor resection margin, location, and recurrence.
Urist and colleagues reported a review of 3445 patients with stage I cutaneous melanoma. In a single-factor analysis, tumor thickness, tumor ulceration, and increasing patient age were associated with an increased rate of local recurrence. Local recurrence rates at 5 years were 13.2% for patients with lesions of greater than 4 mm in size compared with 0.2% for lesions 0.76 mm. Tumors with ulceration had a local recurrence rate of 11.5% versus 1.9% without the presence of ulceration at 5 years.
Balch and colleagues performed a prospective study evaluating the optimal surgical margin with resection of primary cutaneous melanoma. Four hundred and sixty-eight patients with melanomas of the trunk or proximal extremity were randomized to a 2-cm or 4-cm radial excisional margin. A second group comprised 278 patients with melanoma of the head, neck, or distal extremities who received a 2-cm radial margin. Local recurrence rates were same whether there was a 2-cm or 4-cm margin. Multivariate analysis showed that ulceration and head or neck location were associated with increased local recurrence. Local recurrences were associated with poor survival, with a 5-year survival rate of 9% to 11% compared with 86% for those without a local recurrence.
Ulceration of cutaneous melanoma determined by microscopic evaluation represents a high-risk feature for metastases. Survival rates of those with tumors with ulcerated melanoma are significantly lower than those with nonulcerated melanomas of the same thickness. Ulceration is the only prognostic feature of primary melanoma that independently predicts outcomes for melanoma of stages I to III. In the American Joint Committee on Cancer (AJCC) staging system for cutaneous melanoma published by Balch and colleagues in 2001, the 5-year survival rate of patients with T2 nonulcerated melanoma was similar to those with a T1 ulcerated lesion (89% vs 91%), illustrating that an ulcerated lesion has the same survival rate as that of a nonulcerated lesion with a higher T stage.
Surgical resection remains the primary treatment for cutaneous melanoma. After surgical excision, recurrence rates vary between 15% and 24% for lesions 4 mm or larger, and up to 50% in those with head or neck desmoplastic melanoma subtype. In the setting of high-risk features, adjuvant radiation treatment should be considered for the primary site as well as the regional lymph nodes ( Table 1 ).
| High-risk features | Series |
|---|---|
| Tumor size >4 mm, ulceration | Urist et al, 1985, Univ. of Alabama at Birmingham |
| Microscopically positive surgical margins, close excision margins, neurotropic desmoplastic histopathology, multiple involved lymph nodes, large involved lymph nodes with extracapsular spread | Stevens et al, 2000, Royal Prince Alfred Hospital |
| Cervical lymph node involvement, extracapsular extension, >3 positive lymph nodes, clinically involved lymph nodes, involved lymph node >3 cm | Lee et al, 2000, Roswell Park Cancer Institute |
| Lymph node extracapsular extension, lymph node ≥3 cm, ≥4 involved lymph nodes, nodal recurrence after previous dissection | Ballo et al, 2003, M.D. Anderson Cancer Center |
Adjuvant radiation therapy to primary site
Surgery has remained the cornerstone of management in most patients with cutaneous melanoma. Primary tumors measuring less than 1 mm in thickness have 5-year survival rates that exceed 90% with surgical resection alone. However, retrospective reviews have revealed prognostic indicators for the likelihood of locoregional recurrence, indicating a role for adjuvant radiation to the primary lesion. High-risk clinicopathologic features include increasing tumor thickness, desmoplastic subtype, ulceration, tumor resection margin, location, and recurrence.
Urist and colleagues reported a review of 3445 patients with stage I cutaneous melanoma. In a single-factor analysis, tumor thickness, tumor ulceration, and increasing patient age were associated with an increased rate of local recurrence. Local recurrence rates at 5 years were 13.2% for patients with lesions of greater than 4 mm in size compared with 0.2% for lesions 0.76 mm. Tumors with ulceration had a local recurrence rate of 11.5% versus 1.9% without the presence of ulceration at 5 years.
Balch and colleagues performed a prospective study evaluating the optimal surgical margin with resection of primary cutaneous melanoma. Four hundred and sixty-eight patients with melanomas of the trunk or proximal extremity were randomized to a 2-cm or 4-cm radial excisional margin. A second group comprised 278 patients with melanoma of the head, neck, or distal extremities who received a 2-cm radial margin. Local recurrence rates were same whether there was a 2-cm or 4-cm margin. Multivariate analysis showed that ulceration and head or neck location were associated with increased local recurrence. Local recurrences were associated with poor survival, with a 5-year survival rate of 9% to 11% compared with 86% for those without a local recurrence.
Ulceration of cutaneous melanoma determined by microscopic evaluation represents a high-risk feature for metastases. Survival rates of those with tumors with ulcerated melanoma are significantly lower than those with nonulcerated melanomas of the same thickness. Ulceration is the only prognostic feature of primary melanoma that independently predicts outcomes for melanoma of stages I to III. In the American Joint Committee on Cancer (AJCC) staging system for cutaneous melanoma published by Balch and colleagues in 2001, the 5-year survival rate of patients with T2 nonulcerated melanoma was similar to those with a T1 ulcerated lesion (89% vs 91%), illustrating that an ulcerated lesion has the same survival rate as that of a nonulcerated lesion with a higher T stage.
Surgical resection remains the primary treatment for cutaneous melanoma. After surgical excision, recurrence rates vary between 15% and 24% for lesions 4 mm or larger, and up to 50% in those with head or neck desmoplastic melanoma subtype. In the setting of high-risk features, adjuvant radiation treatment should be considered for the primary site as well as the regional lymph nodes ( Table 1 ).
| High-risk features | Series |
|---|---|
| Tumor size >4 mm, ulceration | Urist et al, 1985, Univ. of Alabama at Birmingham |
| Microscopically positive surgical margins, close excision margins, neurotropic desmoplastic histopathology, multiple involved lymph nodes, large involved lymph nodes with extracapsular spread | Stevens et al, 2000, Royal Prince Alfred Hospital |
| Cervical lymph node involvement, extracapsular extension, >3 positive lymph nodes, clinically involved lymph nodes, involved lymph node >3 cm | Lee et al, 2000, Roswell Park Cancer Institute |
| Lymph node extracapsular extension, lymph node ≥3 cm, ≥4 involved lymph nodes, nodal recurrence after previous dissection | Ballo et al, 2003, M.D. Anderson Cancer Center |
Adjuvant radiation therapy to regional lymphatics
Regional lymph nodes are the most common site of metastatic disease for melanoma. For those who present with clinical regional nodal involvement, a lymph node dissection is typically performed. Those without clinical nodal involvement may undergo sentinel lymph node biopsy or elective lymph node dissection. Although the overall survival benefit for sentinel lymph node biopsy is yet unproven, sentinel lymph node biopsy with selective dissection has become the standard of care for melanoma. Many high-risk nodal features have been reported in the surgical literature, including nodal extracapsular extension, 4 or more lymph nodes involved, lymph nodes greater than 3 cm in diameter, positive lymph nodes in the cervical basin, and lymph nodes detected at the time of therapeutic dissection. Patients who have any of these high-risk features have a regional recurrence rate of 20% to 80% after nodal dissection. Given these high recurrence rates, one must then consider the use of adjuvant radiation therapy to regional nodes when such high-risk nodal features are present.
In a retrospective review, Creagen and colleagues reviewed 82 patients who had biopsy-proven regional nodal metastases between 1972 and 1977. They compared a control group treated with lymphadenectomy with a group of patients who received adjuvant radiation 4 weeks after lymphadenectomy. No statistically significant difference was seen in disease-free interval or survival for patients with only 1 positive node or those with more than 2 positive nodes. However, there was a significant difference in the time to recurrence, with a median time of 9 months in the control group compared with 20 months in the irradiated group.
Lee and colleagues performed a retrospective analysis of 338 patients with melanoma who had undergone lymph node dissection for pathologically involved lymph nodes at Roswell Park Cancer Institute. Seventy-five percent of patients underwent therapeutic lymph node dissection, whereas the remainder had elective node dissection. No patients were treated with adjuvant radiation to the nodal basin. Overall, nodal basin recurrence was 30% at 10 years. Patients with extracapsular extension had a 10-year nodal recurrence much higher than those without (63% vs 23%, respectively). Nodal basin failure rate was 80% for lymph nodes measuring greater than 6 cm compared with 42% for those 3 to 6 cm in size. Failure rate was 24% in the nodal basin for nodes smaller than 3 cm. They also found that the number of nodes involved predicted for nodal basin failure, with a failure rate of 63% for those with more than 10 nodes involved. Eighty-seven percent of patients with nodal basin failure developed distant metastases compared with 54% of those without nodal failure. The recommendation of these investigators was the consideration of adjuvant radiation if there was cervical lymph node involvement, extracapsular extension, more than 3 positive lymph nodes, clinically involved nodes, or nodal involvement size greater than 3 cm in diameter.
Investigators at the MDACC retrospectively reviewed patients with axillary metastases treated with axillary dissection followed by adjuvant hypofractionated radiation (30 Gy at 6 Gy per fraction over 2.5 weeks). Patients treated with adjuvant radiation had high-risk features including lymph node size 3 cm or larger, 4 or more positive lymph nodes, extracapsular extension, or recurrent disease after initial surgical resection alone. A total of 89 patients were reviewed, with a median follow-up of 63 months. Sixty-six patients were treated for their first axillary disease and 22 patients for recurrent axillary disease after prior surgery. At 5 years, they found an actuarial overall survival of 50%, disease-free survival of 46%, distant metastasis-free survival of 49%, and axillary control rate of 87%. Univariate analysis showed that axillary control was inferior if axillary disease was greater than 6 cm in size (72% vs 93%). It was also found that the distant metastasis-free and disease-free survivals were less if there were more than 2 nodes positive for metastatic disease or if the primary tumor had a Breslow thickness greater than 4 mm.
At MDACC, elective irradiation to regional lymphatics after local excision of primary cutaneous head and neck melanomas that are 1.5 mm or more thick or Clark level IV or higher has been used as an alternative to prophylactic neck dissection. In 2004, Bonnen and colleagues reported a retrospective review of 157 patients with stage I or II cutaneous melanoma of the head and neck who received elective regional radiotherapy after wide local excision of the primary lesion. The elective radiation therapy was delivered to the primary tumor site and the ipsilateral draining lymph nodes, including the supraclavicular fossa. The median prescribed dose was 30 Gy at 6 Gy per fraction delivered twice weekly to the primary site and draining lymphatics. Actuarial 5-year local control and overall survival were 94% and 58%, respectively. Actuarial 5-year regional control was 89% and distant metastasis-free survival was 63%. All local disease recurrences occurred within the radiation field, and 11 of the 15 regional recurrences also occurred within the radiation field.
In conclusion, the standard of care for patients with lymph node metastases is therapeutic lymph node dissection. Patients who have high-risk primary tumors without clinical lymph node metastases should be considered for sentinel lymph node biopsy. Such patients may also be considered as candidates for elective nodal irradiation. Postoperative nodal irradiation is indicated in those patients with high-risk clinicopathologic features including lymph node size, number of involved lymph nodes, and the presence of extracapsular extension. Unfortunately, patients with cutaneous melanoma with high-risk features have a high risk of distant metastases and overall poor prognosis. Therefore, it is appropriate to weigh potential benefits of radiation therapy in the context of the entire burden of treatment, which often also includes the possibility of systemic adjuvant therapy for patients at high risk of local and distant recurrent disease (see Table 1 ).
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
