From 1983 to 1998, 157 patients with Stage I or II cutaneous melanoma of the head and neck received elective regional radiotherapy after wide local excision of the primary lesion. None of the patients had received sentinel lymph node biopsy or dissection of the lymph nodes. Their medical records were reviewed retrospectively and analyzed for outcome.

RESULTS
The median follow-up for the current review was 68 months (range, 7-185 months). The disease recurred locally in 9 patients, in the neck lymph nodes in 15 patients, and distantly in 57 patients. The actuarial regional control rate was 89% at both 5 years and 10 years. The actuarial disease-specific survival and distant metastasis-free survival rates were 68% and 63%, respectively, at 5 years and 58% and 49%, respectively, at 10 years. Breslow thickness was a significant determinant of disease-specific survival and distant metastasis-free survival rates. At 10 years, 6% of patients had developed a symptomatic treatment-related complication. There were no treatment-related deaths.

CONCLUSIONS
The results of the current study confirmed the efficacy and safety of elective regional radiotherapy for patients with cutaneous head and neck melanoma predicted to have a high rate of lymph node involvement. Elective irradiation was a viable alternative to elective lymph node dissection. It may also serve as an alternative to sentinel lymph node biopsy, particularly for patients for whom dissection and systemic therapy are not therapeutic options.

INTRODUCTION  and DISCUSSION

The rate of disease recurrence in regional lymph nodes after local excision of primary malignant melanoma ranges from 20% to 60% and is directly related to pathologic features of the primary lesion, most notably Breslow thickness.In the early 1980s, a protocol of elective irradiation to regional lymphatics after local excision of primary cutaneous head and neck melanomas measuring 1.5 mm thick or Clark level IV was adopted at The University of Texas M. D. Anderson Cancer Center (MDACC) in Houston. The purpose of this treatment approach was to avoid disease recurrence in the neck, as well as unnecessary elective parotidectomy or neck dissection. The results of this protocol, reported by Ang  in 1994, indicated very satisfactory regional control, despite the known risk of clinically significant lymph node involvement in patients with such tumors.

With the maturation of sentinel lymph node biopsy (SLNB) techniques, the risk of lymph node metastasis and its direct relation to Breslow thickness have been more fully delineated. The rate of lymph node involvement has been documented to be < 5% for lesions 0.75 mm in thickness and as high as 50% for lesions > 4 mm in thickness. At many institutions, SLNB has replaced routine elective lymph node dissection (ELND) and at others it has replaced observation of the lymph node basin after local excision alone. Patients with negative SLNBs are observed, whereas patients with positive SLNBs undergo completion lymphadenectomies.

Although this policy has gained widespread acceptance among clinicians for tumors occurring outside of the cervical lymph node basin, its routine application for tumors of the neck remains controversial. Data support the accuracy of intraoperative lymphatic mapping for cutaneous head and neck melanoma. However, multiple sentinel lymph nodes are frequently identified. They can be found at widespread sites and they frequently are located within the parotid gland, which some have suggested require superficial parotidectomy for removal. Even though SLNB provides prognostic information and may direct the use of adjuvant systemic therapies, neither SLNB nor systemic therapy has been shown to be therapeutic in terms of overall survival or to fully overcome the adverse prognostic significance of lymph node metastases. For these reasons, we believe that elective neck irradiation remains a viable treatment option for patients who are not candidates for SLNB or systemic therapy protocols. In the current study, we assess the safety and efficacy of elective lymph node irradiation in providing locoregional control (LRC) in patients with cutaneous melanoma of the head and neck who are at risk for lymph node involvement.

Each patient underwent wide local excision of the primary tumor followed by adjuvant irradiation of the primary tumor site and elective irradiation of the ipsilateral draining lymph nodes, including the supraclavicular fossa. The width of the margin of resection depended on the thickness and location of the primary lesion. The microscopic margin was found to be positive in only one patient. The median length of time between local excision and irradiation was 6 weeks (range, 3-24 weeks).

Radiotherapy was generally delivered with electron beams (generally 9-12 megaelectron volts) to the target volume. The median prescribed dose was 30 grays (Gy) delivered twice weekly (on Mondays and Thursdays or on Tuesdays and Fridays) at 6 Gy per fraction over 2.5 weeks.  The dose was specified at Dmax. Junctions between adjoining fields were moved after the second and fourth fractions to improve dose homogeneity. An intraoral stent was used as necessary to reduce the dose to the oral cavity and a beveled bolus was placed over the temporal lobe to limit the dose to the brain to < 24 Gy. Patients with midline lesions were excluded from this treatment protocol and generally observed.

At MDACC, patients with clinically lymph node-negative cutaneous head and neck melanomas measuring 1.5 mm thick or Clark level IV have routinely been offered elective lymph node irradiation. This policy stemmed from surgical experiences that revealed both high rates of lymph node involvement and substantial incidences of lymph node recurrence after wide local excisions of primary lesions.When the results of this treatment approach were first reported, the risk of lymph node involvement was estimated from the rate of lymph node recurrence in patients observed after local excision of a cutaneous melanoma. The documented rate of regional disease recurrence for these patients was 20-60%. With the maturation of SLNB results, the rate of lymph node involvement (Table 4) is < 5% for lesions 0.75 mm, 10% for lesions 0.76-1.50 mm, 20% for lesions 1.51-4.0 mm, and 30-50% for lesions > 4.0 mm. Using the weighted averages from Table 4, an estimated 33-40 patients had lymph node disease at diagnosis in the current series (depending on the Breslow groupings used but not accounting for Clark level). However, of the 155 patients in our study with known Breslow thicknesses, only 15 patients had lymph node recurrences (15 of 155 vs. 33-40 of 155). Although not every lymph node metastasis will develop into clinically evident disease, this low regional disease recurrence rate is believed to be a result of elective irradiation. In addition, it is likely that the estimated incidence of lymph node involvement would be much higher if weighted averages from studies applying reverse transcriptase-polymerase chain reaction were used. Given this very satisfactory RC rate and low morbidity, the lack of survival benefit observed in the four randomized trials investigating ELND, and the controversy surrounding adjuvant interferon, elective irradiation is not an unreasonable approach.

In recent years, however, it has become increasingly clear that SLNB followed by lymphadenectomy for patients with documented lymph node disease is replacing elective treatment of the regional lymphatics. Although the overall results of four randomized trials investigating ELND were negative for survival benefit, the data indicated that subgroups of patients might have benefited from ELND. Also, there was an inferior survival rate for patients undergoing delayed dissection for regional disease recurrence when compared with patients receiving ELND and found to have lymph node disease. This suggested that lymph node dissection might benefit only patients who are known to have lymph node involvement. SLNB now allows proponents of ELND to perform dissection in patients with known lymph node disease because these are the only patients expected to benefit. By identifying lymph node metastases, SLNB provides prognostic information and allows appropriate stratification of patients' risk factors for clinical trials examining the role of adjuvant therapy.

One critical issue concerning SLNB is the predictive value of the results, particularly when performed in the head and neck. Although many studies evaluating SLNB for cutaneous melanoma have indicated that the incidence of lymph node recurrence after a negative SLNB is very low, to our knowledge few studies included significant numbers of patients with head and neck primary tumors.[30] In series including only head and neck primary melanomas, a sentinel lymph node was identified in 92-99% of patients and the rate of lymph node recurrence after a negative SLNB was < 5%.

Despite these results, there are still some concerns regarding the use of SLNB for cutaneous melanoma of the head and neck. Although the false-negative predictive value of SLNB is low, multiple sentinel lymph nodes are frequently identified. They are found at widespread sites and are frequently located within the parotid gland, requiring superficial parotidectomy for removal. Neither SLNB nor systemic therapy has been shown to be therapeutic in terms of OS or to fully compensate for the adverse prognostic significance of lymph node metastases. Therefore, for patients with cutaneous melanoma of the head and neck, elective irradiation remains a rational treatment, particularly for patients who are not candidates for systemic therapy protocols or dissection of the regional lymphatics. In these patients, knowing the status of the regional lymph nodes will have no bearing on subsequent management.

The current series and series evaluating adjuvant radiotherapy for patients with high-risk clinicopathologic features observed at the time of therapeutic dissection for lymph node disease indicate that radiotherapy is effective at sterilizing microscopic deposits of melanoma cells This suggests that therapeutic irradiation might be an alternative to completion lymphadenectomy for patients who have undergone an SLNB and had positive results. Furthermore, because LRC in the current series decreased as the thickness of the primary lesion increased (along with increasing expected incidence of lymph node involvement), the effectiveness of adjuvant radiotherapy may be related to the microscopic burden of disease. For patients with a positive SLNB, additional lymph node disease is probable, but expected to be of minimal microscopic burden and oftentimes not detected using immunohistochemical techniques. It is at this time that radiotherapy would be most effective. This therapeutic strategy uses SLNB to provide prognostic information and remove the largest burden of disease, but the radiotherapy obviates lymphadenectomy, which can require neck dissection and superficial parotidectomy.

The results of the current study confirm the efficacy and safety of elective regional radiotherapy for patients with cutaneous head and neck melanoma predicted to have a high rate of lymph node involvement. We have continued to advocate the hypofractionated regimen because of the positive experiences at MDACC with hypofractionation, a continued belief that some melanoma cell lines are more sensitive to large doses per fraction, and patient convenience.Although SLNB with directed lymphadenectomy remains the standard of care for most patients with malignant melanoma and the current treatment approach has not been tested at other centers, the fact remains that for elderly patients or those with significant comorbidity, elective lymph node irradiation can provide very satisfactory RC and may be a better option than observation of the neck.

Furthermore, elective irradiation may be an alternative to SLNB, particularly when the disease status of the lymph node basin will have no bearing on subsequent management or dissection is not an option because of patient condition. In addition, in patients who have undergone SLNB with positive results, radiotherapy has a high likelihood of controlling microscopic disease and avoids surgical dissection. We believe that a randomized trial comparing completion lymphadenectomy with therapeutic irradiation for patients with a positive SLNB is justified.