Radiation Oncology/Melanoma/Lymphatics

Melanoma: Management of regional lymphatics

Evaluation of regional lymphatics

Elective LND vs Delayed LND

• Overall, no survival benefit for performing immediate LND vs observation/delayed LND
• Results from WHO Melanoma Trial #14 suggested that for patients who have LN disease, immediate LND improves survival over delayed LND
• Planned subgroup analysis of the Intergroup Melanoma trial showed survival benefit for intermediate risk patients (non-ulcerated, thickness 1.0-2.0 mm, or located in extremity). It was felt that high risk patients (ulcerated, thickness >2.0 mm, and trunk) are at high risk of distant failure, and immediate LND does not provide significant benefit

• Intergroup Melanoma Surgical Trial (1983-1989) -- elective LND vs. observation
  • Randomized. 740 patients, Stage I-II melanoma, intermediate Breslow thickness (1.0-4.0 mm). WLE minimum margin 2cm. . Planned subgroup analysis based on previously known risk factors
  • 2000 PMID 10761786 (Free full text) — "Long-term results of a multi-institutional randomized trial comparing prognostic factors and surgical results for intermediate thickness melanomas (1.0 to 4.0 mm). Intergroup Melanoma Surgical Trial." (Balch CM, Ann Surg Oncol. 2000 Mar;7(2):87-97.) Median F/U 10 years
    • Outcome: 10-year OS ELND 77% vs. observation 73% (NS). In non-ulcerated subgroup, ELND 84% vs observation 77% (SS). In thinner subgroup (Breslow 1.0-2.0), ELND 86% vs. observation 80% (SS). In extremity subgroup, ELND 84% vs. observation 78% (p=0.05)
    • Negative predictors: Breslow thickness, ulceration, age >60
    • Conclusion: No benefit overall; however, survival benefit for select patient populations (non-ulcerated, thickness 1.0-2.0, extremity).

• WHO Melanoma Trial #14 (1982-1989) -- immediate LND vs delayed LND
  • Randomized. 240 patients. Trunk melanoma, Breslow thickness >=1.5 mm (median 3.3 mm). WLE with 3cm margin. Arm 1) immediate LND vs. Arm 2) delayed LND until clinical metastases
  • 1998 PMID 9519951 -- "Immediate or delayed dissection of regional nodes in patients with melanoma of the trunk: a randomised trial. WHO Melanoma Programme." (Cascinelli N, Lancet. 1998 Mar 14;351(9105):793-6.) Mean F/U 11 years
    • Outcome: 5-year OS delayed LND 51% vs. immediate LND 62% (NS)
    • LN Status: LN+ in delayed LND 37% vs. immediate LND 22%. Median TTF 8 months. Survival of patients with immediate LN+ better than with delayed LN+ (p=0.04)
    • Conclusion: No difference. However, immediate LND offers improved survival in patients with LN mets
  • Comment: No lymphoscintigraphy to determine draining lymph node basin; substantial proportion of thickness >4.0 mm lesions which are at high risk of distant mets

• Mayo Clinic (1971-1976) -- immediate LND vs delayed LND
  • Randomized, 3 arms. 173 patients, localized malignant melanoma. Arm 1) delayed LND until clinical progression vs Arm 2) delayed LND 3 months after primary surgery vs. Arm 3) immediate LND. H&N and midline lesions excluded. Breslow depth <1.5mm in 57%
  • 1978 PMID 638981 -- "A prospective randomized study of the efficacy of routine elective lymphadenectomy in management of malignant melanoma. Preliminary results." (Sim FH, Cancer. 1978 Mar;41(3):948-56.)
    • Outcome: 5-year OS delayed LND 63% vs. short delayed LND 56% vs. immediate LND 54% (NS)
    • LN Status: LN+ in delayed LND 16% vs short delayed LND 9% vs. immediate LND 4%
    • Conclusion: Elective LND not beneficial in management of melanoma
  • 1986 PMID 3747613 -- "Lymphadenectomy in the management of stage I malignant melanoma: a prospective randomized study." (Sim FH, Mayo Clin Proc. 1986 Sep;61(9):697-705.)
    • Outcome: No difference in OS or DMFS
    • Prognostic factors: level of invasion, thickness
    • Conclusion: No difference

• WHO Cooperative Trial #1 (1967-1974) -- immediate LND vs delayed LND
  • Randomized. 553 patients, Stage I (T1-3N0) melanoma of the extremities (83% lower extremities), Breslow thickness >=1mm in 90%, >=4mm in 31%. Satellite nodules permitted. Tumors >5 cm excluded. Wide local excision with at least 3cm margin. Arm 1) immediate LND vs Arm 2) delayed LND until clinically positive LNs detected. For lower extremity, inguino-illiac LND; for upper extremity, axillary en bloc resection including pectoralis minor
  • 1982 PMID 7074555 -- "Delayed regional lymph node dissection in stage I melanoma of the skin of the lower extremities." (Veronesi U, Cancer. 1982 Jun 1;49(11):2420-30.) Follow-up 7-13 years
    • Outcome: 5-year OS immediate LND 68% vs delayed LND 69% (NS); 10-year OS 9% vs. 13% (NS). No subset benefited from immediate LND.
    • LN status: Frequency of LN+ in immediate arm 20% vs. frequency of clinical LN failure in delayed arm 22%. Mean TTF 13 months. No survival difference between patients with initially positive LN and with delayed positive LNs. 5-year OS 46% vs. 41% (NS)
    • Conclusion: Delayed LND as effective as immediated dissection in Stage I melanoma of extremities

LND patterns of failure

• Sheba Medical Center, Israel; 2005 PMID 16365238 -- "Patterns of recurrence in patients with melanoma after radical lymph node dissection." (Nathansohn N, Arch Surg. 2005 Dec;140(12):1172-7.)
  • Retrospective. 141 patients, malignant melanoma, radical LND (axilla 61%, groin 44%). Prior intervention ("tampering") in 13%. Elective LND 26%, therapeutic LND 51%, positive SLND 24%. Median F/U 3.4 years
  • Outcome: Regional recurrence 16%, in-transit mets 11%. Tampering only predictor of regional recurrence 10% vs. 83% (SS), but no impact on OS
  • Conclusion: Previous interventions in draining basin increase risk of melanoma recurrence in the surgical field. FNA and SLN are safe

• Halle-Wittenberg, Germany; 2001 PMID 11321665 -- "Superficial inguinal and radical ilioinguinal lymph node dissection in patients with palpable melanoma metastases to the groin--an analysis of survival and local recurrence." (Kretschmer L, Acta Oncol. 2001;40(1):72-8.)
  • Retrospective. 104 patients, melanoma to ilioinguinal LNs. Median F/U 10 years
  • Prognostic factors: >2 LN+, truncal primary, ilio-inguinal vs inguinal location
  • Conclusion: Presence of pelvic mets indicates systemic disease. Local recurrence depend on regional tumor burden

• John Wayne Cancer Center; 2000 (1971-1998) PMID 11005552 -- "Is adjuvant radiotherapy necessary after positive lymph node dissection in head and neck melanomas?" (Shen P, Ann Surg Oncol. 2000 Sep;7(8):554-9; discussion 560-1.)
  • Retrospective. 217 patients, primary H&N melanoma, who underwent RLND. Postop RT 10%. Median F/U 2.7 years
  • Outcome: cervical LN recurrence 14%. 5-year regional control ECE- 87% vs. ECE+ 69% (SS), nonpalpable 96% vs. palpable 87% (p=0.07), 1-3LN+ 82% vs. >3LN+ 91% (NS). On multivariate analysis, ECE+ only predictor. Failure with postop RT 14% (ECE present in 43%) vs. no PORT 14% (ECE present in 23%)
  • Conclusion: Low incidence of cervical recurrence doesn't justify RT, except in patients with ECE
  • Editorial (PMID 11005550): Philosophical differences between different centers; randomized trials such as E3697 critical

• Royal Marsden; 2000 (1984-1998) PMID 10931025 -- "Prognosis and surgical management of patients with palpable inguinal lymph node metastases from melanoma." (Hughes TM, Br J Surg. 2000 Jul;87(7):892-901.)
  • Retrospective. 132 patients, melanoma metastatic to inguinal LNs. 5-year OS 34%.
  • Predictive factors: age, number of involved lymph nodes, ECE+, presence of pelvic LN+
  • Conclusion: Prognosis of metastases to groin is variable

• Roswell Park; 2000 (1970-1996) PMID 10661355 — "Nodal basin recurrence following lymph node dissection for melanoma: implications for adjuvant radiotherapy." (Lee RJ, Int J Radiat Oncol Biol Phys. 2000 Jan 15;46(2):467-74.)
  • Retrospective. 338 patients with Stage III melanoma (H&N 17%, axilla 47%, groin 36%). Clinically positive LN 75%, elective LND 25%. No RT. Mean F/U 4.5 years
  • Outcome: Overall 10-year nodal recurrence 30%. Mean time to recurrence 1 year. Risk of recurrence by site: cervical 43%, axilla 28%, inguinal 23% (SS). ECE 63% vs 23% (SS). Clinical LND 36% vs. elective LND 16%. By size >6 cm 80% vs. 3-6cm 42% vs. <3cm 24% (SS). By number 1-3 LN+ 25% vs. 4-10 LN+ 46% vs. >10 LN+ 63%. No difference if adjuvant systemic therapy given
  • Predictors: Number of nodes, elective vs therapeutic LND. If nodal basin failure, risk of DM 87% vs. 54%
  • Conclusion: Significant risk of nodal failure, especially if cervical involvement, ECE+, >3 LN+, clinically involved LN, or LN >3 cm

• Utah; 1993 (1978-1988) PMID 8377499 -- "Local control following therapeutic nodal dissection for melanoma." (Monsour PD, J Surg Oncol. 1993 Sep;54(1):18-22.)
  • Retrospective. 48 patients, therapeutic LND (axilla 31%, groin 52%, neck 17%). Initial dissection in 15%, delayed dissection in 85%.
  • Outcome: Nodal relapse 52%. Only predictor age >50 (31% vs. 66%). TTR shorter in ECE+ 5 months vs. ECE- 16 months
  • Conclusion: High regional failure rate; further treatment should be considered

• U Pennsylvania; 1992 PMID 1341268 -- "Loco-regional nodal relapse in melanoma." (Miller EJ, Surg Oncol. 1992 Oct;1(5):333-40.)
  • Retrospective. 207 patients, NED after LND. Stratified into Group A (relapse within LND), Group B (regional relapse elsewhere), Group C (no relapse)
  • Outcome: Prognostic factors: subungal/volar location, number of positive LN
  • Conclusion: Prognostic factors describe subset of patients who could benefit from postoperative adjuvant therapy

• Memorial Sloan Kettering; 1992 PMID 1365686 -- "Recurrence patterns and outcome in 1019 patients undergoing axillary or inguinal lymphadenectomy for melanoma." (Gadd MA, Arch Surg. 1992 Dec;127(12):1412-6.)
  • Retrospective. 1019 patients with melanoma, axillary or inguinal LND
  • Outcome: Recurrence in 40% (single site 72%, multipl sites 28%). Median OS if single site 11 months, if multiple sites 3 months. Nonvisceral recurrences 18 months vs. visceral recurrences 6 months. If surgical resection, median OS 15 months vs. 4 months
  • Conclusion: Observations support selective resection

• Sydney Melanoma Unit; 1991 (1960-1990) PMID 1951880 -- "Experience with 998 cutaneous melanomas of the head and neck over 30 years." (O'Brien CJ, Am J Surg. 1991 Oct;162(4):310-4.)
  • Retrospective. 998 patients, cutaneous melanoma of H&N (face 47%, neck 29%, scalp 14%, ear 10%) treated surgically. Superficial spreading 30%, nodular 28%, lentigo maligna melanoma 16%, other 26%. Therapeutic LND 15%, elective LND 23%
  • Outcome: Overall recurrence rate 13%. Neck recurrence rate after LND 24%, parotid recurrence rate 14%. If pN+, recurrence rate 28%. 5-year DSS 77%, 10-year 66%
  • Predictors: age, thickness, ulceration, anatomic site, LN+

• MD Anderson
  • 1986 (1970-1979) PMID 3778208 -- "The role of modified neck dissection in the treatment of cutaneous melanoma of the head and neck." (Byers RM, Arch Surg. 1986 Nov;121(11):1338-41.)
    • Retrospective. 181 patients with melanoma of H&N, modified neck dissection for suspected or proven melanoma. No RT. pN0 76%, 1 LN+ 11%, 2+ LN+ 13%
    • Outcome: neck failure rate 16%; elective dissection & N0 10%; elective dissection &N+ 14% vs. therapeutic dissection & N+ 50%
    • Conclusion: Modified neck dissection appropriate
  • 1989 PMID 2774907 -- "Patterns of relapse in 1001 consecutive patients with melanoma nodal metastases." (Calabro A, Arch Surg. 1989 Sep;124(9):1051-5.)
    • Retrospective. 1001 patients, Stage III melanoma.
    • Outcome: Nodal relapse single LN+ 9%, no ECE 15%, multiple LN+ 15-33%, ECE 28%. 5-year OS 1 LN+ 45%, 2-4 LN+ 37%, 5-10 LN+ 20%, >10 LN+ 5%, ECE+ 14%
    • Conclusion: Higher relapse with more LN+, ECE+

• Institution?; 1986 PMID 3790922 -- "Morbidity, mortality and local recurrence following regional node dissection for melanoma." (Bowsher WG, Br J Surg. 1986 Nov;73(11):906-8.)
  • Retrospective. 86 patients, melanoma with regional LND (prophylactic 28%, therapeutic 72%)
  • Outcome: Local recurrence cervical 33%
  • Toxicity: Significant wound complications short term, infrequent long term

Sentinel lymph node

• For thin (<=1 mm) melanoma, incidence of SLN+ is ~5%
• For intermediate (1-4 mm) melanoma, incidence of SLN+ is 15-20%
• For thick (>4 mm) melanoma, incidence of SLN+ is 30-50%
• If sentinel node positive, there is ~20% chance of other positive lymph nodes
• Impact of completion LND after SLNB+ is not clear; MSLT-II is ongoing and retrospective data suggest there may not be a difference in DFS
• Nodal basin recurrence rate after completion LND is 3-12% in modern series
• Please see anatomy for skin LN drainage heat maps

• MSLT-II (Ongoing)
  • Survival of SLNB-positive melanoma patients with and without complete lymph node dissection: A multicenter, randomized DECOG trial. Leiter, U. ASCO 2015 J Clin Oncol 33, 2015 (suppl; abstr LBA9002).
  • No differences in 5-year RFS, DMFS, or OS (p=0.72, 0.76, 0.86, respectively).
  • Final analysis planned 3 years after last patient accrued, approximately 2022.
• Multicenter Selective Lymphadenecomy Trial (MSLT-I) (1994-2002) -- SLNB vs. observation
  • Randomized. 1,269 patients. Clinically localized primary cutaneous melanoma, wide local excision, intermediate Breslow thickness (1.2-3.5 mm). Arm 1) SLNB + LND if node positive vs. Arm 2) observation + LND if nodal relapse
  • 2006 PMID 17005948 -- "Sentinel-node biopsy or nodal observation in melanoma." (Morton DL, N Engl J Med. 2006 Sep 28;355(13):1307-17.) Median F/U 5 years
    • Outcome: 5-year DFS SLNB 78% vs. observation 73% (SS), but this was predominately due to expected nodal relapse in the observation arm. 5-year DSS 87% vs. 87% (NS). LN+ SLNB 16% immediately vs. observation 5-year failure 16% (NS), however, higher LN+ burden in delayed arm SLNB 1.4 vs observation 3.3 (SS) indicating disease progression during observation
    • SLN Status: 5-year DSS SLN- 90% vs. SLN+ 72% (SS). If LN+, 5-year DSS SLNB+ 72% vs. observation LN+ 52% (SS). If SLN+, 70% had only that LN+, 28% had 2-3 LN+, and 2% had 4+ LN+
    • Conclusion: No overall survival difference. However, staging by SLNB can identify patients with LN+, whose survival is improved by immediate LND
  • Lessons learned (from NCCN.org 2009 conference): Similar proportion will recur in regional node basin as that with positive SLN, status of SLN most powerful predictor of outcome in intermediate thickness melanoma, if SLN+ survival better if LNs removed immediately, but there is no improvement in disease-specific survival and thus SLNB is purely a staging procedure. Cost to find one SLN+ is ~$90,000 in intermediate-thickness melanoma

• Royal Marsen; 2008 PMID 18097453 -- "Prognostic false-positivity of the sentinel node in melanoma." (Thomas JM, Nat Clin Pract Oncol. 2008 Jan;5(1):18-23.)
  • Review. Evidence that not all positive SLNs progress to clinical disease, resulting in incorrect upstaging
  • Multiple comments, links available under PMID 18097453

Completion Lymph Node Dissection

• American College of Surgeons; 2008 (2004-2005) PMID 18414952 -- "Complete lymph node dissection for sentinel node-positive melanoma: assessment of practice patterns in the United States." (Bilimoria KY, Ann Surg Oncol. 2008 Jun;15(6):1566-76. Epub 2008 Apr 15.)
  • Population survey. 44,548 patients in National Cancer Data Base, Stage I-III melanoma. SLNB done in 39%; of these 17% SLNB+.
  • Outcome: Completion LND in 50%. Less likely to undergo CLND if >75 years or lower extremity melanoma. Only 42% underwent CLND as a separate procedure (NCCN guideline). 69% had at least 10 LN examined (NCCN guideline)
  • Conclusion: Only half of patients undergo completion LND; quality surveillance measures are needed

• Multi-Institutional; 2006 PMID 16604476 -- "Melanoma patients with positive sentinel nodes who did not undergo completion lymphadenectomy: a multi-institutional study." (Wong SL, Ann Surg Oncol. 2006 Jun;13(6):809-16. Epub 2006 Apr 12.)
  • Retrospective. 134 patients with SLN+ without completion LND (patient refusal 49%, patient/physician decision 48%). Median thickness 2.6 mm, 33% ulcerated. Compared with reference 164 patients treated at MSKCC who underwent CLND. Median F/U 20 months & 30 months
  • Outcome: Nodal recurrence as component of first recurrence 15% (similar to expected non-SLN rate in patients with SLN+). TTF 11 months. Nodal recurrence-free survival CLND- 80% vs. MSKCC CLND+ 88% (p=0.07), and similar DSS (NS)
  • Conclusion: Outcome in SLN+ patients who did not undergo completion LND was comparable to matched patients with completion LND

• John Wayne Cancer Institute, California; 2004 PMID 15365064 — "Factors predictive of tumor-positive nonsentinel lymph nodes after tumor-positive sentinel lymph node dissection for melanoma." (Lee JH, J Clin Oncol. 2004 Sep 15;22(18):3677-84.)
  • Retrospective. 191 patients, SLNB positive and followed by completion LND
  • Outcome: Non-SLN+ in 24%. Predictors: Breslow thickness (>=3 mm), SLN size (>=2 mm)
  • Model: If 0 risk factors, 12%; if 1 risk factor, 31%; if both risk factors, 42%
  • Conclusion: Thicker primary and larger SLN correlate with non-SLN lymph nodes

RT to regional lymphatics

• Administration of adjuvant RT to regional lymphatics remains controversial
• Risk of relapse after lymph node dissection is 15-20%, but may be as high as 30-50% with certain high risk factors
• One randomized trial using substandard technique in the mid 1970's showed no benefit. Two further randomized trials in the U.S. were closed due to non-accrual (RTOG 93-02 and ECOG 3697). In Australia/New Zealand, TROG 02.01 has completed and showed improvement in regional control but not OS or RFS (see below).
• While it appears that adjuvant RT improves regional control, it has no impact on survival. On the other hand, a regional recurrence can be significantly morbid
• Based on multiple surgical series (see above) and retrospective institutional reviews, for now it is probably reasonable to at least discuss adjuvant RT with the following patient subsets:
  • Extracapsular extension
  • Multiple (>2-4) involved lymph nodes (fewer needed in cervical regions, more in groin regions)
  • Large lymph nodes (>=3 cm)
  • Recurrent disease after prior complete lymph node dissection

Randomized

• ANZMTG 1-02 / TROG 02.01 (Australia / New Zealand) (2002-2007) -- Observation vs Adjuvant RT
  • Randomized. 217 patients, 16 centers. High risk patients: One nodal basin, completely resected. Significant risk for relapse defined as parotid >= 1 LN+, or axilla >= 2 LN+, or groin >=3 LN+, or ECE+, or LN size neck >= 3cm or axilla/groin >= 4cm. RT to nodal basin 48 Gy/20 fx. Relapses in OBS arm could be offered RT. Primary endpoint regional relapse.
  • 2015 PMID 26206146 -- "Adjuvant lymph-node field radiotherapy versus observation only in patients with melanoma at high risk of further lymph-node field relapse after lymphadenectomy (ANZMTG 01.02/TROG 02.01): 6-year follow-up of a phase 3, randomised controlled trial." (Henderson MA, Lancet Oncol 2015 Sep;16(9):1049-60.) Median F/U 73 monoths
    • Outcome: No improvement in OS (HR: HR 1·27 [95% CI 0·89–1·79], p=0·21) or RFS (HR: 0·89 [0·65–1·22], p=0·51).
    • Toxicity: 22% G3-4, predominately skin/subcutaneous. 15% increase in limb volume vs observation (8%) over 5 years, SS.
    • Conclusion: Adjuvant RT does not improve survival or recurrence free survival.

• • 2012 PMID 22575589 -- "Adjuvant radiotherapy versus observation alone for patients at risk of lymph-node field relapse after therapeutic lymphadenectomy for melanoma: a randomised trial" (Burmeister BH, Lancet Oncol 2012 Jun;13(6):589-97.) Median F/U 40 months
    • Outcome: LN field failure improved with XRT (HR 0.56, SS). Both RFS and OS unaffected by nodal XRT.
    • Conclusion: Adjuvant RT improves regional control, but no impact on survival

• RTOG 9302 (1993-1997) -- Lymphadenectomy +/- RT
  • Randomized. Stopped due to non-accrual. Protocol Lymphadenectomy, positive cervical LNs. Arm 1) adjuvant RT 30/5 vs. Arm 2) observation

• Mayo Clinic (1972-1977) -- Lymphadenectomy +/- RT
  • Randomized. 56 patients. Lymphadenectomy with LN+; primary on trunk, extremities, or unknown. RT supervoltage, 50/28 split-course with 3-4 week after 25 Gy
  • 1978 PMID 363255 -- "Adjuvant radiation therapy for regional nodal metastases from malignant melanoma: a randomized, prospective study." (Creagan ET, Cancer. 1978 Nov;42(5):2206-10.)
    • Outcome: No difference for DFS, or OS
    • Conclusion: Radiation made no substantial contribution to results
  • Comment: Substandard split-course regimen

General

• MDACC & Roswell Park; 2009 (1983-2003) PMID 19701906 -- "The benefits of adjuvant radiation therapy after therapeutic lymphadenectomy for clinically advanced, high-risk, lymph node-metastatic melanoma." Agrawal S, Cancer. 2009 Dec 15;115(24):5836-44.
  • Comparison of 106 Roswell Park pts meeting Ballo criteria (see below), where RT was not delivered, vs 509 MDACC pts who received adjuvant RT after TL if meeting Ballo criteria (30 Gy, 5 fx's over 2.5 weeks).
  • Regional recurrence 10% in RT group vs 40% in TL alone group (SS).
  • Distant recurrence (55% vs 74%) (SS) and disease-specific survival (30% vs 51%) (SS) improved in RT group
  • Conclusion: Adjuvant RT associated w/ improved LC and may impact DSS.

• TROG 96.06
  • Phase II. 234 patients, melanoma involving regional lymph nodes or extranodal soft tissues in the basin, with high risk for recurrence (>1 LN+, ENE+, recurrence after prior LND, or tumor spill at time of surgery). Adjuvant RT 48/20. Systemic therapy during RT permitted but discouraged
  • 2002 PMID 12028092 -- "Radiation therapy following nodal surgery for melanoma: an analysis of late toxicity." (Burmeister BH, ANZ J Surg. 2002 May;72(5):344-8.)
    • Interim report. 130 patients
    • Late toxicity: Acceptable
    • Conclusion: RT regimen used could form basis of a randomized trial
  • 2006 PMID 17064803 -- "A prospective phase II study of adjuvant postoperative radiation therapy following nodal surgery in malignant melanoma-Trans Tasman Radiation Oncology Group (TROG) Study 96.06." (Burmeister BH, Radiother Oncol. 2006 Nov;81(2):136-42. Epub 2006 Oct 24.)
    • Outcome: First relapse regionally 7%. 5-year regional control 91%. 5-year PFS 27%, OS 36%. If >2 LN+, significantly worse DMFS, PFS, OS
    • Regional control subsets: Known primary 90% vs. unknown primary 89% (NS). H&N 90% vs. axilla 88% vs. ilio-inguinal 93% (NS); 1-2 LN+ 89% vs >2 LN+ 93% (NS); ECE- 90% vs. ECE+ 89% (NS); 1 operation 91% vs. >1 operation 79% (NS)
    • Late Toxicity: Grade 3 lymphedema axilla 9% vs. groin 19%; no Grade 4
    • Conclusion: Adjuvant RT after nodal surgery could offer benefit in terms of regional control

• University of Florida; 2006 (1980-2004) PMID 16973303 -- "Adjuvant radiotherapy for cutaneous melanoma: comparing hypofractionation to conventional fractionation." (Chang DT, Int J Radiat Oncol Biol Phys. 2006 Nov 15;66(4):1051-5. Epub 2006 Sep 12.)
  • Retrospective. 56 patients (H&N 87%), high risk disease (recurrent 52%, cervical LNs, lymph nodes >3cm, >3 LN+, ECE, gross residual disease, close/positive SM, satellitosis), treated with adjuvant RT. Hypofractionated RT 30/5 (73%) or conventional with median dose 60/30 (27%). Median F/U 4.4 years
  • Outcome: 5-year LRC 87%, CSS 57%, OS 46%. No difference between schedules (p=0.97) for any endpoint. In-field failure worse with satellitosis (SS)
  • Toxicity: late toxicity 4% (1 osteoradionecrosis of external auditory canal, 1 radiation plexopathy), both in hypofractionated regimens
  • Conclusion: Hypofractionation and conventional fractionation equally efficacious

• M.D.Anderson; 2006 (1983-2003) PMID 16182463 — "Combined-modality therapy for patients with regional nodal metastases from melanoma." (Ballo MT, Int J Radiat Oncol Biol Phys. 2006 Jan 1;64(1):106-13.)
  • Retrospective. 466 patients, nodal mets from melanoma, managed with LND + adjuvant RT. Therapeutic LND in 93%. Systemic therapy in 33%. Median F/U 4.2 years
  • Outcome: 5-year regional control 89%, DFS 42%, DMFS 44%, DSS 49%. No predictors of worse regional control. Predictors for DSS number of LN, ulceration
  • Toxicity: Grade 2 in 16%, Grade 3 <1%. 10-year symptomatic lymphedema 11%, worse with groin LN (27%)
  • Conclusion: Regional disease can be controlled with LND+RT, but risk of DM and melanoma death remains high
  • Refined indications:
    • For most patients: ECE+, LN >=3 cm, 4+ LN+, recurrent disease
    • For cervical patients: ECE+, smaller LNs, fewer LNs, recurrent disease
    • For groin patients: 2 of the high-risk features
    • For >10 LN+: 2 of the high-risk features, given their risk of systemic disease is very high

• Munster/Tubingen, Germany; 2001 (1980-1988) PMID 11167684 -- "Should adjuvant radiotherapy be recommended following resection of regional lymph node metastases of malignant melanomas?" (Fuhrmann D, Br J Dermatol. 2001 Jan;144(1):66-70.)
  • Retrospective. 116 patients with Stage III melanoma (58 patients from one center who underwent adjuvant RT compared with 58 controls from another center who did not; matched on number of lymph nodes and gender). RT variable, typically 50-65 Gy, in 2-3.8 Gy/fx
  • Outcome: Regional recurrence RT 15% vs. no RT 28% (NS). OS RT 17% vs. control 26% (NS). Regional mets usually accompanied by distant mets
  • Conclusion: Present study doesn't support adjuvant RT

• New York University; 2001 (1993-1999) PMID 11769862 -- "Elective radiation therapy for high-risk malignant melanomas." (Cooper JS, Cancer J. 2001 Nov-Dec;7(6):498-502.)
  • Retrospective. 40 patients, surgery, adjuvant RT for high risk melanoma. High risk criteria: primary >4mm thick (10%), close/involved SM+ (22%), multiple involved LNs (52%), ECE (15%), recurrent disease in resected region (7%). RT 30/5 or 36/6. Median F/U 1.5 years
  • Outcome: 5-year LRC 84%. 5-year OS 39%
  • Toxicity: skin erythema acutely, no late sequelae
  • Conclusion: Elective RT effectively controlled residual subclinical disease, but better adjuvant therapy necessary

• Sydney Melanoma Unit; 2000 (1989-1998) PMID 10618610 -- "Locally advanced melanoma: results of postoperative hypofractionated radiation therapy." (Stevens G, Cancer. 2000 Jan 1;88(1):88-94.)
  • Retrospective. 174 patients, Stage I-III melanoma, treated with surgery and adjuvant RT. typically for SM+ or high risk lymph node disease. Primary site 20%, lymph nodes 80%. Dose 30-36 Gy in 5-7 fractions BIW
  • Outcome: In-field recurrence 11%. Median DSS 2 years, 5-year OS 41%. Decreased survival predicted by in-field recurrence
  • Toxicity: symptomatic arm lymphedema 58%
  • Conclusion: Locoregional control remains an important goal; postop RT effective in reducing local recurrence compared to surgical data

• Oregon; 2000 (1992-1999) PMID 11034246 -- "Prevention of local recurrence after surgical debulking of nodal and subcutaneous melanoma deposits by hypofractionated radiation." (Morris KT, Ann Surg Oncol. 2000 Oct;7(9):680-4.)
  • Retrospective. 41 patients (neck 64%, axilla 22%), bulky melanoma deposits treated with surgery + postop RT. RT 30/5 BIW. Untreated cohort comparison (n=21). Mean F/U 1.9 years
  • Outcome: In-field recurrence rate 5%. Untreated cohort LR 19% (p=0.06)
  • Toxicity: minimal
  • Conclusion: Addition of RT is a well-tolerated method of providing excellent local control

• Peter MacCallum Institute, Australia; 1999 (1985-95) - PMID 10421540 — "Nodal radiation therapy for metastatic melanoma." (Corry J, Int J Radiat Oncol Biol Phys. 1999 Jul 15;44(5):1065-9.)
  • Retrospective. 113 patients. 42 no residual disease after surgery (adjuvant group); 71 pts palliative (63 macroscopic disease after surgery, 8 no surgery). Adjuvant RT 50/25 or 60/30. Palliative RT 30/6, 50/25, or 60/30
    • Adjuvant group: nodal failure 22%. 5-yr OS 33%.
    • Palliative group: Clinical CR in 23%, 68% with symptomatic response. 5-yr OS 8%. 68% failed first in nodes.
  • Conclusion: Recommend postop RT for patients at high risk of regional recurrence (multiple nodes, ECE, or recurrent nodal disease)

• Queensland Radium Institute, Australia; 1995 (1989-1993) PMID 7638990 -- "Radiation therapy for nodal disease in malignant melanoma." (Burmeister BH, World J Surg. 1995 May-Jun;19(3):369-71.)
  • Retrospective. 57 patients, isolated resectable and nonresectable nodal disease treated with RT. Variety of RT schedules. Median F/U 15 months
  • Outcome: Response rate 84%. Regional relapse 12%, but none in absence of distant mets. Median OS 1.6 years.
  • Conclusion: Need for additional studies, but definitive role for RT in patients with bulky inoperable disease

• Princess Margaret; 1983 (1975-1980) PMID 6185198 -- "0-7-21 radiotherapy in nodular melanoma." (Johanson CR, Cancer. 1983 Jan 15;51(2):226-32.)
  • Retrospective. 54 patients, nodular melanoma. Three clinical subsets: R1 surgery (41%), R2 surgery (17%), and recurrent (43%), vast majority in lymph nodes. RT 24/3 on day 0, day 7, and day 21
  • Outcome: local control R1 group 82%, R2 group 78%, and recurrent group 39%
  • Complications: 6%
  • Conclusion: Nodular melanoma not radioresistant, and large dose per fraction produces high response rate

Cervical nodes

• Sydney Melanoma Unit
  • 2008 (1990-2004) PMID 18958539 -- "Adjuvant postoperative radiotherapy to the cervical lymph nodes in cutaneous melanoma: is there any benefit for high-risk patients?" (Moncrieff MD, Ann Surg Oncol. 2008 Nov;15(11):3022-7. Epub 2008 Oct 30.)
    • Retrospective. 716 patients, cervical LN surgery. High risk disease offered RT (n=129).
    • Outcome: RT did not improve regional control
    • Conclusion: No evidence to support use of adjuvant RT for high-risk melanoma
  • 1997 (1987-1995) PMID 9323147 -- "Adjuvant radiotherapy following neck dissection and parotidectomy for metastatic malignant melanoma." (O'Brien CJ, Head Neck. 1997 Oct;19(7):589-94.)
    • Retrospective. 143 patients with metastatic melanoma to cervical LNs or parotid, 152 dissected necks/parotids. Adjuvant RT 33/6 given in 32%.
    • Outcome: Regional recurrence RT group 6% vs. no RT group 19% (p=0.055). Irradiated group more extensive nodal involvement (2+ LN 65% vs 40%, ECE 48% vs 19%). By multivariate analysis, RT trend to regional control (p=0.06), no impact on OS
    • Conclusion: Adjuvant RT trend to improved regional control

• MD Anderson
  • RT replacing LND; 2005 (1983-2003) PMID 15952196 — "Melanoma metastatic to cervical lymph nodes: Can radiotherapy replace formal dissection after local excision of nodal disease?" (Ballo MT, Head Neck. 2005 Aug;27(8):718-21.)
    • Retrospective. 36 patients, with surgical excision of parotid or cervical LN met (median 1 LN removed), no completion LND. Initial presentation (56%) or recurrence but no prior LND (44%). Treated with RT 30/5 to primary, excised nodal area, and undissected ipsilateral neck. Median F/U 5.3 years
    • Outcome: 5-year regional control 93%, failures in-field. 5-year DFS 59%, DMFS 59%
    • Toxicity: 5-year late toxicity 8%
    • Conclusion: Select patients may receive RT to regional lymphatics instead of completion LND
  • Elective nodes; 2004 (1983-1998) PMID 14716775 -- "Elective radiotherapy provides regional control for patients with cutaneous melanoma of the head and neck." (Bonnen MD, Cancer. 2004 Jan 15;100(2):383-9.)
    • Retrospective. 157 patients, Stage I-II cutaneous melanoma of H&N, WLE + elective nodal irradiation. No SLNB or elective LND. RT given to primary tumor, ipsilateral draining LNs, and SCV fossa. RT dose 30/5 BIW as below. Patients with midline lesions generally observed. Median F/U 5.7 years
    • Outcome: LR 6%, regional recurrence 10%, DM 36%. Both 5-year and 10-year regional control 89%, DSS 68% and 58%. Median TTF 17 months. Based on SLNB data, estimated that 33-40 patients had LN+, but only saw 15 LN recurrences
    • Predictors: LRC: Breslow thickness <=2 mm 92% vs. >2 mm 82% (p=0.05). Breslow thickness also predictor for DSS and DM rate
    • Toxicity: 10-year late toxicity 6%
    • Conclusion: Safety and efficacy of elective regional RT confirmed, may serve as alternative to SLNB
  • Involved nodes; 2003 (1983-1998) PMID 12655537 — "Adjuvant irradiation for cervical lymph node metastases from melanoma." (Ballo MT, Cancer. 2003 Apr 1;97(7):1789-96.)
    • Retrospective. 160 patients, involved cervical nodes (93% palpable), either at diagnosis (44%) or recurrent (56%). Median time to LN recurrence 13 months. Treated with surgery (neck dissection or local excision) and adjuvant RT 30/5. Most referred for RT due to ECE (49%), multiple >=4 LN+ (26%), large >3cm LN (21%), or recurrent neck disease after initial dissection (6%); 27% had no risk factors. RT given to primary site, ipsilateral lymphatics including SCV fossa. Median F/U 6.5 years
    • Outcome: 10-year LC 94%, RC 94%, LRC 91%, DSS 48%, DMFS 43%, DFS 42%, OS 39%
    • Predictors: No predictors for LRC
    • Toxicity: 5-year 10%, but no Grade 3+
    • Conclusion: Adjuvant RT excellent LRC. Authors recommend for ECE, LN >3cm, multiple LNs, recurrent disease, or selective neck dissection
  • Phase II; 1994 (1983-1992) - PMID 7960981 — "Postoperative radiotherapy for cutaneous melanoma of the head and neck region." (Ang KK, Int J Radiat Oncol Biol Phys. 1994 Nov 15;30(4):795-8.)
    • Phase II. 174 patients. Three groups with projected LR rate ~50%: Group A (n=79) wide local excision and elective neck RT if depth >= 1.5 mm or Clark's level IV-V. Group B (n=32) wide local excision, limited neck dissection, and adjuvant neck RT. Group C (n=63) neck dissection for nodal relapse, followed by salvage RT. RT given 30/5, prescribed at Dmax, BIW. Median F/U 2.9 years
    • Outcome: 5-year LRC 88%, 5-year OS 47%. No predictors for LRC after RT. Predictors for survival were thickness <1.5 mm 100% vs 1.5-4.0 mm 72% vs >4 mm 30%; 1-3 LN+ 39% vs. 4+ LN+ 23%
    • Toxicity: Acute tolerance excellent, late fibrosis in 3 patients
    • Conclusion: Hypofractionated regimen safe, and more effective than historical controls
  • Phase II; 1990 (1983-1988) PMID 2297407 -- "Regional radiotherapy as adjuvant treatment for head and neck malignant melanoma. Preliminary results." (Ang KK, Arch Otolaryngol Head Neck Surg. 1990 Feb;116(2):169-72.)
    • Phase II. 83 patients, high risk melanoma of head & neck (primary thicker >1.5mm or palpable LNs). RT 24/4 or 30/5
    • Outcome: 2-year LRC Group A 95%, Group B 90%, and Group C 83%. 2-year OS 80%, 71%, and 69%. Majority of failures DM
    • Toxicity: Minimal

Axillary nodes

• MD Anderson
  • 2009 (1984-2005) PMID 18774657 -- "Radiation therapy field extent for adjuvant treatment of axillary metastases from malignant melanoma." (Beadle BM, Int J Radiat Oncol Biol Phys. 2009 Apr 1;73(5):1376-82. Epub 2008 Sep 5.)
    • Retrospective. 200 patients, mucosal melanoma, Level I-III dissection, postop RT if high risk disease (typically LN size >=3 cm in 69%, >= 4 LN+ in 45%, ECE in 83%, recurrent disease after initial LND in 18%). RT 30/5 BIW to axilla only 48%, or axilla + SCV 52%. Median F/U 4.9 years
    • Outcome: 5-year axilla control 88% (no difference between axilla RT 89% vs axilla + SCV RT 86%). 5-year DMFS 46%, DFS 43%, OS 51%
    • Toxicity: No Grade 3 toxicity; Grade 1 14%, Grade 2 18%. Treatment with axilla + SCV fields associated with higher toxicity
    • Conclusion: Adjuvant axilla RT effective for local control; axillary fields sufficient with lower toxicity
  • 2002 - PMID 11958890 — "Adjuvant irradiation for axillary metastases from malignant melanoma." Ballo MT et al. Int J Radiat Oncol Biol Phys. 2002 Mar 15;52(4):964-72.
    • Retrospective. 89 patients, high risk features (LN or axillary mass >=3cm, >=4 positive LNs, ECE+, recurrent disease). RT dose 30/5 twice weekly. Portal included axillary and SCLV. Systemic therapy in 51 patients. Median F/U 5.2
    • Outcome: 5-yr axillary control 87%, OS 50%, DFS 46%, DM free 49%. Poor axillary control for nodes > 6 cm.
    • Toxicity: 30% developed treatment-related arm edema (Grade 2 or higher in 19%).
    • Conclusion: Improved axillary control with RT compared to surgery alone (expected rate of control 50-70%).

Inguinal nodes

• MD Anderson; 2004 PMID 15576833 — "A critical assessment of adjuvant radiotherapy for inguinal lymph node metastases from melanoma." (Ballo MT, Ann Surg Oncol. 2004 Dec;11(12):1079-84.)
  • Retrospective. 40 patients, inguinal/pelvic mets, treated with RT. Indications for RT: ECE+, LN >=3 cm, >=4 LN+, LN recurrence. RT dose 30/5 BIW. Median F/U 1.9 years
  • Outcome: 3 year regional control 74% (78% of recurrences were dermal lymphatics). DM free 35%, OS 38%.
  • Toxicity: Lymphedema 37%; was present before RT in 17%
  • Conclusion: RT may prevent nodal disease, but in-field dermal recurrences occur