Surgical Issues in Melanoma Mark B. Faries, MD, FACS Director, - PowerPoint PPT Presentation

Surgical Issues in Melanoma Mark B. Faries, MD, FACS Director, Donald L. Morton Melanoma Research Program Director, Surgical Oncology Training Program Professor of Surgery John Wayne Cancer Institute

Surgical Issues • Margins • How narrow? • Sentinel Lymph Node Biopsy • Who • Why • Completion Lymph Node Dissection • Why? • Why not? • Metastatic Disease (Stage IV) • Where does surgery fit?

Margin Recommendations:pre-1970* 2 cm – Cooling (1966) 5 cm – Hadley (1907) Raven (1953) Petersen (1962) Olsen (1966) 8 cm – Pack (1953) 15 cm – Petersen (1962) “As wide as possible” - Veronesi (1966) * Wong CK, Dermatologica 141: 215, 1970

Randomized Trials: <2 mm DFS French Cooperative Group (n=326) < 2 mm Swedish Melanoma Trial Group (n=989) 2 cm 5 cm WHO #10 (n= 712) 1cm 8 vs. 3 local recurrences (NS) 3 cm Khayat et al, Cancer , 2003 Apr; 97(8): 1941-6 Cohn-Cedermark, Cancer , 2000; 89: 1495 Veronesi U, Arch Surg, 1991 Apr; 126(4): 438-441

Randomized Trials: Intergroup • n=468 • Median follow up >10 years 2 cm 1-4 mm 4 cm • No difference in local recurrence • 2.6% (4cm) vs. 2.1% (2cm) • Skin grafts 46% (4cm) vs. 11% (2cm) • Risk of LR based on primary tumor

Randomized Trials: UK Trial Sweden • n=900 • n = 936 pts 1cm 2 cm 3 cm 4 cm > 2 mm Gillgren et al, Lancet, November 2011 Thomas et al. NEJM 2004

Answer Key: Current (NCCN) Recommendations 5 mm Melanoma-in-situ Breslow <1mm 1 cm Breslow 1.01-2mm 1-2 cm Breslow 2.01-4mm 2 cm Breslow >4mm 2 cm

Clinical vs. Pathological Margins

Lymph Node Treatment

Lymph Node Treatment

Regional Lymph Nodes

Elective Lymph Node Dissection: WHO #14 All (>1.5mm) 1.5- 4.0mm >4.0mm

Intergroup ELND: Overall Survival Balch, Ann Surg Oncol , 2000

Sentinel Node

Problem: Identification of patients 80% of patients undergoing ELND had negative nodes Others have concomitant systemic spread – not cured by ELND Only a subset can benefit from nodal surgery

MSLT-I Melanoma >1 mm or > Clark IV (primary analysis 1.2-3.5 mm) Randomization Wide excision alone 40% 60% Wide excision + SLN SLN - SLN + Immediate CLND CLND for Recurrence No recurrence: Observation observation

MSLT-I prognosis

SLN Biopsy and Disease-Free Survival: MSLT-I Thick (≥3.5mm) Intermediate Thickness (1.2-3.5mm)

Delayed treatment  metastatic spread within the regional nodal basin 3.5 3.3 ± 0.5 Mean # Pos. Nodes 3 2.5 2 Watch & 1.4 ± 0.1 1.5 Wait SNB 1 0.5 0 Immediate CLND Delayed CLND

Impact of Clinical Recurrence: Morbidity MSLT 1

Overall Melanoma Related Survival (Breslow 1.20 – 3.5mm) Final Dataset 100 SNB Survival (%) 75 OBS HR: 0.84 50 P=0.18, 95% CI (0.64-1.09) Group # Event / Estimate S(t) ± SE Total N 5-year 10-year 25 OBS 97 / 500 85.7 ± 1.6 % 78.3 ± 2.0% 86.6 ± 1.3 % 81.4 ± 1.5 % SNB 125 / 770 0 0 2 4 6 8 10 12 Time (years)

MSLT-I Melanoma >1 mm or > Clark IV (primary analysis 1.2-3.5 mm) DSS: Primary Endpoint DFS: Secondary Endpoint Randomization Wide excision alone 40% 60% Wide excision + SLN SLN - SLN + Occult Stage III Immediate CLND CLND for Recurrence No recurrence: Observation observation

24 Melanoma Specific Survival – Node+ Morton A 50 Year Odyssey 111509 (1.2-3.5mm) Final Dataset Group # Event / Estimate S(t) ± SE % Total N 5-year 10-year OBS, had nodal recur. 48/87 57.5 ± 5.4 41.5 ± 5.6 100 SNB+ 70 / 193 69.8 ± 4.4 62.1 ± 4.8 75 SNB+ Survival (%) 50 OBS HR: 0.56 25 95% C.I. (0.37, 0.84) Log Rank P=0.006 0 0 2 4 6 8 10 12 Time (years)

Latent Subgroup Analysis

26 Melanoma Specific Survival – Node+ Morton A 50 Year Odyssey 111509 (1.2-3.5mm) Final Dataset Group # Event / Estimate S(t) ± SE % Total N 5-year 10-year OBS, had nodal recur. 48/87 57.5 ± 5.4 41.5 ± 5.6 100 SNB+ 70 / 193 69.8 ± 4.4 62.1 ± 4.8 75 SNB+ Survival (%) 50 OBS HR: 0.56 25 95% C.I. (0.37, 0.84) Log Rank P=0.006 0 0 2 4 6 8 10 12 Time (years)

Selection for SLN: Thick Melanoma? Overall Survival

Thin Melanoma? Melanoma-specific Survival

Node-Positive Thin Melanoma: Outcomes

Thin Melanoma SLN predictors Problems: – SLN population is selected – SLN has false negatives – SLN has shorter follow up – Use clinical nodal recurrence instead

Predictors 10.0 7.0 10.0 Clark Ulceration Breslow 6.0 8.0 8.0 5.0 6.0 6.0 4.0 3.0 4.0 4.0 2.0 2.0 2.0 1.0 0.0 0.0 0.0 I II III IV V UNK Yes No Unknown 0.01-0.25 0.26-0.50 0.51-0.75 0.76-0.99 Primary Site 5.0 5.0 Gender Age 4.0 4.0 4.0 3.5 3.0 3.0 3.0 2.5 2.0 2.0 2.0 1.5 1.0 1.0 1.0 0.5 0.0 0.0 0.0 Extremity Head/neck Trunk Female Male <30 30-39 40-49 50-59 60-69 >=70

Predicted probabilities of Nodal Recurrence Predicted % Breslow Age Sex node recurrence >70 50-70 <50 <0.5 >70 female 0.1 <0.5 >70 male 0.4 Male Female <0.5 50-70 female 0.3 <0.5 50-70 male 0.9 <0.50 0.51-0.75 0.76-0.99 <0.5 <50 female 0.6 <0.5 <50 male 2.1 0.51-0.75 >70 female 0.5 0.51-0.75 >70 male 1.7 0.51-0.75 50-70 female 1.2 0.51-0.75 50-70 male 4.1 0.51-0.75 <50 female 2.9 0.51-0.75 <50 male 9.2 0.76-0.99 >70 female 1.0 0.76-0.99 >70 male 3.4 0.76-0.99 50-70 female 2.5 0.76-0.99 50-70 male 8.1 0.76-0.99 <50 female 5.8 Concordance index = 0.79 0.76-0.99 <50 male 17.4

CLND: Rationale and Data

MSLT2 : Is CLND necessary in SN(+) LN basins? 79-88% of patients have Negative NSN nodes in CLND specimen CLND(+) NSN(-) # SN(+) Stain n (%) % MSLT-I 187 22 (11.8%) H&E 88% JWCI 322 39 (12.1%) H&E 88% Cochran 90 19 (21.1%) IHC 79%

Equipoise: ? Disadvantages Advantages • Additional surgery • Potential removal of – Larger incision more cancer (10-20%) – JP drain • Complete Staging • Potential Information complications: – Lymphedema • Clinical trial eligibility • Disease may already be systemic • Ultrasound may pick up any recurrence at an early time point

Is CLND necessary in SN(+) LN basins? RFS MSS Multivariable: HR 1.51, p=0.09

JWCI Retro Data

DeCOG Trial • Randomized 1:1 to CLND or observation • Powered to detect 10% absolute survival difference with 80% power • No Head/Neck Melanomas • Median Breslow 2.4 mm • About 2/3 of patients’ SLN disease <1 mm

DeCOG Trial: Discussion/Conclusions • Better nodal recurrence rate (14.6 vs 8.3%) • Not better MSS “Based on our findings, complete lymphadenectomy cannot be recommended in melanoma patients with micro- metastases.” • Difficult recruitment - High refusal/dropout • Did not achieve target accrual -Decreased statistical power • Follow up <3 years

MSLT-II and MILND

MSLT II: Trial Design Melanoma >1.2 mm or > Clark IV, n=3500 LM/SL: standard and molecular assessment Melanoma: + SLN - + Observation (Outside Center) n=700 Stratification: MSLT1 Center Breslow Randomization n=1926 Ulceration SLN H&E vs. PCR Immediate CLND Nodal Ultrasound No Recur Recur Observation Delayed CLND Observation

64

Accrual: Complete 2000 1800 All North Am 1600 Europe Australia 1400 Target 1200 1000 800 600 400 200 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 0 1 2 3 4 5 6 7 8 9 10

MSLT-II Possible Outcomes Morton SSO PI MSLT-II 5Mar11 45

Minimally Invasive: MILND

Minimally Invasive: MILND

Minimally Invasive: MILND

Minimally Invasive: MILND

Minimally Invasive: MILND

Distant Metastases

Surgery for Metastatic Melanoma: Heresy? • It’s too late for surgery, a local therapy • Surgery is morbid and complicated • Risk/Benefit Ratio very high

Meta-analysis of Phase 2 Trials Korn et al. J Clin Oncol . Korn et al. J Clin Oncol . Feb 1 2008, 527-34. Feb 1 2008, 527-34.

Better Staging 2008 • CT scanning 2003 Circa 1990

Vaccines: CancerVax AJCC Stage IV Melanoma Resection of Metastatic Lesions Stratification Factors • Site of metastasis : M1a: soft-tissue & nodal mets M1b: visceral mets • # individual lesions : 1, 2-3, 4-5 Randomize N=496 BCG + Canvax. BCG + Placebo

MMAIT-IV Overall Survival (Intent To Treat) 1.0 Canvaxin TM Placebo Median Survival (months) 32 39 0.8 Survival at 5 years Overall Survival 40% 45% 0.6 BCG + Placebo n=250 0.4 BCG + Canvaxin TM n=246 0.2 HR=1.18 P=0.245 BCG/Pl BCG/Cv 0.0 0 12 24 36 48 60 72 84 96 108 120 132 Time (months)