Update on Lymph Node Management in Melanoma John T. Vetto MD, FACS Professor of Surgery Division of Surgical Oncology Oregon Health & Science University Portland, Oregon
Lymph Nodes in Melanoma Outline Can we better predict who is sentinel lymph node (SLN) positive? Can we predict who has tumor in the non-sentinel nodes (NSLNs) Heraclitus Is there still a role for CLNDx?
Halstedian Model Primary (T) Nodal (N) Distant (M) Non-Halstedian Model Primary Nodal Distant
OHSU Prospective SLN Database: Adults SLN+ Rate=12% (T1-T4)-16% (T2-T4) SLN- SLN+ p<0.001 Predictors of survival SLN status Ulceration Gender SLNWG, OHSU, 2017 n>1500 cases 4
Can We Better Predict Who is SLN Positive? 6 of 7 (all T stages) or 5 of 6 (MSLT-1; T2-T4) sentinel nodes are negative Clinical Factors: Thickness Chance of a + SLN (nonulceratedulcerated) <1 mm 3-13%* 1.01-2.0 mm 13-24% 2.01-4.0 mm 24-34% >4.0 34-54% For T1: Breslow thickness 0.75 mm, Clark level IV, and ulceration Han D, et al. Clinicopathologic predictors of sentinel lymph node metastasis in thin melanoma. J Clin Oncol. 2013 Dec 10;31(35):4387-93
NCCN Version 3.2 (2018) Recommends SNB Should be Based on Risk of a +SLN Risk Recommendation Examples 5%< Do not recommend T1a with no negative features* 5-10% Discuss and Consider T1a with negative features or T1b with no negative features >10% Discuss and Offer T1b with negative features or >T2a Negative features: young age, mitoses >2/mm2, LVI, transection https://www.nccn.org/professionals/physician/melanoma
Cellular Functions Represented in the DecisionDx-Melanoma Signature Migration/chemotaxis/ metastasis Chemokine/secreted molecules CXCL14 SPP1 CLCA2 S100A9 S100A8 BAP-1 CXCL14 MGP SPP1 Differentiation/ proliferation Cell surface receptors Structural proteins CRABP2 SPRRIB BTG1 TACSTD2 CLCA2 ROBO1 CST6 KRT6B KRT14 Gap junction/cellular adhesion Extracellular matrix protein GJA1 DSC1 PPL MGP ARG1 Immune response LTA4H S100A8 S100A9 TYRP1 ARG1 CXCL14 Transcription factor Gerami et al, Clin Cancer Res; 21(1), 2015 TRIM29 ID2 Other SAP130 EIF1B AQP1 RBM23
DecisionDx-Melanoma Test Workflow CM tumor tissue RNA isolation cdna generation and amplification (14X) Microfluidics PCR gene card 28 discriminant gene targets and 3 control genes Analysis of GEP with a proprietary algorithm to determine class and metastatic risk Class 1 low metastatic risk Class 2 high metastatic risk
Risk prediction for Stage I/II patients is refined using sub-classification Class 1 Low Risk Class 2 High Risk 0 0.41 0.5 0.59 1 Probability score Class 1A Class 1B Class 2A Class 2B Recurrence-Free Survival (n=356)
Can We Use the GEP Score to Determine Risk of +SLN? Zager et al. BMC Cancer 2018 Vetto et al. Amer Acad Derm Meeting 2018 NCCN Thresholds for SLNB (2.2018) Guideline Discuss and offer Discuss and consider Do not recommend SLN+ (positivity) rate >10% 5% to 10% <5% Develop Optimal Algorithm: Model development with retrospective cohort totaling 946 patients 1-3 Class 1 Breslow s depth 2mm and Age Independently Validate: Two prospective, multicenter cohorts totaling 1,421 patients
Demographics for prospective validation cohort (#1 and #2) for SLNB guidance Attribute Age (years) Cohort #1 (n= 584) Castle prospective multi-center studies 1,2 Cohort #2 (n=837) Independent prospective multi-center study 3 Median (range) 61 (18 100) 63 (12-101) Breslow depth (mm) Median (range) 1.2 (0 18) 1.16 (0-60) Ulceration present 18% 24% Mitotic rate 1/mm 2 65% 64% Node status positive 14% 12% T Stage T1 44% 42% T2 31% 32% T3 17% 17% T4 7% 9% GEP Class 2 25% 29% 1 Hsueh et al. J Hematol Oncol 2017 ; 2 Dillon et al. SKIN J Cutan Med 2018; 3 Vetto et al. AAD Meeting 2018
GEP subclass can predict SLNB positivity risk for patients with T1-T2 tumors and inform SLNB guidance SLN Positivity Rate 30% 20% 10% 5% 0% <55 55-64 65 Age (years) Class 1A Class 1B/2A Class 2B Thresholds based on NCCN Guidelines (v3.2018) n=1,065 NCCN Recommendations for SLNB (v3.2018) Discuss and Offer Discuss and Consider Do not Recommend GEP Result Probability of a Positive Sentinel Lymph Node for T1-T2 Patients <55 years (n=370) 55-64 years (n=247) 65 years (n=448) Class 1A 7.6% 4.9% 1.6% Class 1B/2A 19.6% 7.7% 6.9% Class 2B 24.0% 30.8% 11.9% SLN+ probability in T1-T2 patients: Is below the 5% threshold established by guidelines in those 55 years old with a Class 1A result Is above the 10% threshold established by guidelines in all age groups with a Class 2B result
Completion Lymph Node Dissection Historically the standard of care for patients with positive sentinel nodes. MSLT-II: Associated with increased disease-fee overall survival. Non-sentinel node status is an important prognostic factor (hazard ratio for death: 1.78). Faries, M. B., et al. New England Journal of Medicine 2017:23: 2211-2222.
MSLT-2: Three Questions Would an improved OS survival be seen in arm contained only patients with +NSNs? What will happen to +NSNs left in patients (in the era of new adjuvant therapies)? Can we predict which patients have +NSNs?
Results Overall Incidence of Positive NSNs in CLND Specimens 17.6% 82.4% Schuitevoerder D, Am J Surg, 2018, in press
Findings Increased tumor thickness and anatomic location (neck,groin) of the nodal basin were associated with metastasis in NSNs. Higher numbers of harvested NSNs were associated with higher rates of NSN positivity (13 vs. 20, p=0.005). Supports other studies: plus SLN tumor burden, GEP score (SSO abstract). Schuitevoerder D, Am J Surg, 2018, in press
Halstedian Model Primary (T) Nodal (N) Distant (M) Non-Halstedian Model Primary Nodal Distant
OHSU Multidiciplinary Melanoma Team Shared Beliefs Patient centered care; Platinum Rule Decisions are shared Consider clinical trials at every step of the way Exciting time for melanoma patients and providers
Shameless Plug: OHSU/Knight Multidisciplinary Melanoma Conference (Thursdays, &am, 3 rd floor CHH) Surgical Oncology Medical Oncology Radiation Oncology Dermatology Nuclear Medicine/Radiology Dermatopathology Surgical Pathology Medical Genetics Clinical Trial staff Data Managers