Egger, Michael E.

Importance Contemporary guidelines recommend sentinel lymph node biopsy (SLNB) for patients with melanoma with predicted risk of SLN metastasis greater than 10% and consideration of SLNB when the risk is 5% to 10%. A gene expression profile (GEP)–based test that can accurately identify patients with a low risk of SLN metastasis would refine selection for SLNB. Objective To establish the predictive capability of the combined clinicopathological factors and GEP (CP-GEP) test to identify patients with primary cutaneous melanoma who can safely forgo SLNB and to investigate the prognostic value of CP-GEP regarding the outcome in patients with cutaneous melanoma after a negative SLNB (not reported herein). Design, Setting, and Participants This prognostic study was conducted from September 2021 to June 2024 at 9 academic medical centers with experienced melanoma surgeons. Included were patients with biopsy-proven invasive cutaneous melanoma with T1 to T3 tumors and clinically negative regional LNs. All patients were deemed candidates for SLNB using standard clinical criteria. GEP was performed on formalin-fixed, paraffin-embedded tissue from the primary melanoma biopsy. Analyses were performed from December 2024 to August 2025. Intervention CP-GEP testing to determine risk of having a positive SLNB in patients with T1 to T3 cutaneous melanoma who were considered appropriate candidates for the procedure based on standard clinical parameters. Main Outcomes and Measures CP-GEP results were reported as low risk or high risk; the primary outcome measure was negative predictive value (NPV) in low-risk cases. Analyses included NPV assessment by tumor (T) subcategory, primary site, and age. Results A total of 1761 patients (median [IQR] age, 64 [53-72] years; 997 male [56.6%]) underwent SLNB (310 [17.6%] SLN positive) and had a successful CP-GEP test; GEP was successful in 97.7% of samples. A total of 651 patients (37.0%) were considered low risk by CP-GEP. Among low-risk cases, 46 (7.1%) were SLN positive, and NPV was 92.9% (95% CI, 90.7%-94.8%). High-risk cases had a 23.8% (264 of 1110) SLN-positive rate. The percentage of cases with low-risk CP-GEP declined with increasing T category (T1 = 346 of 507 [68.2%]; T2 = 295 of 897 [32.9%]; T3 = 10 of 357 [2.8%]). CP-GEP results were consistent in discriminating SLN-positive rates across primary sites, histologic subtypes, and mitotic count categories. In 2 large subsets, clinical stage IB (n = 1187) and patients 65 years and older (n = 832), 6.5% (95% CI, 4.6%-8.8%) of low-risk clinical stage IB cases and 6.6% (95% CI, 4.2%-9.7%) of all low-risk cases in patients 65 years and older were SLN positive vs 18.3% (95% CI, 15.3%-21.6%) and 20.3% (95% CI, 16.8%-24.2%) for high-risk cases, respectively. Conclusions and Relevance In this multicenter, prospective, blinded, prognostic study, the CP-GEP test reliably identified patients with melanoma with less than 10% SLN metastasis risk. SLN metastasis rates were approximately 3-fold greater for high-risk vs low-risk CP-GEP cases. In appropriately selected patients, the CP-GEP test may add value for shared patient-surgeon decision-making.

Abstract Background: Sentinel lymph node biopsy (SLNB) provides important staging and prognostic information that guides surveillance and adjuvant systemic therapy decisions for patients with cutaneous melanoma. At most centers, SLNB is indicated for patients with cutaneous melanoma with at least a 5% risk of having nodal metastases, typically melanomas ≥0.8 mm in thickness or selected thinner lesions with high-risk features such as elevated mitotic rate and/or ulceration. However, up to 85% of patients who undergo SLNB have a negative result. Currently, there is an unmet clinical need to better identify patients with a low risk of nodal metastasis who may safely forgo SLNB, but otherwise meet established criteria for undergoing SLNB. Previously, a model using primary tumor gene expression profile (GEP) data combined with clinicopathological features (CP) was developed to identify melanoma patients with a low risk of having a positive SLN. The model has been externally validated in multiple retrospective studies. The aim of the MERLIN_001 observational registry study is to prospectively validate the CP-GEP model in an independent multicenter cohort of primary cutaneous melanoma patients, who undergo lymphatic mapping and SLNB per current clinical guidelines. Methods: A total of 9 centers across the US are included in the study with planned enrollment of 2,340 patients, allowing for a loss of up to 30% due to screen failure, tissue loss, low RNA yield, no gene expression profile or failure to perform the planned SLNB. Patients with clinically node-negative cutaneous melanoma and planned SLNB using current guideline indications are eligible for the study and will be followed for five years. Both patients and investigators are blinded to the results (NCT04759781). FFPE material from the initial melanoma biopsy is collected and the GEP of the primary melanoma is assessed. Subsequently, CP-GEP probability scores are calculated and expressed as a binary classification (Low Risk or High Risk for nodal metastasis), which will be compared to SLNB pathology results. Performance metrics for CP-GEP will be evaluated and will include: SLNB Reduction Rate based on Negative Predictive Value, Positive Predictive Value, Sensitivity and Specificity, and the corresponding 95% confidence intervals. Finally, the performance of CP-GEP to stratify patients according to risk of recurrence (recurrence-free survival, distant metastasis-free survival, overall survival) will also be assessed, based on five-year outcomes data. Additional analyses will be performed using the data collected throughout the study. Enrollment of patients started in September 2021 and is ongoing. As of January 2023, 890 patients have been enrolled, representing 46% of the targeted number of patients with a successful SLNB and CP-GEP test result. Citation Format: Tina J. Hieken, Michael E. Egger, Christina V. Angeles, Erin E. Burke, Michael C. Lowe, John R. Hyngstrom, Georgia M. Beasley, Edmund K. Bartlett, Vernon K. Sondak. Merlin_001: a prospective registry study of a primary melanoma gene-signature to predict sentinel node (SN) status and determine its prognostic value for more accurate staging of SN-negative melanoma patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 2 (Clinical Trials and Late-Breaking Research); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(8_Suppl):Abstract nr CT053.