key: cord-0018754-tfzckppd
authors: Yousaf, Ahmed; Tjien‐Fooh, Félicia J.; Rentroia‐Pacheco, Barbara; Quattrocchi, Enrica; Kobic, Ajdin; Tempel, Dennie; Kolodney, Michael; Meves, Alexander
title: Validation of CP‐GEP (Merlin Assay) for predicting sentinel lymph node metastasis in primary cutaneous melanoma patients: A U.S. cohort study
date: 2021-04-29
journal: Int J Dermatol
DOI: 10.1111/ijd.15594
sha: 3f772c6c7bde08c0ddd3f6d4b645407ed8137af5
doc_id: 18754
cord_uid: tfzckppd

BACKGROUND: Approximately 85% of melanoma patients who undergo a sentinel lymph node biopsy (SLNB) are node‐negative. Melanoma incidence is highest in patients ≥65 years, but their SLNB positivity rate is lower than in younger patients. CP‐GEP, a model combining clinicopathologic and gene expression variables, identifies primary cutaneous melanoma (CM) patients who may safely forgo SLNB due to their low risk for nodal metastasis. Here, we validate CP‐GEP in a U.S. melanoma patient cohort. METHODS: A cohort of 208 adult patients with primary CM from the Mayo Clinic and West Virginia University was used. Patients were stratified according to their risk for nodal metastasis: CP‐GEP High Risk and CP‐GEP Low Risk. The main performance measures were SLNB reduction rate (RR) and negative predictive value (NPV). RESULTS: SLNB positivity rate for the entire cohort was 21%. Most patients had a T1b (34%) or T2a (31%) melanoma. In the T1‐T2 group (153 patients), CP‐GEP achieved an SLNB RR of 41.8% (95% CI: 33.9‐50.1) at an NPV of 93.8% (95% CI: 84.8‐98.3). Subgroup analysis showed similar performance in T1‐T2 patients ≥65 years of age (51 patients; SLNB positivity rate, 9.8%): SLNB RR of 43.1% (95% CI: 29.3‐57.8) at an NPV of 95.5% (95% CI: 77.2‐99.9). CONCLUSION: We confirmed the potential of CP‐GEP to reduce negative SLNB in all relevant age groups. Our findings are especially relevant to patients ≥65 years, where surgery is often elective. CP‐GEP may guide SLNB decision‐making in clinical practice.

The incidence rate of cutaneous melanoma in the U.S. is rising, with more than 100,350 invasive new cases and 6,850 deaths expected in 2020. 1 Currently, sentinel lymph node biopsy (SLNB) is the standard of care for staging melanoma patients. [2] [3] [4] Referral for SLNB is currently guided by tumor thickness and ulceration. 4 For very thin melanomas, other risk factors may be taken into account, such as age and mitotic rate. 3 Despite these selection criteria, about 85% of all patients undergoing an SLNB are not found to have nodal metastasis. Therefore, a non-invasive test that could avoid putting these patients at risk for SLNB-associated complications would provide substantial clinical benefit. [5] [6] [7] In elderly patients, referral for SLNB surgery must be carefully weighed against their higher risk for surgery-related complications and comorbidities. 5, [8] [9] [10] Also, while the incidence of melanoma is highest among the elderly, SLNB positivity rates decrease with age, making the elderly a patient population for which decision-making for SLNB can be challenging. 5, 8 A tool that can deselect elderly patients for SLNB is beneficial to patients and physicians.

diagnosis. 11 This model combines Breslow thickness and patient age with the expression of eight genes in the primary melanoma to identify patients who may safely forgo SLNB due to their low risk of nodal metastasis. This model has recently been validated in a European cohort. 12 Here, we describe the first validation of CP-GEP (Merlin Assay) in a U.S. cohort with a subgroup analysis of patients 65 years or older. The validated CP-GEP model may aid in deselecting patients for SLNB, specifically patients 65 years or older, where the SLNB procedure is often elective. Quantitative polymerase chain reaction (qPCR) and CP-GEP model We performed the RNA extraction and qPCR measurements as previously described. 12 Cycle threshold (Ct) values for all target genes (GDF15, CXCL8, LOXL4, TGFBR1, ITGB3, PLAT, SERPINE2, and MLANA) were normalized by the average Ct of two housekeeping genes (RLP0 and ACTB), yielding the DCt.

We excluded patients with low RNA yield or insufficient expression of housekeeping genes. The CP-GEP probability score was calculated by combining DCt values with clinicopathologic factors (Breslow thickness and patient age at diagnosis). The CP-GEP model has a binary output: CP-GEP High Risk and CP-GEP Low Risk. Patients whose CP-GEP score was higher than the predefined cut-off value were considered High Risk. Otherwise, patients were classified as Low Risk. 12, 13 The CP-GEP model is commercially developed as the Merlin Assay.

We characterized the performance of the CP-GEP model by calculating sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), SLNB reduction rate (RR), and the corresponding 95% Clopper-Pearson CI. 14 SLNB RR was calculated as described by Mocellin et al. and represented the fraction of patients who are not selected for an analyses were performed in R (version 3.6.1). 16 We considered P-values <0.05 statistically significant. Patient characteristics were summarized using the gtsummary package in R (version

Forty-four (21%) of the 208 patients included in this study were SLNB positive. The majority of patients were diagnosed with a T1-T2 tumor (73.6%), with the largest patient groups having a T1b (34%) or T2a (31%) melanoma ( Table 1) . (Table 4 ).

We present an independent validation study of CP-GEP in a U.S. cohort, a model designed to identify patients who may safely forgo SLNB. CP-GEP performance assessment showed that the SLNB reduction rate (RR) was highest for T1 melanoma patients and then decreased as lesions increased in thickness. This trend is in agreement with previous studies. 11, 12 CP-GEP achieved an SLNB RR of 41.8% in T1-T2 melanoma patientsa group of patients who stand to benefit the most from CP-GEP molecular testing. This finding is similar to the results of a European validation study, which reported an SLNB RR of 36% for 105 T1-T2 melanoma patients (NPV of 92.1%). 12 Findings are also similar to the discovery cohort, which reported an overall SLNB RR of 42% at an NPV of 96%. 11 Since older patients have an up to four times higher incidence of melanoma with higher risks of complications and comorbidities, 5 

Cancer Stat Facts: Melanoma of the Skin. National Cancer Institute: Surveillance, Epidemiology, and End Results Program

Final trial report of sentinel-node biopsy versus nodal observation in melanoma

Melanoma: Cutaneous. 1.2021, National Comprehensive Cancer Network

Melanoma staging: evidence-based changes in the American joint committee on cancer eighth edition cancer staging manual

Identification of risk factors in lymphatic surgeries for melanoma: a national surgical quality improvement program review

Complications of sentinel lymph node biopsy for melanoma -A systematic review of the literature

Deselecting melanoma patients for sentinel lymph node biopsy during COVID-19: clinical utility of tumor molecular profiling

Melanoma in older patients: declining gap in survival between younger and older patients with melanoma

Stratifying SLN incidence in intermediate thickness melanoma patients

Trends in sentinel lymph node biopsy enactment for cutaneous melanoma

Model combining tumor molecular and clinicopathologic risk factors predicts sentinel lymph node metastasis in primary cutaneous melanoma

Validation of a clinicopathological and gene expression profile model for sentinel lymph node metastasis in primary cutaneous melanoma

Primary cutaneous melanoma risk stratification using a clinicopathologic and gene expression model: a pilot study

Two-sided confidence intervals for the single proportion: comparison of seven methods

Sentinel node status prediction by four statistical models: results from a large Bi-Institutional series (n = 1132)

R: A language and environment for statistical computing

Presentation-Ready Data Summary and Analytic Result Tables. 2020

Recent trends in cutaneous melanoma incidence and death rates in the United States, 1992-2006