Impact of a 31-gene Expression Profiling Test for Cutaneous Melanoma on Dermatologists’ Clinical Management Decisions

May 2017 | Volume 16 | Issue 5 | Original Article | 428 | Copyright © May 2017


Aaron S. Farberg MD,a Alex M. Glazer MD,b Richard White MS,c and Darrell S. Rigel MD MSd

aDepartment of Dermatology, Icahn School of Medicine at Mount Sinai, New York, NY bNational Society for Cutaneous Medicine, New York, NY cIris Interactive System, Cody, WY dDepartment of Dermatology, NYU School of Medicine, New York, NY

Table 1investigate the effect of prognostic molecular profiling of CMM in an attempt to further understand its effect on the clinical management decision-making process of dermatology resident physicians. 

METHODS

Attendees at a national dermatology conference were asked to respond to a series of questions following an educational presentation on the 31-GEP test. Provided with patient information exclusive of tumor thickness, respondents were asked to identify the Breslow thickness (BT; ranging from 0.7-1.5 mm in 0.1 mm increments) at which decisions about SLNBx, imaging (including X-ray, ultrasound, computed tomography, and/or positron-emission tomography) and oncology referral would be made. Additionally, responses were obtained about willingness to utilize SLNBx or imaging based on six patient vignettes with variable clinical characteristics (Table 1). Of 172 attendees, 169 dermatology resident physicians completed the survey (99% response rate). Results were quantified and differences between response groups were assessed using a T-test or Fisher’s exact test where appropriate. Institutional Review Board approval was waived for this study. 

RESULTS

Impact of 31-GEP on Tumor Thickness-based Referral to SLNBx, Oncology, and ImagingRespondents were provided the description of a 30-year-old male with a thigh lesion that was biopsy confirmed melanoma, without ulceration or atypical features, and no family or personal history of skin cancer. We evaluated the BT at which respondents would recommend SLNBx, imaging, and medical oncology referral given that results of the 31-GEP were either not provided, a Class 1 outcome, or a Class 2 outcome. When 31-GEP results were not provided, 62%, 57%, and 55% of respondents used a BT of 1.0 mm as the inflection point to manage a patient with SLNBx, oncology referral or imaging, respectively (Figure 1). When respondents were provided with a Class 1 result in addition to clinical data, similar numbers of respondents (62%, 54% and 53%, respectively) again chose 1.0mm as the inflection point for implementing SLNBx, oncology referral and imaging. In contrast the most commonly selected BT with a Class 2 result was 0.7mm, where 52%, 42%, and 41% of respondents indicated that they would refer for SLNBx, oncology or imaging, respectively compared to 17%, 17%, and 11% with no GEP results (P<0.05). Overall, responses reflecting the BT inflection points for guiding SLNBx, oncology referral and imaging were changed 23%, 18%, and 19%, respectively, after inclusion of a Class 1 result, with risk-appropriate changes (increased BT) of 87%, 83%, and 59%, respectively, for each modality. Following the addition of a Class 2 outcome to patient characteristics, the initial BT used to guide SLNBx, medical oncology referral, and imaging was changed in 47%, 50%, and 47% of the responses, respectively, with 95%, 84%, and 97% of the cases, respectively, changed in a risk-appropriate direction (decreased BT). Recommendations for SLNbx and Imaging Using Six Patient VignettesWe examined the number of respondents who would recommend SLNBx and imaging for each of six patient vignettes of varying tumor characteristics (Table 1), comparing the number of recommendations without 31-GEP information to those based on a Class 1 or Class 2 result (Figure 2). For each vignette, a Class 1 designation resulted in a significant decrease in recommendations for both SLNBx and imaging (P<0.05) with the exception of patient #1. When given a Class 2 result, a significantly larger number of respondents recommended imaging in all cases and SLNBx in five of the six cases (P<0.005), as this procedure would already be considered for patient #6 given a BT>1 mm in accordance with current guidelines.