PowerPoint Presentation Background • Significant variability exists within the established guidelines for cutaneous melanoma patient follow-up and surveillance.1 • A validated prognostic 31-gene expression profile (GEP) test has been shown to accurately classify a patient’s risk of metastasis within five years post-diagnosis as either low (Class 1) or high (Class 2).2,3 • The test has been shown to impact management decisions, including frequency of clinical visits, imaging and blood work recommendations, and physician referrals as measured by changes in surveillance practices following receipt of the test result.4-6 Methods • A retrospective case review was performed following IRB approval at Desert Surgical Oncology, Rancho Mirage, CA. Data were collected from October 2015 through June 2016. • Eligible patients had a diagnosis of stage I-III cutaneous melanoma and underwent GEP testing as part of their routine clinical care. • Medical records were reviewed by the managing surgical oncologist. A questionnaire was completed for each patient describing the intended management strategy prior to and following the receipt of a GEP test result. • Recommendations for follow-up were categorized as blood work (labs), imaging, frequency of clinical visits, and referral to medical oncology. • Documented management changes were categorized as increased intensity, decreased intensity, or no change, based on comparison of management plans before and after receipt of GEP test result. Group comparisons were evaluated using Fisher’s exact tests. Table 4. Review of clinical impact studiesTable 2. Pre-test management plan Results Table 1. Cohort demographics Conclusions • The inclusion of GEP testing as part of the management strategy at our institution has resulted in significant risk-driven follow-up and surveillance differences between low- and high-risk patients. • Results of this study are consistent with previously published reports of the GEP’s impact on clinical management. • GEP testing in combination with conventional staging methods can be employed to develop a more efficient and individualized follow-up plan based on clinical factors as well as intrinsic biological risk. References 1. Cromwell KD, Ross MI, Xing Y, et al. Melanoma Res 2012;22:376-85. 2. Gerami P, Cook RW, Russell MC, et al. Clin Cancer Res 2015;21:175-83. 3. Gerami P, Cook RW, Wilkinson J, et al. J Am Acad Dermatol 2015;72:780-5 e783. 4. Berger AC, Davidson RS, Poitras JK, et al. Curr Med Res Opin 2016;32:1599-604. 5. Farberg AS, Glazer AM, White R, Rigel DS. J Drugs Dermatol 2017;16:428-31. 6. Schuitevoerder D, Heath M, Massimino K, et al. Ann Surg Oncol 2017;24:S144. Disclosures RWC and FAM are employees and stockholders of Castle Biosciences, Inc. The proprietary GEP test is clinically available through Castle Biosciences as the DecisionDx®-Melanoma test (www.SkinMelanoma.com). A Retrospective Case Series to Evaluate the Clinical Utility of a 31-Gene Expression Profile Test in Cutaneous Melanoma Patients Robert W. Cook, PhD1, Federico A. Monzon, MD1, David Hyams, MD2 1Castle Biosciences, Inc., Friendswood, TX, 2Desert Surgical Oncology, Rancho Mirage, CA Clinical Characteristic Overall (n = 70) Class 1 (n = 45) Class 2 (n = 25) AJCC stage (v7) I 39 (56%) 36 (80%) 3 (12%) II 29 (41%) 7 (16%) 22 (88%) III 2 (3%) 2 (4%) 0 (0%) Breslow thickness Median (range), mm 1.3 (0.4-6.8) 1.0 (0.4-2.5) 2.5 (0.8-6.8) ≤1 mm 25 (36%) 21 (47%) 2 (8%) >1 mm 45 (64%) 24 (53%) 23 (92%) Mitotic index <1/mm2 18 (26%) 15 (33%) 3 (12%) ≥1/mm2 52 (74%) 30 (67%) 22 (88%) Regression Absent 67 (96%) 43 (96%) 24 (96%) Present 3 (4%) 2 (4%) 1 (4%) Ulceration Absent 48 (69%) 39 (87%) 9 (36%) Present 22 (31%) 6 (13%) 16 (64%) Table 3. Changes by class for each surveillance method Objective • To determine differences in management strategies and surveillance between Class 1 and Class 2 patients at a single surgical oncology center. Class 1 Class 2 Decrease Increase Decrease Increase Labs 45 0 0 0 Imaging* 13 0 0 25 Visits 45 0 0 0 Referral 1 1 0 5 *p<0.0001, Fisher’s exact test Figure 1. Schematic showing management changes after inclusion of GEP test result to existing surveillance plans. GEP class was a significant predictor of change in management (p<0.0001, Fisher’s exact test). C/A/P: chest, abdomen and pelvis. Management modality Frequency Labs q3 months x 2 years and q6 months x 3 years Imaging CT scan q1 year x 5 years or none Office visits q3 months x 2 years and q6 months x 3 years Referral none Study n Result Berger (2016) Prospective, multicenter 163 patients 53% changed management after inclusion of GEP result Farberg (2017) Dermatologist survey 169 physicians 47-50% changed management after inclusion of GEP result Schuitevoerder (2017) Prospective, single center 90 patients 52% of management decision based on GEP result using decision tree model Current study Retrospective, single center 70 patients 100% changed management after inclusion of GEP result Slide Number 1