PowerPoint Presentation Performance and clinical decision-making using the prognostic 40-gene expression profile (40-GEP) test in 1,018 patients with high-risk cutaneous squamous cell carcinoma (SCC) Jennifer J. Siegel, PhD1; Anesh Prasai, PhD1; Aaron S. Farberg, MD2; Matthew S. Goldberg, MD1,3 1Castle Biosciences, Inc. Friendswood, TX; 2Baylor Scott & White Health System, Dallas, TX; 3Icahn School of Medicine at Mount Sinai, New York, NY Background › The 40-gene expression profile (40-GEP) test categorizes patients with a primary SCC who have one or more clinicopathologic risk factors into three biological risk groups based on the likelihood of regional, nodal, or distant metastasis (Low= Class 1; Moderate= Class 2A; High= Class 2B).6 › The 40-GEP test has been shown to improve risk assessment and provide more precise, individualized metastatic risk stratification when combined with current risk prediction methods.6,7 › Clinical utility studies have demonstrated the ability of 40-GEP test results to guide risk-aligned patient management including changes to follow-up, surveillance imaging, SLNB, and ART following 40-GEP testing ( Table 1).8-10 This study was sponsored by Castle Biosciences, Inc. (CBI), which provided funding to the contributing centers for tissue and clinical data retrieval. JJS, AP, and MSG are employees and options holders of CBI. ASF is a consultant for CBI. 40-GEP Risk Class Overall Cohort 3-year MFS (95% CI) Overall Event Rate Univariate Cox regression analysis (HR) Class 1 95.3% (93.6-97.0%) 5.1% 1 Class 2A 81.9% (78.2-85.8%) 18.8% 4.0 Class 2B 56.9% (44.8-72.2%) 43.1% 11.4 Without 40-GEP 88.2% (86.3-90.2%) 12.3% -- Methods References 1. Rogers, JAMA Derm 2015 2. Karia, JAAD 2013 3. Cañueto, JAAD 2019 4. Ruiz, JAMA Derm 2019 5. NCCN Guidelines V2.2022 6. Wysong, JAAD 2021 7. Ibrahim, Future Onc 2021 8. Hooper, Cancer Inv 2022 9. Teplitz, JDD 2019 › The two independent cohorts (n=420, validation7 n=598, performance11) were compared and then combined (n=1,018). › Overall event (i.e., metastasis) rates, Kaplan-Meier analysis for metastasis-free survival, and univariate Cox regression analysis were performed for both the individual and combined cohorts. › Clinical utility and management changes in the combined cohort were modeled based on the pre- and post-40-GEP test results of a published clinical impact study of real-world cases.8 Results n=1,018 p<0.0001M e ta s ta s is -F re e S u rv iv a l 100% 80% 0% 20% 40% 60% Years 0 2 3 4 51 Class 1 Class 2A Class 2B Ibrahim et al.7 (n=420) Overall event rate = 15.0% Arron et al.11 (n=598) Overall event rate = 9.9% Relative fold change (Class/Overall) Relative fold change (Class/Overall) Class 1 0.4x 0.4x Class 2A 1.3x 1.7x Class 2B 3.5x 3.6x Table 2. 40-GEP accurately and consistently stratifies metastatic risk Presented at Fall Clinical Dermatology Conference, October 20-23, 2022, Las Vegas NV For more information: jsiegel@castlebiosciences.com 10. Litchman, Future Onc 2021 11. Arron, presented at ACMS 2022 › In two independent high-risk SCC cohorts, the 40-GEP consistently demonstrates significant metastatic risk stratification. › When cohorts are combined, the 40-GEP continues significant performance in identifying metastatic risk, as shown with patients receiving Class 2A and Class 2B results incurring a 4- and 11-fold increase in metastasis, respectively, when compared to those with Class 1 results. › The impact of the 40-GEP on management decisions modeled across this large cohort represents a substantial improvement in risk-aligned clinical decision-making. Conclusions Disclosures Figure 1. Performance of the 40-GEP to stratify high-risk SCC patients by risk for regional or distant metastasis (n=1,018) p < 0.0001; CI= confidence interval; HR= Hazard Ratio; MFS= metastasis free survival Figure 2. 60% of patients in the combined cohort would have a risk-aligned change in management plans post-40-GEP results Clinical Issue and Objective SCC is a common skin cancer with overall favorable prognosis. However, due to its high incidence, mortality exceeds that of melanoma.1,2 Broad guidelines, along with lack of standardized and accurate risk assessment methods complicates treatment planning for SCC patients.3-5 This study reports on 40-GEP performance metrics in risk stratification of >1,000 patients and highlight its impact on guiding risk-aligned decisions for patients with SCC › When management changes from Hooper et al.8 were used on the combined cohort to model clinical utility of the 40-GEP, results showed that 60% of high-risk SCC patients would have risk-aligned (i.e., escalation of treatment due to Class 2A or 2B results; de-escalation due to a Class 1 result) alterations in at least one of the following modalities: follow-up frequency, adjuvant radiation therapy, nodal imaging, and surveillance imaging. › Kaplan-Meier survival analysis of the combined cohort (n=1,018) demonstrated statistically significant 3-year metastasis-free survival between all classes. Cox regression analysis established hazard ratios for Class 2A and Class 2B as 4.0 and 11.4, respectively (p<0.0001). › Statistically significant 40-GEP-based risk stratification was observed in each of the two independent cohorts (p<0.001, log- rank). R is k -a li g n e d c h a n g e i n t re a tm e n t p la n Escalation due to Class 2A or 2B De-escalation due to Class 1 % Patients with change due to 40-GEP Follow-up Frequency Adjuvant Radiation Therapy Nodal Imaging Surveillance Imaging Overall Management Change 36% 37% 57% 39% 57% Table 1. Individualized risk-aligned changes in overall management plans are made post-40-GEP testing Clinical Impact Studies of 40-GEP8-10 Clinicians Patients Specific clinical recommendation changed with 40-GEP Overall change in management plan recommended with 40-GEP 34 real-world test users 6 real-world cases F/U, SLNB, baseline nodal imaging, adjuvant radiation (ART), adjuvant chemotherapy, surveillance imaging Integration of the 40-GEP Class call significantly impacted recommended patient management plans in a risk- appropriate manner while staying within guidelines. 162 dermatologists* 2 patient vignettes F/U, SLNB, nodal imaging, adjuvant radiation, adjuvant chemotherapy 402 dermatologists 3 patient vignettes F/U, SLNB referral, radiation, chemotherapy, immunotherapy F/U = follow up schedule; SLNB = sentinel lymph node biopsy; *Majority dermatologists with 8.6% dermatology NP/PA, 1.2% dermatopathologist, 1.9% other