PowerPoint Presentation


The integrated 31-gene expression profile (i31-GEP) test for cutaneous melanoma outperforms CP-GEP at
identifying patients who can safely forego sentinel lymph node biopsy.

Background

Methods

› Management of patients with melanoma involves multiple decision points
during clinical care, all of which, in line with guidelines, should be aligned

with a patient’s risk for poor outcomes. The 31-gene expression profile

(GEP) was developed and validated to predict a patient’s risk of

recurrence and further validated to precisely predict a patient’s

individualize risk for a positive SLNB.

› An SLNB risk threshold weighs surgical risks against those of missing a
positive SLN. Current guidelines recommend a 5% risk threshold for

considering SLNB in patients with cutaneous melanoma (T1a with high-

risk features, T1a-HR-T4).1

› A 5% threshold indicates that, in a group of 20 similar patients foregoing
SLNB, 19 would have a negative SLNB, with one missed positive SLN

(19:1 negative:positive ratio).2,3 Any novel test to identify patients who

can forego SLNB should increase the ratio of negative-to-missed positive

nodes (Figure 1).

› A second GEP test was developed to identify patients at low risk of SLN
metastasis, CP-GEP, but is not available for survival prognostication.11-12

Acknowledgments & Disclosures

References

› Funding provided by Castle Biosciences.
› AJ and PP are on the speaker's bureau for Castle Biosciences. BM is an employee and stock and options holder at

Castle Biosciences.

Results

Presented at the 2022 Fall Clinical Dermatology Conference.

› Standard of care suggests that at a 5% risk
threshold, for every 20 patients not getting an

SLNB, one positive node will be missed (19:1

true-to-false negative). To be safe and

clinically useful, any new test must do better.

› i31-GEP: 30:1 true-to-false negative SLNB
ratio is better than using standard of care for

identifying patients who may safely forego

SLNB.

› CP-GEP: 15:1 true-to-false negative SLNB
ratio is worse than using standard of care.

› The i31-GEP is the only test to offer both
SLNB risk prediction and risk of recurrence,

metastasis, or death prognostication.

Conclusions

› We compared the performance of two GEP tests, the i31-GEP (n=763)4 and the
CP-GEP (U.S. validation cohort; n=153 [includes three T1a]),12 in patients with T1b-

T2 tumors, with known SLNB results, to determine if either test increased the ratio

of negative-to-missed positive nodes.

Scan or click here  
for more info

Figure 2. Only the i31-GEP performs better than standard of 
care at identifying those who can safely forgo SLNB (T1b-T2)

Figure 1. Current guidelines suggest considering SLNB when the 
risk of a positive biopsy is ≥5% (T1aHR– T4)

1. National Comprehensive Cancer Guidelines, v2, 2022. 2. Vickers et al. BMJ 2015. 3. Vickers et 
al. Diagn Progn Res. 2019. 4. Whitman et al. JCO PO 2021. 5. Vetto et al. Future Oncology. 2019. 
6. Hsueh et al. JCO PO 2021. 7. Jarell et al.  JAAD. 2022. 8. Arnot et al. AJS. 2021. 9. Dillon et al. 
CMRO.2022. 10. Ahmed et al.  Cancer Med. 2022. 11. Bellomo et al. JCO PO 2020. 12. Yousaf et 
al. IJD 2021. 13. Mulder et al. BJD 2020.

19:1 15:130:1

i31-GEP Standard of Care CP-GEP

Missed positive SLNTrue negative SLN (safely forego SLNB)

CP-GEP would miss more positive nodes per 100 ‘low-risk’ patients 
(n~6; 100/15) than using the current standard of 5% (n=5), while 
i31-GEP would miss less than the standard (n~3; 100/30) and half as 
much as CP-GEP.

1 missed positive SLN per 
31 considered low-risk

1 missed positive SLN 
per 20 not done

1 missed positive SLN per 
16 considered low-risk

31-GEP: Better than 
standard of care

CP-GEP: Worse than 
standard of care

Clinical Impact and Objective
• Patient management decisions, including the decision to undergo SLNB, 

should be risk-appropriate to the individual being considered for 
treatment. Currently, national guidelines recommend patients consider 
SLNB when risk reaches a 5% threshold, broadly identified by T-stage 
(T1a with high-risk features and greater).  Thus, by guidelines, an 
allowable threshold for true negatives to false negatives when foregoing 
SLNB is 19:1, and any test use to guide this decision should be superior 
to this benchmark.

• To compare the utility of the i31-GEP and CP-GEP for SLNB guidance 
with the current standard of care in T1b-T2 cutaneous melanoma.

i31-GEP results adapted from Whitman et al. JCO PO 2021.4 CP-GEP results obtained from Yousaf et al. IJD 2021.12

TN: True negative. FN: False negative.

Test TN FN Ratio (TN:FN)
i31-GEP 154 5 30:1 (154/5)
Standard 19 1 19:1 (19/1)
CP-GEP 60 4 15:1 (60/4)

Abel Jarell, MD,1 Brian Martin, PhD,2 Peter Prieto, MD, MPH3

1Northeast Dermatology Associates, PC, Portsmouth, NH, 2Castle Biosciences, Inc., Friendswood, TX, 3University of Rochester Medical Center, Rochester, NY

https://castlebiosciences.com/research-development/publications/