In a prospective multicenter analysis, Castle Biosciences’ i31-SLNB algorithm had an AUC of 0.74, compared with 0.61 for the clinicopathologic-only MIA nomogram (p=0.001), for predicting sentinel lymph node (SLN) positivity in cutaneous melanoma. Among patients categorized as low risk (<5% probability), the observed SLN positivity rate was 2.6% by i31-SLNB versus 5.8% with MIA. In the >10% predicted risk stratum, actual positivity was 21.4% with i31-SLNB versus 13.8% with MIA. In a separate multicenter cohort of 810 stage I–II patients with negative SLNs, the 31-gene expression profile independently predicted poorer five-year outcomes, adding prognostic value beyond AJCC staging (ANOVA χ²=7.75, p=0.02), with high-risk Class 2B results associated with significantly lower five-year recurrence-free survival (p<0.001). The company is presenting both data sets at the 2nd European Congress on Dermato-Oncology. One study focuses on pre-surgical decision support, positioning i31-SLNB to refine which patients cross the 5% referral threshold for SLN biopsy. The second study addresses the long-standing gap in risk stratification among SLN-negative patients, where up to one in six can still recur within five years. Together, the data aim to broaden DecisionDx-Melanoma’s role from nodal triage to longitudinal management, integrating a gene-expression score with clinicopathologic features to inform both biopsy selection and post-surgical surveillance. Strategically, this is a push to move melanoma staging beyond thickness and ulceration toward biologically anchored, risk-aligned care. It targets two pressure points: the high proportion of negative SLNBs that strain surgical capacity and expose patients to procedural morbidity, and the clinical ambiguity in SLN-negative cases where current tools underperform. The improved discrimination versus a widely used nomogram is a competitive message for sites that have standardized on clinicopathologic calculators. But the more consequential question is utility, not just accuracy: can the algorithm reduce unnecessary SLNBs without compromising outcomes, and can it meaningfully redirect follow-up intensity for SLN-negative patients in ways that affect recurrence detection and resource use? For research sites and surgical programs, implementation would shift pre-operative workflows upstream. Tissue processing, test ordering, and turnaround times need to fit within tight windows between diagnosis and planned SLNB. If adopted, volumes of SLNB could decline in lower-risk cohorts, changing OR block planning and ancillary revenue patterns. At the same time, surveillance and imaging could intensify for SLN-negative patients flagged as high risk. Sponsors and CROs may leverage the assay for stratification or enrichment in adjuvant and neoadjuvant studies, particularly where nodal status is inadequate for risk definition. Payers and guideline bodies remain the gatekeepers; improved AUC and incremental prognostic value must translate into prospective clinical utility, cost offsets, and reproducible workflows across community and academic settings. In Europe, where MIA-derived tools are entrenched, this evidence creates an opening but not yet a standard. The next milestone is real-world and prospective utility data that quantify SLNB avoidance rates, downstream complications, and non-inferiority on recurrence and survival when the test informs omission decisions. For the SLN-negative cohort, evidence that test-guided surveillance alters detection timing or adjuvant decisions will matter more than p-values. Regulatory scrutiny of laboratory-developed tests and evolving NCCN/ESMO language will shape uptake. Watch for health-economic analyses, turnaround-time commitments that align with surgical schedules, and head-to-head integrations with emerging modalities like ctDNA. The operational signal to track in 2026 is whether SLNB rates decline at test-adopting centers without a penalty on early relapse detection—if that holds, practice patterns could shift quickly.
Jon Napitupulu is Director of Media Relations at The Clinical Trial Vanguard. Jon, a computer data scientist, focuses on the latest clinical trial industry news and trends.
