Abstract 1058: Immune biomarkers for recurrence in stage II and III melanoma to guide patient stratification
Bibliographic record
Abstract
Abstract Stage II-III melanoma patients lack needed biomarkers to avoid over-treatment and determine need for combination immunotherapy. A “hot” tumor immune micro-environment (TIME) is predictive of survival and immunotherapy benefit in multiple tumor types. We have previously identified the ratio of CD8+/CD68+ cells as well as a 53-gene melanoma immune panel (MIP) as indicative of a favorable “hot” TIME. CD8/CD68 ratio and MIP were evaluated in specimens from 213 previously untreated patients including 164 from Roswell Park Comprehensive Cancer Center (RPCCC) and 49 from Geisinger Health Systems (GHS, n=36). Tumors were stained with Sox10, CD8, CD68, MPO, PDL1 and HLA-DR using the MOTiF workflow (Akoya Biosciences, AB). Image analysis was completed using inForm software (AB). RNA was extracted and NanoString analysis performed using published methods. Receiving Operator Curves (ROC) were generated and KM curves using cutoffs selected using maximally selected log-rank statistics to quantify survival benefit. We tested whether the CD8/CD68 ratio correlated with disease specific survival (DSS) and distant metastatic recurrence (DMR). In the RPCCC cohort (n=164), 138 specimens were evaluated by a pathologist and stained of which 15 were excluded based on tissue quality, and inForm analysis was completed on 80. Of these 80, 16 had DMR of whom 8 had died. ROC showed CD8/CD68 ratio to correlate with DMR (p=0.0397). KM analysis showed high CD8/CD68 ratio correlated with DMR free survival (p=0.00079). Within the GHS samples, 36 of 49 patients had sufficient quality tissue, of whom 13 had died. DMR data was not available. A high CD8/CD68 ratio correlated with DSS by ROC curve (p=0.027) and KM analysis (p=0.0009). MIP was run on 166 patients of whom 147 passed quality control for analysis. MIP correlated with DSS in the RPCCC cohort by ROC analysis (p=0.025) and cutoff defined based on prior cohort showed correlation with DSS (p=0.01) and trend towards correlation with DMRFS (p=0.065). To include diverse samples in the training set, MIP was retrained with the addition of 81 patients from Columbia University (CUMC). KM analysis performed using the retrained signature showed a favorable signature correlated with DSS (p=0.032) and DMRFS (p=0.00077). Immune based biomarkers including CD8/CD68 ratio and immune gene signature have potential to assist clinicians in selecting early-stage melanoma patients for more intensive immunotherapy regimens. Full analysis of the data from the RPCCC and GHS cohorts will be presented at the conference. Citation Format: Gerardo A. Espinoza, Chenxin Zhang, Jee-Young Moon, Divya B. Kenchappa, Matteo Abbruzzese, Lydia Bioh, Yadriel M. Bracero, Sharmin Sultana, Ajay Singh, Thazin Aung, John J. Krolewski, Robyn D. Gartrell, Larisa J. Geskin, Lawrence W. Leung, Tammie Ferringer, Rui Chang, Basil Horst, Kent Nastiuk, Yvonne M. Saenger. Immune biomarkers for recurrence in stage II and III melanoma to guide patient stratification [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1058.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".