Abstract B11: Prognostic biomarkers in stage II-III melanoma
Bibliographic record
Abstract
Abstract Introduction: Biomarkers that predict clinical response in stage II-III melanoma are critical to helping clinicians weigh the potential benefits of immunotherapy against its potential toxicities, as well as allowing for clinical trial population enrichment strategies. We previously defined and validated a 53-immune-gene panel, Melanoma Immune Profile (MIP), predictive of disease-specific survival (DSS). Combination of genomic and tissue-based immune biomarkers may help in the identification of more precise and predictive biomarkers. Methods: We performed quantitative multiplexed immunofluorescence (qmIF) on 64 primary stage II-III melanomas and 14 metastatic melanomas diagnosed at Columbia University Irving Medical Center (CUIMC) between 2000-2014 and 2014-2017, respectively. Formalin-fixed, paraffin-embedded (FFPE) slides were stained and analyzed for DAPI (nuclei), CD3 (T cells), CD8 (cytotoxic T cells, or CTLs), CD68 (macrophages), SOX10 (tumor), HLA-DR (antigen presentation and activation), and Ki67 (proliferation) according to published methods. Spatial relationships were quantified via the inhomogeneous pair correlation function (PCF). NanoString analysis of MIP was performed on extracted RNA from 78 stage II-III melanomas diagnosed between 2000-2014 with known DSS. Results: We find that high CTL and low macrophage density in the stroma correlates with DSS (p<0.001). The CTL/macrophage ratio in patients with primary stage II-III melanoma who subsequently died from melanoma is different from that in patients who did not die from melanoma (p<0.05). Interestingly, the CTL/macrophage ratio in patients with recurrence is similar to that of patients with metastatic tumors. Total immune cells and CD8+ T cells are diminished in metastatic samples compared to primary samples (p<0.01). Using PCF, we find that greater clustering among CTLs (<10-micron radius) is associated with nonrecurrence (p=0.042). HLA-DR+ macrophages and tumor cells are more dispersed than CTLs (>10-micron radius), and these cells are more repulsive of one another in recurrent patients (p=0.008). Further, we validate MIP’s prediction of nonprogression on CUIMC samples (AUC=0.695) and find that MIP correlates with DSS (AUC=0.719, p=0.003). In close evaluation of MIP, we find greater than 1-fold increase in expression of 9 immune genes (p<0.05 with Benjamini-Hochberg correction), including CD2 and CD27. Combination of low CD2 or CD27 expression with low CTL/macrophage ratio in the stroma predicts poor DSS by ROC curve (AUC=0.741, p<0.05 and AUC=0.756, p<0.05, respectively) and KM curve (p<0.01and p<0.001, respectively). Conclusion: QmIF of stage II-III melanomas shows that the stromal CTL/macrophage ratio is associated with DSS. Further, MIP identifies patients with a low risk of death from melanoma. Combination of the stromal CTL/macrophage ratio with the top genes of MIP also predicts DSS. Combining qmIF and mRNA analyses can therefore stratify stage II-III melanoma patients to receive immunotherapy. Note: This abstract was not presented at the conference. Citation Format: Emanuelle M. Rizk, Robyn D. Gartrell, Andrew Chen, Rui Chang, Douglas K. Marks, Camden L. Esancy, Bret Taback, Basil A. Horst, Charles G. Drake, Raul Rabadan, Yvonne M. Saenger. Prognostic biomarkers in stage II-III melanoma [abstract]. In: Proceedings of the AACR Special Conference on Melanoma: From Biology to Target; 2019 Jan 15-18; Houston, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(19 Suppl):Abstract nr B11.
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 | 0.000 | 0.001 |
| 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 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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".