A community challenge to predict clinical outcomes after immune checkpoint blockade in non-small cell lung cancer.

BACKGROUND: Predictive biomarkers of immune checkpoint inhibitor (ICI) efficacy are currently lacking for non-small cell lung cancer (NSCLC). Here, we describe the results from the Anti-PD-1 Response Prediction DREAM Challenge, a crowdsourced initiative that enabled the assessment of predictive models by using data from two randomized controlled clinical trials (RCTs) of ICIs in first-line metastatic NSCLC. METHODS: Participants developed and trained models using public resources. These were evaluated with data from the CheckMate 026 trial (NCT02041533), according to the model-to-data paradigm to maintain patient confidentiality. The generalizability of the models with the best predictive performance was assessed using data from the CheckMate 227 trial (NCT02477826). Both trials were phase III RCTs with a chemotherapy control arm, which supported the differentiation between predictive and prognostic models. Isolated model containers were evaluated using a bespoke strategy that considered the challenges of handling transcriptome data from clinical trials. RESULTS: A total of 59 teams participated, with 417 models submitted. Multiple predictive models, as opposed to a prognostic model, were generated for predicting overall survival, progression-free survival, and progressive disease status with ICIs. Variables within the models submitted by participants included tumor mutational burden (TMB), programmed death ligand 1 (PD-L1) expression, and gene-expression-based signatures. The best-performing models showed improved predictive power over reference variables, including TMB or PD-L1. CONCLUSIONS: This DREAM Challenge is the first successful attempt to use protected phase III clinical data for a crowdsourced effort towards generating predictive models for ICI clinical outcomes and could serve as a blueprint for similar efforts in other tumor types and disease states, setting a benchmark for future studies aiming to identify biomarkers predictive of ICI efficacy. TRIAL REGISTRATION: CheckMate 026; NCT02041533, registered January 22, 2014. CheckMate 227; NCT02477826, registered June 23, 2015.

Location matters: spatial dynamics of tumor-infiltrating T cell subsets is prognostic in colon cancer.

Background: Colon cancer is a heterogeneous disease and consists of various molecular subtypes. Despite advances in high-throughput expression profiling, limitations remain in predicting clinical outcome and assigning specific treatment to individual cases. Tumor-immune interactions play a critical role, with tumors that activate the immune system having better outcome for the patient. The localization of T cells within tumor epithelium, to enable direct contact, is essential for antitumor function, but bulk DNA/RNA sequencing data lacks spatial distribution information. In this study, we provide spatial T cell tumor distribution and connect these data with previously determined genomic data in the AC-ICAM colon cancer patient cohort. Methods: Colon cancer patients (n=90) with transcriptome data available were selected. We used a custom multiplex immunofluorescence assay on colon tumor tissue sections for quantifying T cell subsets spatial distribution in the tumor microenvironment, in terms of cell number, location, mutual distance, and distance to tumor cells. Statistical analyses included the previously determined Immunologic Constant of Rejection (ICR) transcriptome correlation and patient survival, revealing potential prognostic value in T cell spatial distribution. Results: T cell phenotypes were characterized and CD3+CD8-FoxP3- T cells were found to be the predominant tumor-infiltrating subtype while CD3+FoxP3+ T cells and CD3+CD8+ T cells showed similar densities. Spatial distribution analysis elucidated that proliferative T cells, characterized by Ki67 expression, and Granzyme B-expressing T cells were predominantly located within the tumor epithelium. We demonstrated an increase in immune cell density and a decrease in the distance of CD3+CD8+ T cells to the nearest tumor cell, in the immune active, ICR High, immune subtypes. Higher densities of stromal CD3+FoxP3+ T cells showed enhanced survival outcomes, and patients exhibited superior clinical benefits when greater spatial distances were observed between CD3+CD8-FoxP3- or CD3+CD8+ T cells and CD3+FoxP3+ T cells. Conclusion: Our study's in-depth analysis of the spatial distribution and densities of major T cell subtypes within the tumor microenvironment has provided valuable information that paves the way for further research into the intricate relationships between immune cells and colon cancer development.

Modulation of SLFN11 induces changes in DNA Damage response in breast cancer.

BACKGROUND: Lack of Schlafen family member 11 (SLFN11) expression has been recently identified as a dominant genomic determinant of response to DNA damaging agents in numerous cancer types. Thus, several strategies aimed at increasing SLFN11 are explored to restore chemosensitivity of refractory cancers. In this study, we examined various approaches to elevate SLFN11 expression in breast cancer cellular models and confirmed a corresponding increase in chemosensitivity with using the most successful efficient one. As oncogenic transcriptomic downregulation is often driven by methylation of the promotor region, we explore the demethylation effect of 5-aza-2'-deoxycytidine (decitabine), on the SLFN11 gene. Since SLFN11 has been reported as an interferon inducible gene, and interferon is secreted during an active anti-tumor immune response, we investigated the in vitro effect of IFN-γ on SLFN11 expression in breast cancer cell lines. As a secondary approach to pick up cross talk between immune cells and SLFN11 expression we used indirect co-culture of breast cancer cells with activated PBMCs and evaluated if this can drive SLFN11 upregulation. Finally, as a definitive and specific way to modulate SLFN11 expression we implemented SLFN11 dCas9 (dead CRISPR associated protein 9) systems to specifically increase or decrease SLFN11 expression. RESULTS: After confirming the previously reported correlation between methylation of SLFN11 promoter and its expression across multiple cell lines, we showed in-vitro that decitabine and IFN-γ could increase moderately the expression of SLFN11 in both BT-549 and T47D cell lines. The use of a CRISPR-dCas9 UNISAM and KRAB system could increase or decrease SLFN11 expression significantly (up to fivefold), stably and specifically in BT-549 and T47D cancer cell lines. We then used the modified cell lines to quantify the alteration in chemo sensitivity of those cells to treatment with DNA Damaging Agents (DDAs) such as Cisplatin and Epirubicin or DNA Damage Response (DDRs) drugs like Olaparib. RNAseq was used to elucidate the mechanisms of action affected by the alteration in SLFN11 expression. In cell lines with robust SLFN11 promoter methylation such as MDA-MB-231, no SLFN11 expression could be induced by any approach. CONCLUSION: To our knowledge this is the first report of the stable non-lethal increase of SLFN11 expression in a cancer cell line. Our results show that induction of SLFN11 expression can enhance DDA and DDR sensitivity in breast cancer cells and dCas9 systems may represent a novel approach to increase SLFN11 and achieve higher sensitivity to chemotherapeutic agents, improving outcome or decreasing required drug concentrations. SLFN11-targeting therapies might be explored pre-clinically to develop personalized approaches.

The immune landscape of solid pediatric tumors.

BACKGROUND: Large immunogenomic analyses have demonstrated the prognostic role of the functional orientation of the tumor microenvironment in adult solid tumors, this variable has been poorly explored in the pediatric counterpart. METHODS: We performed a systematic analysis of public RNAseq data (TARGET) for five pediatric tumor types (408 patients): Wilms tumor (WLM), neuroblastoma (NBL), osteosarcoma (OS), clear cell sarcoma of the kidney (CCSK) and rhabdoid tumor of the kidney (RT). We assessed the performance of the Immunologic Constant of Rejection (ICR), which captures an active Th1/cytotoxic response. We also performed gene set enrichment analysis (ssGSEA) and clustered more than 100 well characterized immune traits to define immune subtypes and compared their outcome. RESULTS: A higher ICR score was associated with better survival in OS and high risk NBL without MYCN amplification but with poorer survival in WLM. Clustering of immune traits revealed the same five principal modules previously described in adult tumors (TCGA). These modules divided pediatric patients into six immune subtypes (S1-S6) with distinct survival outcomes. The S2 cluster showed the best overall survival, characterized by low enrichment of the wound healing signature, high Th1, and low Th2 infiltration, while the reverse was observed in S4. Upregulation of the WNT/Beta-catenin pathway was associated with unfavorable outcomes and decreased T-cell infiltration in OS. CONCLUSIONS: We demonstrated that extracranial pediatric tumors could be classified according to their immune disposition, unveiling similarities with adults' tumors. Immunological parameters might be explored to refine diagnostic and prognostic biomarkers and to identify potential immune-responsive tumors.