SITC 2025

Tempus LensQuickly define detailed patient cohorts using natural language processing and transform complex criteria into actionable datasets. Explore unstructured clinical data and extract nuanced insights. With Lens, the power of Tempus data is at your fingertips.

An analysis of 190,189 patients, utilizing T-cell receptor (TCR) sequencing data from Tempus’ de-identified real-world database, investigated the prevalence and distribution of mucosal-associated invariant T (MAIT) cells across numerous solid tumor types. The study determined MAIT cell abundance using TCRα chains, with comparisons adjusting for overall T cell infiltration or total TCR’s detected where relevant. MAIT cell abundance was highest in tumors of mucosal origin, such as colorectal and gastroesophageal cancers, and higher MAIT cell levels were statistically significantly in younger patients, males, and non-smokers. Furthermore, the prevalence of MAIT cells showed associations with important clinical and molecular factors, including reduced MAIT cell abundance in PD-L1 high (TPS ≥50%; pan-cancer), microsatellite instability (MSI)-high (pan-cancer and colorectal cancer), and microbial factors such as lower bacterial load (pan-cancer). These findings suggest that MAIT cell prevalence may have important clinical implications and may serve as a promising biomarker and immunotherapy target.

This study investigates the utility of the Immune Profile Score (IPS) as a molecular signature to predict the effectiveness of immune checkpoint inhibitor (ICI) therapy in rare, advanced solid cancers. Given the clinical unmet need for rare, heterogeneous cancers, the study evaluated 90 eligible patients from Tempus’ de-identified real-world database who had a rare, advanced solid cancer diagnosis and received off-label ICI treatment, excluding those with high TMB or MSI. Patients were further categorized as IPS-high or IPS-low. The findings demonstrated that IPS-high patients experienced significantly longer overall survival compared to those categorized as IPS-low. Importantly, IPS maintained its prognostic significance across all patient subgroups and clinically relevant confounders. These results support IPS as a pan-cancer biomarker capable of accurately stratifying ICI treatment outcomes and potentially supporting the label expansion of ICIs to various rare cancer types.

A detailed computational framework, TargetR, was developed by integrating public and real-world multi-omics datasets to accelerate the discovery and validation of immunotherapy targets. This approach unified pharmacologic, genomic, transcriptomic, and proteomic data, utilizing advanced analytics to assess target potential based on criteria such as normal tissue expression, mutational load, and copy number variation (CNV) correlation with expression. Validation of findings can then be performed using Tempus’ de-identified real-world database, which provides clinical features and outcome data. The framework generated an automated, user-friendly report with actionable insights, enabling the identification and prioritization of targets for treatments like bispecific T-cell engagers and CAR-T therapies, thereby supporting the development of next-generation immunotherapies.

A2 Biotherapeutics reports the first patient with non-small cell lung cancer to have a complete response after CAR T-cell therapy. Using Tempus’ xT CDx assay, A2 Bio identified HLA-A*02 loss of heterozygosity, which qualified the patient for a clinical trial. The patient achieved a durable complete response, highlighting the utility of Tempus’ molecular profiling in matching patients with difficult-to-treat cancers to innovative therapies.

This study investigated the expression of preferentially expressed antigen in melanoma (PRAME), a cancer-testis antigen, across advanced and metastatic solid tumors to help prioritize indications for PRAME-targeting T cell therapy. The methods involved analyzing PRAME mRNA expression in 4,622 patient samples using Tempus’ de-identified real-world database and The Cancer Genome Atlas, supplemented by PRAME protein assessment via immunohistochemistry on tumor microarrays. Results showed that PRAME prevalence and expression levels were highest in melanoma, endometrial cancer, ovarian cancer, lung squamous cell carcinoma, and triple-negative breast cancer. Importantly, PRAME protein expression demonstrated homogeneity within primary tumors and consistency across matched metastases, with no apparent effect from prior anti-PD1/PDL1 therapies. Findings suggest that PRAME may be a broadly applicable therapeutic target for immunotherapy, regardless of disease stage, location, or prior treatment history.

This exploratory study investigated the potential of Tempus’ Immune Profile Score (IPS), a DNA- and RNA-based molecular signature, to act as a predictive biomarker for immune checkpoint inhibitor (ICI) benefit in patients with microsatellite stable (MSS) advanced colorectal cancer (CRC). Tempus’ de-identified real-world database was utilized to identify 46 eligible MSS CRC patients who received an ICI alone or ICI-containing regimen in the third line (3L) or beyond. Using pre-ICI tissue, patients were stratified into IPS-High versus IPS-Low groups. The results demonstrated a clinically meaningful improvement in real-world overall survival (rwOS) for the IPS-High group compared to the IPS-Low group. Furthermore, a comparison of IPS risk stratification on ICI therapy versus prior non-ICI regimen provided additional insight about IPS's utility as an ICI-specific biomarker. This hypothesis-generating data address an unmet need for patients whom an ICI therapy and predictive biomarker are urgently needed.

A detailed analysis using Tempus’ de-identified real-world database examined the relationship between androgen receptor (AR) alterations and the immune microenvironment in 1,556 patients with castration-resistant prostate cancer (CRPC). The study specifically investigated AR mutations, amplifications, and ARv7 splicing detected via DNA (Tempus xT) and RNA (Tempus xR) sequencing. Over half of the CRPC patients exhibited these AR alterations, which were associated with significantly decreased immune infiltration and reduced expression of key immunotherapy targets like PD−1, PD−L1, and CTLA−4. Regression analysis confirmed this link to decreased immune infiltration was independent of tumor mutational burden (TMB) and tumor purity. These findings suggest that AR mutations contribute to a reduced immune response, potentially serving as a mechanism of resistance to treatment, and underscore the necessity of using AR status to stratify CRPC patients for immunotherapy.

This research evaluates the prognostic value of defining an "ultrahigh" tumor mutational burden (TMB) threshold (≥40 mutations/MB) compared to the standard 10 mt/MB cutoff for patients receiving immune checkpoint inhibitor (ICI) therapy. Using Tempus Lens, the research team defined a cohort of 17,449 patients with five different cancer types (melanoma, lung, GI, non-melanoma skin, and uterine) from Tempus’ de-identified multimodal database. The analysis sought to compare real-world objective response rates (rwORR) and overall survival (rwOS) across low, high, and ultrahigh TMB groups. The findings indicate that patients in the ultrahigh TMB group experience significantly improved clinical outcomes, including enhanced rwORR and better rwOS. This ultrahigh TMB status is also linked to a distinct tumor microenvironment, specifically showing a higher degree of regulatory T cell and myeloid cell infiltration, suggesting that ultrahigh TMB may serve as a novel marker for predicting ICI responsiveness.