Biology subjects

Besse, B.

Publications and source records attributed to Besse, B..

4 recordsLinked to original sources

bioRxiv2026

SHERLOC: An interpretable deep learning model for longitudinal circulating tumor DNA data in survival analysis

Longitudinal circulating tumor DNA (ctDNA) measurements offer a noninvasive means to monitor treatment response, but clinical trial data present substantial methodological challenges due to high-dimensional short longitudinal ctDNA sequences and limited sample sizes. We introduce SHERLOC, a deep learning framework specifically designed for survival analysis using longitudinal on-treatment ctDNA data, which integrates shared temporal representations of gene-level variant allele frequencies, feature-specific temporal trajectories of panel-level ctDNA biomarkers, and survival-aware genomic representations pre-trained on a large pan-cancer tissue-biopsy dataset (MSK-CHORD), within an interpretable Cox proportional hazards framework. Benchmarked against diverse statistical, ensemble, and deep learning approaches in a non-small-cell lung cancer cohort from the phase III IMpower150 trial, SHERLOC consistently achieved superior survival discrimination and calibration, while remaining interpretable and robust to reductions in the number of available longitudinal liquid biopsy time points per patient. The resulting ctDNA-based risk score provided prognostic information both independent of and complementary to standard radiographic response assessments, and enabled patient stratification within homogeneous RECIST response groups--highlighting its potential as an early, non-invasive decision-support tool to guide treatment adaptation and patient management.

cancer biology↗

bioRxiv2026

Single-cell trajectories in metastatic urothelial carcinoma reveal tumor-immune reprogramming and macrophage-driven resistance to PD-(L)1 blockade

Immune checkpoint inhibitors (ICI) improved outcomes in metastatic urothelial carcinoma (mUC), but primary and acquired resistance remain poorly understood. We performed single-nuclei RNA sequencing on sequential metastatic biopsies from ICI-treated mUC patients. Tumor cells showed transcriptomic heterogeneity within individual lesions, basal cells being associated with increased immune infiltration and response. Myeloid and lymphoid compartments exhibited features of immune dysfunction in non-responders. Longitudinal analyses revealed convergent adaptive resistance mechanisms, dominated by polarization toward pro-tumoral macrophage states, but also including downregulation of the antigen presentation machinery in tumor cells, increased checkpoint expression with loss of cytotoxicity in T cells. Individual trajectories point to distinct evolutionary routes under ICI pressure. Across pivotal ICI trials, bulk expression of the M2-like macrophage marker HES1 predicted ICI resistance. Our study provides the first single-cell longitudinal atlas of ICI-treated mUC, revealing macrophage reprogramming as a dominant driver of resistance, establishing a framework for individualized immunotherapy strategies.

cancer biology↗

bioRxiv2025

Single- and double-strand circulating DNA fragmentomics for enhanced cancer detection performance

In early detection of cancer, the use of circulating cell-free DNA (cirDNA) obtained from blood samples is notable for its minimally invasive nature. We have developed an algorithm designed to discriminate cancer patients and healthy individuals based on cirDNA fragment end motif analysis assisted by machine learning, using data obtained from shallow whole genome sequencing (a method we call EMA). We applied EMA to cirDNA from the plasma of patients with stage II-III breast cancer, stage I-III non-small cell lung cancer, and metastatic colorectal cancer (mCRC). CirDNA from 158 individuals was prepared following the conventional double-stranded DNA library preparation (DSP). Using 3 bp end motifs, each tumor type was detected with a sensitivity of 0.87-1.00, a specificity of 0.95, and an AUC above 0.96. The three selected cancer types could be differentiated with an accuracy (ACC) above 0.94. Multi-cancer detection by pooling samples from the three cancer types showed ACC, AUC and sensitivity of 0.98, 0.99 and 0.98, respectively. Comparisons with 4 and 2 bp end motifs were conducted, and our main observations were confirmed using an external public dataset (N=366). We also performed a single-stranded DNA library preparation (SSP) using mCRC patients and healthy control cirDNA, which allowed us to make the first ever end motif analysis in the literature which compares the use of DSP and SSP. As compared to EMAD (use of DSP), EMAS (use of SSP) produced a very significant difference in end motif frequency and an improved cancer detection performance (ACC, AUC and sensitivity of 0.97, 1.00 and 0.99, respectively). Furthermore, optimal performance was produced when the full-size range was used for EMAS, whereas when the dataset was restricted to fragments of 115 - 220 bp for EMAD. EMAS strong performance, coupled with its compatibility with cost-effective shallow cirDNA sequencing, positions this methodology as a potentially transformative tool in early cancer screening.

genomics↗

bioRxiv2023

The PI3K-AKT-mTOR axis persists as a therapeutic dependency in KRASG12D-driven non-small cell lung cancer

IntroductionKRASG12C and KRASG12D inhibitors represent a major translational breakthrough for non-small cell lung cancer (NSCLC) and cancer in general by directly targeting its most mutated oncoprotein. However, resistance to these small molecules has highlighted the need for rational combination partners necessitating a critical understanding of signaling downstream of KRAS mutant isoforms. MethodsWe contrasted tumor development between KrasG12Cand KrasG12D genetically engineered mouse models (GEMMs). To corroborate findings and determine mutant subtype-specific dependencies, isogenic models of KrasG12C and KrasG12D initiation and adaptation were profiled by RNA sequencing. We also employed cell line models of established KRAS mutant NSCLC and determined therapeutic vulnerabilities through pharmacological inhibition. We analysed differences in survival outcomes for patients affected by advanced KRASG12C or KRASG12D-mutant NSCLC. ResultsKRASG12D exhibited higher potency in vivo, manifesting as more rapid lung tumor formation and reduced survival of KRASG12D GEMMs compared to KRASG12C. This increased potency, recapitulated in an isogenic initiation model, was associated with enhanced PI3K-AKT-mTOR signaling. However, KRASG12C oncogenicity and downstream pathway activation were comparable with KRASG12D at later stages of tumorigenesis in vitro and in vivo, consistent with similar clinical outcomes in patients. Despite this, established KRASG12D NSCLC models depended more on the PI3K-AKT-mTOR pathway, while KRASG12C models on the MAPK pathway. Specifically, KRASG12D inhibition was synergistically enhanced by AKT inhibition. ConclusionsOur data highlight a unique combination treatment vulnerability and suggest that patient selection strategies for combination approaches using direct KRAS inhibitors should be i) contextualised to individual RAS mutants, and ii) tailored to their downstream signaling.

cancer biology↗