Gu Y, Fu J, Liu X
… +29 more, Valanarasu JMJ, Codella NCF, Tan R, Liu Q, Jin Y, Zhang S, Wang J, Wang R, Song L, Qin G, Usuyama N, Wong C, Cheng H, Lee H, Sanapathi P, Hilado S, Naumann T, Alvarez-Valle J, Bian J, Wei M, Malik K, Zhou L, Gao J, Horvitz E, Lungren MP, Burger D, Topol E, Poon H, Vozila P
Large frontier models such as GPT-5 and Gemini have demonstrated remarkable performance in a wide range of health application benchmarks. However, underneath the seemingly promising results lie salient growth areas, espe...Large frontier models such as GPT-5 and Gemini have demonstrated remarkable performance in a wide range of health application benchmarks. However, underneath the seemingly promising results lie salient growth areas, especially in cutting-edge frontiers such as multimodal reasoning. Here we systematically apply and integrate a series of adversarial stress tests to assess the robustness of flagship models and health benchmarks. Our study reveals prevalent brittleness in the presence of simple adversarial transformations: leading systems can guess the correct answer even with key inputs removed yet may get confused by the slightest prompt alterations while fabricating convincing but flawed reasoning traces. Using clinician-guided rubrics, we demonstrate that popular health benchmarks vary widely in what they truly measure. Our study reveals considerable gaps between benchmark performance and the robustness evidence needed to support claims about multimodal medical reasoning in health applications.
Immune-checkpoint inhibitors benefit a subset of patients with advanced cancer, and the metabolic determinants of response remain unclear. Here, using targeted metabolomics and metagenomics, we profiled 4,336 plasma samp...Immune-checkpoint inhibitors benefit a subset of patients with advanced cancer, and the metabolic determinants of response remain unclear. Here, using targeted metabolomics and metagenomics, we profiled 4,336 plasma samples from 1,714 patients across five tumor types and 16 cohorts spanning Europe and North America, longitudinally sampled during five immune-checkpoint inhibitor-based treatment modalities, including fecal microbiota transplantation. A multimodal machine-learning framework integrating 154 metabolites with clinical variables identified five metabolites, age, body mass index and renal function as predictors of 12-month progression-free survival. The model achieved areas under the curve of 0.88 in training and 0.73 in validation cohorts of 105 and 30 patients, respectively and generalized across seven external cohorts. Histidine was a favorable prognostic feature of survival, whereas long-chain fatty acids and succinate were negatively associated with outcome. Histidine supplementation enhanced antitumor immunity in mice. Histidine-rich diets improved progression-free survival in patients lacking dysbiotic microbiome signatures associated with histidine catabolism.
De Caluwé A, Desmoulins I, Cao K
… +22 more, Remouchamps V, Baten A, Longton E, Peignaux K, Joaquin Garcia A, Venet D, Arecco L, Agostinetto E, Nader-Marta G, Denis Z, Dhont J, Kristanto P, Catteau X, Larsimont D, Salgado R, Poortmans P, Stagg J, Sotiriou C, Piccart M, Ignatiadis M, Romano E, Buisseret L
Patients with estrogen receptor-positive (ER), HER2-negative, early breast cancer (BC) have low pathologic complete response (pCR) rates following neoadjuvant chemotherapy. Immune checkpoint inhibitors (ICIs) provide lim...Patients with estrogen receptor-positive (ER), HER2-negative, early breast cancer (BC) have low pathologic complete response (pCR) rates following neoadjuvant chemotherapy. Immune checkpoint inhibitors (ICIs) provide limited benefit in programmed death-ligand 1 (PD-L1)-negative tumors, characterized by an immune-cold tumor microenvironment. Here we hypothesized that immune-modulating stereotactic body radiation therapy (iSBRT; 3 × 8 Gy) could enhance response through tumor microenvironment reprogramming, and that CD73 blockade could further improve efficacy. We conducted a phase 2, randomized, multicenter trial (Neo-CheckRay) in 147 female patients with high-risk, ERHER2 early BC. Patients received neoadjuvant chemotherapy plus iSBRT alone (No_ICI), with anti-PD-L1 durvalumab (Single_ICI) or with durvalumab plus anti-CD73 oleclumab (Double_ICI). In the intention-to-treat population, the primary endpoint, residual cancer burden 0/1 rate, was 35.4% with No_ICI, 45.1% with Single_ICI and 47.9% with Double_ICI, without statistically significant differences. pCR rates were 16.7%, 29.4% and 33.3%, respectively (P = 0.059). In the per-protocol population (MammaPrint High Risk, n = 131), pCR rates were 16.3%, 32.6% and 35.6%, respectively (P = 0.040). Among PD-L1-negative tumors (n = 91), pCR rates were 3.4%, 28.1% and 30.0%, respectively. No new safety signals were observed. Baseline transcriptomic analysis showed low immune signature expression in PD-L1-negative tumors. Paired baseline and on-treatment biopsies obtained 1 week after iSBRT demonstrated tumor microenvironment reprogramming toward an inflamed phenotype in the iSBRT + anti-PD-L1 arms. These findings suggest that iSBRT + anti-PD-L1 may convert immune-cold ERHER2 BC into more inflamed tumors and improve response, particularly in PD-L1-negative disease. ClinicalTrials.gov registration: NCT03875573 .
Advancements in frontline therapies have substantially improved outcomes in newly diagnosed multiple myeloma (NDMM); however, many patients will not achieve deep responses and will relapse. Teclistamab, a BCMA×CD3 bispec...Advancements in frontline therapies have substantially improved outcomes in newly diagnosed multiple myeloma (NDMM); however, many patients will not achieve deep responses and will relapse. Teclistamab, a BCMA×CD3 bispecific antibody, in combination with daratumumab, has demonstrated strong efficacy in relapsed/refractory multiple myeloma versus standard of care as early as first relapse. This ongoing phase 2 GMMG-HD10/DSMM-XX (MajesTEC-5) study evaluates teclistamab-based regimens in transplant-eligible NDMM. In this prespecified pooled analysis of three cohorts, 49 patients received teclistamab/daratumumab/lenalidomide (Tec-DR; arms A and A1) or Tec-DR with bortezomib (Tec-DVR; arm B). Primary endpoints were incidence and severity of adverse events (AEs) and serious AEs; secondary endpoints included overall response rate (ORR), minimal residual disease (MRD) negativity and MRD-negative complete response (CR). The current analysis spans the induction and autologous stem cell transplantation phases until the premaintenance timepoint. Grade 3 or 4 treatment-emergent AEs (TEAEs) occurred in 91.8% (45/49); most were hematologic (lymphopenia (59.2%; 29/49), neutropenia (59.2%; 29/49) and leukopenia (18.4%; 9/49)). No grade 5 TEAEs were reported. Serious AEs occurred in 55.1% (27/49); pyrexia (12.2% (6/49)) was most common. Any-grade and grade 3 or 4 infections occurred in 81.6% (40/49) and 36.7% (18/49), respectively, the most common grade 3 or 4 infections being COVID-19 and pneumonia (6.1% (3/49) each). Cytokine release syndrome occurred in 67.3% (33/49); all were grade 1 or 2, all resolved and none led to discontinuation of any study treatment. No treatment-related immune effector cell-associated neurotoxicity syndrome (ICANS) events occurred. Across arms, the MRD-negative CR rate was 91.8% (45/49) by the premaintenance timepoint; the MRD negativity rate was 100% in evaluable samples at postinduction cycle 3 (1 × 10 (46/46)), cycle 6 (1 × 10 (46/46) and 1 × 10 (46/46)) and premaintenance (1 × 10 (40/40)); the ORR was 100% (49/49). Total median stem cell yield was 8.1 × 10 per kg. Data support the feasibility of Tec-D(V)R induction in transplant-eligible NDMM, with a consistent safety profile compared with individual regimen components and notable early MRD negativity rates. ClinicalTrials.gov identifier: NCT05695508 .
Rosenstock J, Lingvay I, Ryan D
… +16 more, Jastreboff AM, Kushner R, Acosta A, Blevins TC, Choi M, Holmes FL, Smith T, Connery L, Li Y, Liu MK, Barth A, Butcher J, Civitarese A, Yue H, Coll B, ACCESS Trial Investigators
Huang Z, Zeng L, Ruan Z
… +25 more, Zeng Q, Yan H, Jiang W, Xiong Y, Zhou C, Yang H, Liu L, Dai J, Zou N, Xu S, Wang Y, Wang Z, Deng J, Chen X, Wang J, Xiang H, Li X, Duchemann B, Chen G, Xia Y, Mok T, Scheiermann C, Lévi F, Yang N, Zhang Y
Fallah MP, Wayengera M, Adewole IF
… +24 more, Dereje N, Temfack E, Bamatura M, Tabaro CA, Shaweno T, Duga A, Kuonza F, Tekle BI, Kisera N, Mesfin AA, Sayed Osman SO, Wasunna M, Nyansulu PS, Ogutu B, Adu-Sarkodie Y, N'jai AU, Rashid A, Nyan DC, Mangu C, Raji T, Mbala P, Muyembe JJ, Sewankambo N, Kaseya J
Reanalysis of genomic data in rare disease is highly effective in increasing diagnostic yields but remains limited by manual approaches. Automation and optimization for high specificity will be necessary to ensure scalab...Reanalysis of genomic data in rare disease is highly effective in increasing diagnostic yields but remains limited by manual approaches. Automation and optimization for high specificity will be necessary to ensure scalability, adoption and sustainability of iterative reanalysis. We developed Talos, an open-source tool that automates variant prioritization by integrating dynamically updated gene-disease and variant-level evidence with inheritance-aware filtering and validated its performance using data from 1,089 individuals with rare disease. Trio-based analysis identified 90% of known diagnoses, returning 1.3 variants per case on average. Variant burden reduced to one variant per 200 cases on iterative monthly reanalysis. Application to an unselected cohort of 4,735 undiagnosed individuals identified 241 diagnoses (5.1% yield): 78 (32%) due to new gene-disease relationships, 54 (22%) due to new variant-level evidence and 109 (45%) due to improved analysis strategies. Our automated, iterative reanalysis model demonstrates the feasibility of delivering frequent, systematic reanalysis at scale.
Hartman HS, Finer J, Hartzel D
… +14 more, Kelsey C, Long A, Rocha D, Ruhl JA, vanMaanen D, Malta PP, Castillo M, Roedan Oliver FA, Daniels B, Hughes JW, Leon MB, Probst MA, Poterucha TJ, Elias P
Incidence of early-onset cancer is rising globally in recent generations, which underscores the need to elucidate the influence of emerging generational risk factors. Systemic and organ-specific aging reflects the cumula...Incidence of early-onset cancer is rising globally in recent generations, which underscores the need to elucidate the influence of emerging generational risk factors. Systemic and organ-specific aging reflects the cumulative impact of exposures and may provide an integrative and complementary approach to understand early-onset cancer risk. Here among 154,169 young adults from the United Kingdom Biobank, systemic aging measured by PhenoAge increased across birth cohorts, with 23% s.d. increase for those born 1965-1974 versus 1950-1954, and was associated with early-onset solid cancer risk (hazard ratio (HR) 1.08; 95% confidence interval (CI), 1.03-1.13), driven by lung, gastrointestinal and uterine cancers, independent of genetic risks of aging and cancer. Patterns were consistent using alternative systemic aging measures, including the Klemera-Doubal method-defined age gap and metabolomic-based age gap. These findings were validated partially among 10,262 participants in the United States All of Us Research Program. Proteomics-based organ-specific aging analyses linked immune aging with early-onset lung cancer (HR 1.89; CI, 1.20-2.97) and adipose tissue aging to early-onset colorectal cancer (HR 1.60; CI, 1.11-2.32). Greater age gap, reflecting more advanced biological aging relative to chronological age, may serve as a driver associated with risk of early-onset solid cancers, highlighting the importance of uncovering underlying mechanisms to guide effective prevention strategies.
Wermke M, Ochsenreither S, Jaeger D
… +24 more, Becker H, Bleckmann A, Bozorgmehr F, Chatterjee M, Groepper S, Haenel M, Hecker JS, Häring MF, Heudobler D, Hilf N, Hofmann M, Hutt M, Mayer-Mokler A, Missel S, Ruh M, Schuster H, Veremchuk O, Kleemiss M, Knop S, Laban S, Sebastian M, Spoerl S, Britten CM, Reinhardt C