JCO precision oncology最新文献

Purpose: The primary results of phase III oncology trials may be challenging to interpret, given that results are generally based on P value thresholds. The probability of whether a treatment is beneficial, although more intuitive, is not usually provided. Here, we developed and released a user-friendly tool that calculates the probability of treatment benefit using trial summary statistics.

Methods: We curated 415 phase III randomized trials enrolling 338,600 patients published between 2004 and 2020. A phase III prior probability distribution for the treatment effect was developed on the basis of a three-component zero-mean mixture distribution of the observed z-scores. Using this prior, we computed the probability of clinically meaningful benefit (hazard ratio [HR] <0.8). The distribution of signal-to-noise ratios and power of phase III oncology trials were compared with that of 23,551 randomized trials from the Cochrane Database.

Results: The signal-to-noise ratios of phase III oncology trials tended to be much larger than randomized trials from the Cochrane Database. Still, the median power of phase III oncology trials was only 49% (IQR, 14%-95%), and the power was <80% in 65% of trials. Using the phase III oncology-specific prior, only 53% of trials claiming superiority (114 of 216) had a ≥90% probability of clinically meaningful benefits. Conversely, the probability that the experimental arm was superior to the control arm (HR <1) exceeded 90% in 17% of trials interpreted as having no benefit (34 of 199).

Conclusion: By enabling computation of contextual probabilities for the treatment effect from summary statistics, our robust, highly practical tool, now posted on a user-friendly webpage, can aid the wider oncology community in the interpretation of phase III trials.

Purpose: Cisplatin-based combination chemotherapy (CHT) is standard of care in metastatic urothelial cancer (mUC); however, no predictive molecular biomarkers are available for clinical use. The aim of this study was to investigate the impact of molecular subtypes in relation to treatment response and survival in patients with mUC treated with first-line CHT.

Patients and methods: Molecular subtype classification according to the Lund Taxonomy (LundTax) was performed by tumor transcriptomic profiling and immunostaining in a retrospective cohort. Molecular subtypes were investigated in relation to the primary end point overall response rate (ORR) and secondary end points progression-free survival (PFS) and overall survival (OS). Differential gene expression and association to treatment response were explored.

Results: Ninety-five patients with mUC were classified into urothelial-like (Uro, 43%), genomically unstable (GU, 26%), basal squamous-like (Ba/Sq, 20%), mesenchymal-like (Mes-like, 8%), and small cell neuroendocrine-like (Sc/NE, 3%) subtypes. Patients with Mes-like tumors had lower ORR (14%) compared with Uro (70%), GU (77%), Ba/Sq (75%), and Sc/NE (67%; odds ratio, 0.06 [95% CI, 0.01 to 0.54], P = .012). Furthermore, patients with Mes-like tumors had significantly shorter PFS (hazard ratio [HR], 5.18 [95% CI, 2.28 to 11.76], P < .001) and OS (HR, 3.19 [95% CI, 1.45 to 7.03], P = .004). Patients with Uro and GU showed the longest survival. In responders, an enrichment of downregulated stromal- and immune-related genes was seen. Downregulation of interferon-induced transmembrane protein 2 was associated with increased ORR and improved OS.

Conclusion: This study identifies different CHT responses by LundTax molecular subtypes in patients with mUC, where the Mes-like subtype was associated with lower response rate and shorter survival.

Polyclonal convergent evolution to PARPi resistance in a patient with metastatic breast cancer with gPALB2.

Case series describing excellent outcomes for patients with dMMR GI cancer after resection of a single progressive lesion under immunotherapy.

Purpose: The clinical and research FPG500 program (ClinicalTrials.gov identifier: NCT06020625) is currently ongoing at the Fondazione Policlinico Universitario Agostino Gemelli IRCCS to tailor matched targeted therapies (MTTs) according to biomarkers predictive of response identified by comprehensive genome profiling (CGP).

Materials and methods: The non-small cell lung cancer (NSCLC) cohort results from the FPG500 program are outlined. CGP was performed by TruSight Oncology 500 High Throughput (TSO500HT) assay or Oncomine Focus Assay plus Archer's FusionPlex Lung Panel according to tumor cell content and DNA/RNA quantity. Relevant issues for Molecular Tumor Board (MTB) evaluation included uncommon genomic findings, evaluation for off-label therapies, uncertain result confirmation, and variants of suspect germline origin requiring genetic counseling. Progression-free survival (PFS) and overall survival (OS) for the enrolled patients were assessed using Kaplan-Meier analysis.

Results: In 2022, 283 patients with NSCLC were considered for sequencing, with 93% meeting eligibility criteria. TSO500HT sequencing was conducted in 76% of patients. Follow-up data were obtained for 187 patients, among whom 81% received treatment. Potential driver alterations were identified in 59% of patients, with 41% receiving MTT: 25% were prescribed approved MTTs, whereas 16% gained access to experimental drugs post-MTB evaluation; of note, 18% did not receive any MTT because the regimen was not yet reimbursed in our country. Median PFS and OS varied among treatment groups, with standard chemotherapy/immunotherapy at 7.7 and 10.7 months, approved tyrosine kinase inhibitors at 18.8 and 23.9 months, and MTT post-MTB discussion at 14 and 23.4 months, respectively.

Conclusion: The early data of the FPG program (NSCLC cohort) support the implementation of CGP and MTB in clinical practice to grant access to patients harboring actionable molecular alterations to the most effective and individualized available treatment options, thus improving their survival outcomes.

Purpose: Individuals with Lynch syndrome (LS) are at a high lifetime risk of colorectal cancer (CRC) and other cancers. Aspirin (ASA), a nonsteroidal anti-inflammatory drug (NSAID), has proven chemopreventive benefits in LS, with the CAPP2 randomized double-blind placebo-controlled trial demonstrating a 60% relative risk reduction for CRC among participants who adhered to ASA for 2 years or more. This study sought to characterize uptake of ASA/NSAIDs among individuals with LS and to understand factors associated with use.

Methods: Individuals with LS were invited (June 2020-August 2022) to complete a one-time electronic survey about LS screening behaviors, uptake of ASA/NSAIDs, and current/emerging cancer prevention options. Participants were recruited from the Fox Chase Cancer Center (FCCC) Risk Assessment Program Registry and through a research invitation posted to two patient-facing LS advocacy websites.

Results: Two hundred and ninety-six participants completed the survey including 116 (39.2%) from FCCC and 180 (60.8%) recruited via the Internet, including 14.9% non-US based individuals. Uptake of regular ASA or NSAIDs was modest at 34.8% and was even lower (25.7%) when focusing on individuals taking ASA or NSAIDs solely for chemoprevention of LS. More than half (55%) were taking <100 mg ASA daily. In multivariable modeling, lower perceived threat of LS (odds ratio [OR], 0.84 [95% CI, 0.72 to 0.98]), lower concern for side effects (OR, 0.86 [95% CI, 0.76 to 0.99]), and higher likelihood of recommending ASA/NSAIDs to family or a friend were all associated with ASA/NSAIDs use (OR, 1.70 [95% CI, 1.37 to 2.10]).

Conclusion: Uptake of ASA/NSAIDs chemoprevention is modest among individuals with LS. Patient perceptions of the pros and cons of ASA, more so than demographic and disease-related factors, were associated with chemoprevention uptake.

Purpose: Rapidly expanding medical literature challenges oncologists seeking targeted cancer therapies. General-purpose large language models (LLMs) lack domain-specific knowledge, limiting their clinical utility. This study introduces the LLM system Medical Evidence Retrieval and Data Integration for Tailored Healthcare (MEREDITH), designed to support treatment recommendations in precision oncology. Built on Google's Gemini Pro LLM, MEREDITH uses retrieval-augmented generation and chain of thought.

Methods: We evaluated MEREDITH on 10 publicly available fictional oncology cases with iterative feedback from a molecular tumor board (MTB) at a major German cancer center. Initially limited to PubMed-indexed literature (draft system), MEREDITH was enhanced to incorporate clinical studies on drug response within the specific tumor type, trial databases, drug approval status, and oncologic guidelines. The MTB provided a benchmark with manually curated treatment recommendations and assessed the clinical relevance of LLM-generated options (qualitative assessment). We measured semantic cosine similarity between LLM suggestions and clinician responses (quantitative assessment).

Results: MEREDITH identified a broader range of treatment options (median 4) compared with MTB experts (median 2). These options included therapies on the basis of preclinical data and combination treatments, expanding the treatment possibilities for consideration by the MTB. This broader approach was achieved by incorporating a curated medical data set that contextualized molecular targetability. Mirroring the approach MTB experts use to evaluate MTB cases improved the LLM's ability to generate relevant suggestions. This is supported by high concordance between LLM suggestions and expert recommendations (94.7% for the enhanced system) and a significant increase in semantic similarity from the draft to the enhanced system (from 0.71 to 0.76, P = .01).

Conclusion: Expert feedback and domain-specific data augment LLM performance. Future research should investigate responsible LLM integration into real-world clinical workflows.