Immunotherapy最新文献

Background: First-line pembrolizumab plus chemotherapy has demonstrated durable, clinically meaningful survival benefits over 5 years, compared with chemotherapy alone, in pivotal clinical trials for patients with metastatic NSCLC. This retrospective study aimed to evaluate 5-year real-world outcomes with pembrolizumab plus chemotherapy at US oncology practices.

Methods: Using a nationwide, deidentified database, we selected adults with unresectable stage IIIB/IIIC/IV NSCLC and ECOG performance status of 0-1, excluding those in a clinical trial, who initiated first-line pembrolizumab plus platinum/pemetrexed from 1 June 2017 through 30 September 2021 (EGFR/ALK-wild-type nonsquamous NSCLC) or pembrolizumab plus carboplatin/(nab)-paclitaxel from 1 November 2018 through 30 September 2020 (squamous NSCLC). Overall survival (OS) from first-line initiation, by histology and PD-L1 expression, was estimated using the Kaplan-Meier method. Data cutoff was 30 September 2024.

Results: Median study follow-up was 60 months. Median (95% CI) OS was 15.0 months (13.4-16.0) and 12.9 months (10.3-17.1) among 1960 patients with nonsquamous and 433 with squamous NSCLC, respectively. At 5 years, OS rates were 21.6% and 18.2%, respectively, with 5-year OS rates by tumor PD-L1 < 1%/1-49%/≥50% expression of 15.8%/19.8%/32.6% in nonsquamous and 15.2%/14.1%/32.6% in squamous cohorts.

Conclusions: First-line pembrolizumab plus chemotherapy demonstrates long-term effectiveness for nonsquamous and squamous advanced/metastatic NSCLC, with 5-year OS rates in real-world settings that are consistent across PD-L1 expression strata with 5-year outcomes from the pivotal clinical trials.

Chimeric antigen receptor (CAR) natural killer (NK) cell therapy represents an emerging frontier in cancer immunotherapy. CAR-NK cells offer the potential for "off-the-shelf" treatments with a favorable safety profile, due in part to their innate capacity for HLA-unrestricted tumor targeting and an absence of graft-versus-host disease. Landmark clinical studies using cord blood - derived CAR-NK cells and another employing iPSC-derived CAR-NK cells have reported impressive response rates, durable remissions, and minimal toxicity. Furthermore, a growing landscape of ongoing clinical trials is exploring innovative strategies to enhance persistence, overcome solid tumor challenges, and prevent antigen escape. These advances collectively underscore the potential of CAR-NK therapies to revolutionize cancer treatment while addressing limitations associated with current CAR-T cell approaches.

Emerging evidence demonstrates the potential efficacy of chimeric antigen receptor (CAR) T cells in autoimmune diseases. Currently, this therapy relies primarily on patient-derived autologous CAR-T cells. However, the application of autologous CAR-T cells is hindered by several limitations, such as high costs, time-consuming, and vulnerability to failure during the manufacturing process. Allogeneic CAR-T cells are genetically engineered T cells derived from healthy donors that hold great promise for expanding the accessibility of CAR-T cell therapy and may overcome the limitations of autologous CAR-T cells. Recent studies with small sample sizes have suggested the potential of allogeneic CAR-T cells in autoimmune diseases. These results have motivated researchers and physicians to further investigate the application of allogeneic CAR-T cells. In this review, we discuss the development of allogeneic CAR-T cells, current studies on CAR-T cell therapy for autoimmune diseases, and potential toxicities of allogeneic CAR-T cells. In addition, we discuss current challenges and future directions for implementing this therapy in clinical practice.

Background: Immune checkpoint inhibitors (ICIs) are used to treat various cancers. Serotonin receptors are expressed by immune cells. In preclinical models, serotonergic drugs affect T-cell cytokine production, proliferation and apoptosis. Interactions between clinical serotonergic drug use and dual ICI treatment remain unknown.

Methods: Individual patient data were pooled from 4 Canadian Cancer Trials Group (CCTG) trials of patients treated with dual ICI ± chemotherapy (n = 684). Serotonergic drug use was correlated with clinicopathologic characteristics, best overall response (BOR)/iBOR per RECIST 1.1/iRECIST, progression-free survival (PFS)/iPFS, overall survival (OS) and immune-related adverse events (irAEs) using Cochran - Mantel - Haenszel and log-rank tests.

Results: Eighty-three (12%) patients used serotonergic drugs at baseline and 118 (17%) at any time on trial. By multivariate analysis, serotonergic drug use at baseline was significantly associated with decreased iBOR (p = 0.04), but not PFS (p = 0.21), iPFS (p = 0.28), OS (p = 0.30) or incidence of grade 1/2 or 3/4 irAE (p = 0.85 and 0.99 respectively). Results were not significantly different with serotonergic drug use at any time on trial.

Conclusion: Use of serotonergic drugs did not impact PFS or OS in patients treated with dual ICI ± chemotherapy. This study supports the safe use of serotonergic drugs in the context of dual ICI therapy.

Background: Multidisciplinary teams (MDTs) are fundamental to cancer care but face increasing burdens. This pilot study evaluates a large language model (LLM) simulating a tumor board for a clinically complex cohort of non-small cell lung cancer (NSCLC) patients, using a full-context guideline injection methodology to ground its reasoning in authoritative standards.

Methods: Ten real-world NSCLC cases were presented to Google's Gemini 2.5 Pro using prompt engineering. The model was primed by providing the complete National Cancer Institute guidelines as in-context source data in a structured JSON file. AI-generated recommendations were scored against the institutional human MDT.

Results: The LLM demonstrated high performance, achieving mean scores of 4.9/5.0 for content accuracy, 5.0/5.0 for internal consistency, and 4.4/5.0 for clinical applicability. Importantly, no safety concerns were identified in the AI's recommendations. However, the model did not generate any novel insights beyond those considered by the human MDT.

Conclusions: An LLM primed with comprehensive guidelines can accurately and safely replicate MDT recommendations for complex NSCLC cases. The combination of guideline injection and meticulous prompt engineering is a critical strategy for ensuring LLM reliability. This positions these models as powerful decision-support tools to augment, not replace, expert clinical workflow.