Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma and remains curable for many patients with frontline immunochemotherapy; however, optimal treatment continues to evolve. The Canadian Hematology Consensus Group convened a national panel of lymphoma experts to develop evidence-informed recommendations for the first-line management of adults with DLBCL in Canada. Recommendations were developed using a modified Delphi process and address frontline regimen selection, treatment intensity, central nervous system (CNS) risk assessment and prophylaxis, interim response assessment, and end of treatment management. The panel recommends Pola-R-CHP for patients with international prognostic index 2–5 and activated B-cell subtype by gene expression profiling (or if unavailable, non-germinal centre B-cell subtype by immunohistochemistry). All patients should be assessed for fitness to receive full-dose therapy, with dose attenuation considered when clinically appropriate. Routine high-dose methotrexate prophylaxis is not recommended, and CNS screening should be individualized. Interim imaging may be used to confirm response, but treatment escalation should not be based on PET positivity alone. Patients achieving complete metabolic response at end of treatment require no further therapy, and routine surveillance imaging is not recommended. These recommendations support consistent, evidence-informed, and equitable care for patients with newly diagnosed DLBCL across Canada.
This bibliometric review mapped publication growth, intellectual and conceptual structure, thematic evolution, and toxicity-related nursing and supportive care dimensions across selected cancer immunotherapies. Scopus, Web of Science Core Collection, PubMed, and CINAHL were searched for English-language articles and reviews published from 2011 to 15 August 2026. Records underwent deduplication, title-and-abstract screening, bibliometric analysis, citation and collaboration mapping, and modality-specific coding. The corpus comprised 608 publications: 340 on immune checkpoint inhibitors (ICIs), 156 on chimeric antigen receptor T-cell (CAR-T)/immune effector therapies, 40 on mixed modalities, 29 on bispecific/T-cell engagers, and 43 on other or unspecified immunotherapies. Complete-year publication growth was 39.6% annually between 2011 and 2025. ICI publications emphasized organ-specific toxicities and patient education, whereas CAR-T/immune effector publications emphasized cytokine release syndrome, neurotoxicity, and monitoring. Explicit nursing role and patient-reported outcome terms were less visible than toxicity management terms. Additional author verification confirmed the sampled classifications; this checking did not constitute independent blinded validation. The findings primarily characterize the retrieved ICI and CAR-T literature. Smaller modality groups remain exploratory because comprehensive retrieval was not established. These patterns describe bibliographic visibility and research structure, rather than clinical incidence, care quality, or validated clinical priorities.
Background: Idecabtagene vicleucel (ide-cel) has demonstrated substantial activity in relapsed/refractory multiple myeloma (RRMM), but outcomes in real-world patients remain heterogeneous. Extramedullary disease (EMD) is an adverse phenotype, and prolonged cytopenias may complicate post-CAR-T recovery. We synthesized real-world standard-of-care ide-cel evidence to characterize EMD-associated efficacy outcomes and prolonged hematotoxicity. Methods: We performed a focused evidence synthesis of real-world ide-cel studies reporting EMD-stratified outcomes, cytopenias, hematopoietic interventions, or safety in RRMM. Data were extracted into a structured workbook. Because several US datasets were derived from potentially overlapping registry or consortium populations, primary EMD prevalence and progression-free survival (PFS) synthesis was restricted to non-overlapping EMD-stratified cohorts. Progression-free survival (PFS) was analyzed in parallel across contributing cohorts rather than pooled mathematically. Cytopenia outcomes were summarized descriptively at day 30/month 1 and day 90/month 3. Results: Five real-world studies were included. Across two non-overlapping ide-cel-only EMD-stratified cohorts, 111 of 510 infused patients had reported EMD, corresponding to a combined study-reported prevalence of 21.8% (17.0% in the French registry and 23.9% in the US cohort). Median PFS was shorter with EMD in both contributing cohorts: 6.21 versus 14.82 months in FENIX/DESCAR-T and 5.3 versus 11.1 months in the US EMD-focused cohort. Across both independent cohorts, presence of extramedullary disease was consistently associated with significantly inferior progression-free survival: median PFS was 6.21 months (95% CI: 4.01–NR) versus 14.82 months (95% CI: 11.30–18.37) in the French FENIX/DESCAR-T cohort (adjusted HR 3.15, 95% CI: 1.82–5.47; p < 0.0001), and 5.30 months (95% CI: 4.10–6.90) versus 11.10 months (95% CI: 9.20–12.60) in the US cohort (adjusted HR 1.50, 95% CI: 1.10–2.20; p = 0.02). In the US EMD-focused cohort, day-90 ORR was 52% with EMD versus 82% without EMD. EMD independently predicted inferior PFS in both adjusted analyses. A multicenter cohort evaluating both ide-cel and ciltacabtagene autoleucel (cilta-cel) contributed supportive evidence, with ide-cel-treated EMD patients achieving a best overall response rate (ORR) of 61% (≥PR per IMWG criteria). In the cytopenia-focused cohort, grade ≥ 3 cytopenia was present in 64.7% at day 30 and 40.4% at day 90. Grade ≥ 3 thrombocytopenia was present at day 90/month 3 in 20.8–27.7% across ide-cel-only hematotoxicity studies. Conclusions: Real-world ide-cel data identify EMD as a high-risk phenotype associated with inferior response durability and approximately halved median PFS. Prolonged grade ≥ 3 cytopenias are common after standard-of-care ide-cel, particularly thrombocytopenia and neutropenia. These findings support pre-infusion risk stratification incorporating true EMD status, marrow disease burden, and hematopoietic reserve.
Background: How treatment recommendations generated by large language models (LLMs) compare with real-world decisions in oncology remains poorly characterized, particularly in resource-limited settings. Methods: In this single-center feasibility study, four LLMs (ChatGPT, Claude, Gemini, and Grok, accessed via consumer interfaces in June 2026) generated treatment recommendations for 27 consecutive newly diagnosed lung cancer cases treated at a Romanian oncology unit (June 2024–May 2026). Each model’s output and the treatment actually delivered were formatted identically (135 recommendations total), presented in randomized order, and scored for appropriateness (1–5) by a reviewer blinded to the source, with three quality flags. A second independent reviewer re-scored all recommendations to assess inter-rater reliability. Scores were compared using the Friedman test. Results: Mean appropriateness was highest for the real-world decision (4.52), followed by ChatGPT (4.48), Claude (4.44), Grok (4.41), and Gemini (4.37), with no significant difference (Friedman χ24 = 1.99, p = 0.74). No recommendation scored below 3. Inter-rater agreement was high in raw terms (93.3% within one point) but only fair after chance correction (weighted κ = 0.26), reflecting a ceiling effect. Conclusions: AI-generated recommendations were rated similarly to real-world decisions. The evaluation method proved feasible in a resource-limited setting; the small sample and ceiling effect preclude equivalence claims, and the findings are hypothesis-generating.
Body composition and routinely available blood biomarkers may help characterize prognosis in extensive-stage small-cell lung cancer (ES-SCLC). We retrospectively evaluated 81 patients who received first-line carboplatin, etoposide, and atezolizumab. Baseline L3 skeletal muscle index (SMI) was derived from [18F]FDG PET/CT and classified using the Martin criteria. Overall survival (OS) was assessed using Kaplan–Meier and Cox regression. Of 80 classifiable patients, 33 (41.3%) had low SMI. Median OS was 10.51 months (95% CI, 7.98–14.26) and was 8.48 versus 14.26 months in patients with low versus normal SMI (log-rank p = 0.050). In the primary multivariable model, low SMI was associated with higher mortality (HR, 2.03; 95% CI, 1.07–3.86; p = 0.030); lower albumin and higher lactate dehydrogenase were also associated with mortality. In exploratory analyses, the low-SMI association persisted after post hoc adjustment for bone metastasis (HR, 2.16; 95% CI, 1.05–4.42; p = 0.035), although diagnostics suggested time-varying effects. Continuous SMI and alternative thresholds were not associated with OS, and low SMI was not associated with progression-free survival. Martin-defined low SMI may complement routine biomarkers in prognostic assessment, but its cutoff- and time-dependent association requires cautious interpretation and external evaluation.
