Background: SCLC remains an aggressive malignancy with limited therapeutic progress over the past few decades. It remains unclear how the sequence of immunotherapy initiation influences overall survival (OS) in SCLC; we performed a population-based analysis using NCDB to evaluate its association with survival. Methods: SCLC patients in NCDB from 2016 to 2021 were identified to evaluate the impact of immunotherapy on OS. Among ES-SCLC patients, we conducted subsequent analyses to clarify the relationship between the sequence of immunotherapy initiation and OS in the context of chemotherapy and CRT. Results: Among 69,820 eligible patients, 9242 received CRT plus immunotherapy (CRT + IO), and 11,755 received chemotherapy plus immunotherapy (Chemo + IO). In the overall population, adding immunotherapy to chemotherapy or CRT was associated with modestly improved survival. In ES-SCLC, immunotherapy was associated with longer survival in both the Chemo and CRT cohort, while the addition of immunotherapy did not confer benefits in limited-stage SCLC (LS-SCLC). Within the ES-SCLC Chemo + IO cohort, altering the initiated immunotherapy interval (0–90 days) did not show any meaningful difference in survival. By contrast, in the CRT + IO cohort, survival showed benefit in a time-dependent pattern: patients who initiated immunotherapy within 4–7 days after CRT had a trend of survival, which was consistent with the proposed immune activation window. Conclusions: This real-world analysis suggests that immunotherapy was associated with longer survival in ES-SCLC, CRT + IO is associated with improved OS, with a more obvious survival benefit when immunotherapy is initiated within 4–7 days after CRT. These findings hint at the potential importance of immunotherapy initiation sequence and warrant further prospective validation.
Current methods for pathology image segmentation typically treat 2D slices independently, ignoring valuable cross-slice information. We present PathSeqSAM, a novel approach that treats 2D pathology slices as sequential video frames using SAM2's memory mechanisms. Our method introduces a distance-aware attention mechanism that accounts for variable physical distances between slices and employs LoRA for domain adaptation. Evaluated on the KPI Challenge 2024 dataset for glomeruli segmentation, PathSeqSAM demonstrates improved segmentation quality, particularly in challenging cases that benefit from cross-slice context. We have publicly released our code at https://github.com/JackyyyWang/PathSeqSAM.