BACKGROUND: Antibiotic therapy (ABT)-induced dysbiosis may affect the efficacy of immune checkpoint inhibitors (ICI) therapy. We investigated the association between ABT and real-world overall survival (rwOS) and progression-free survival (rwPFS) in patients with metastatic urothelial carcinoma (mUC) receiving ICI or cisplatin-based chemotherapy (CIS). METHODS: Three thousand, one hundred seventy-nine patients were included from a nationwide electronic health record-derived de-identified database. Three-month landmark Kaplan-Meier methods and log-rank tests were used to estimate rwOS/PFS between treatment modalities based on ABT groups (stratified by exposure, timing, excretion mode, and administration route). Cox proportional models with time-varying coefficients were used to investigate the associations between ABT, treatment modality, and rwOS/PFS. RESULTS: A total of 402 (27.1%) ICI and 655 (38.6%) CIS patients received ABT (p < 0.001). ICI receipt (OR 0.65, p < 0.001) and advanced age (OR 0.98, p < 0.001) were associated with lower ABT use. ICI exclusive findings included a negative correlation with rwOS in patients who received post-treatment initiated (ICI median: pre-13.2 vs post-7.9 vs none-13.3 months; p = 0.009), oral (median oral-9.6 vs none-13.3 months, p = 0.03), and renally cleared (median renal-9.9 vs none-13.3 months, p = 0.04) ABT. ABTs effect was negatively associated with rwOS in ICI patients within first 6 months (HR 1.36, 95% CI 1.107-1.74, p = 0.01) but not thereafter (p = 0.7). CONCLUSIONS: This study identified a potential ICI-specific negative correlation between ABT and rwOS in patients with mUC, specifically those exposed to ABT pills and receipt before treatment initiation. These results emphasize the importance of antibiotic stewardship and continued investigation of the role of gut microbiome in mUC treatment efficacy.

Red fluorescent proteins are useful as morphological markers in neurons, often complementing green fluorescent protein-based probes of neuronal activity. However, commonly used red fluorescent proteins show aggregation and toxicity in neurons or are dim. We report the engineering of a bright red fluorescent protein, Crimson, that enables long-term morphological labeling of neurons without aggregation or toxicity. Crimson is similar to mCherry and mKate2 in fluorescence spectra but is 100 and 28% greater in molecular brightness, respectively. We used a membrane-localized Crimson-CAAX to label thin neurites, dendritic spines and filopodia, enhancing detection of these small structures compared to cytosolic markers.