Clinical Publications & Research Output

Background: AZT-4092, a novel bispecific T-cell engager targeting PD-L1 and EGFR, has demonstrated unprecedented progression-free survival in early-phase studies. This Phase III trial evaluated efficacy and safety in treatment-naïve metastatic non-small cell lung cancer (NSCLC) patients compared to standard platinum-doublet chemotherapy. Methods: 1,240 patients across 89 sites were randomized 1:1 to receive AZT-4092 IV every 3 weeks or investigator's choice chemotherapy. Primary endpoint was PFS per blinded independent central review. Results: Median PFS was 14.2 months vs 7.8 months (HR 0.41, 95% CI 0.35–0.48; p<0.001). Objective response rate was 68% vs 42%. Grade ≥3 adverse events were comparable between arms. Conclusion: AZT-4092 significantly improves PFS and OS in metastatic NSCLC with a manageable safety profile, supporting regulatory submission and potential standard-of-care adoption.

Background: Digital therapeutic platforms integrating wearable ECG, AI-driven risk stratification, and telemedicine have shown promise in AF management. This registry evaluates real-world outcomes of the ZenthCardio™ platform across 12 international health systems. Methods: Retrospective cohort of 8,432 patients enrolled between 2022–2024. Primary endpoint: composite of stroke, cardiovascular hospitalization, or death at 18 months. Results: Platform utilization reduced composite endpoint by 28% (HR 0.72, 95% CI 0.65–0.80). Anticoagulant adherence improved from 64% to 89%. No increase in major bleeding observed. Conclusion: Integrated digital cardiovascular monitoring significantly improves secondary prevention outcomes in AF patients, supporting scalable deployment in public and private health networks.

Background: AZT-NEU-77 is a small-molecule modulator of TREM2 and microglial phenotypic switching. This study elucidates its neuroprotective mechanisms via PET imaging, CSF biomarkers, and digital cognitive endpoints. Methods: 342 mild-to-moderate AD patients randomized to AZT-NEU-77 or placebo. 24-month follow-up with serial [18F]PI-2620 PET, amyloid/tau CSF assays, and validated digital cognitive batteries. Results: Significant reduction in neuroinflammation (TREM2-PET SUVr: -18.4%, p<0.001). Stabilization of MMSE scores (-1.2 vs -4.8, p<0.001). CSF p-tau217 decreased by 22% in treatment arm. Conclusion: AZT-NEU-77 demonstrates disease-modifying potential via microglial modulation and tau pathway suppression, marking a paradigm shift in neurodegenerative therapeutics.

Background: Prime editing enables precise correction of pathogenic mutations without double-strand breaks. This first-in-human trial evaluates intravitreal delivery of PE3-RET for RHO and CRB1 mutations causing Leber congenital amaurosis and RP. Methods: Open-label, dose-escalation (n=18). Primary endpoint: safety up to 12 months. Secondary: visual acuity, ERG amplitude, and OCT structural changes. Results: No grade ≥3 ocular toxicity. Mean best-corrected VA improved by 6.5 lines. ERG amplitudes stabilized in 78% of patients. No off-target edits detected in ocular tissue biopsies. Conclusion: Intravitreal prime editing is safe and clinically meaningful in inherited retinal disease, paving the way for broader genomic medicine applications.

Background: Early prediction of sepsis progression remains a critical ICU challenge. We developed ZenthSepsisAI, a multimodal transformer integrating longitudinal EHR, proteomics, and microbiome data. Methods: Training on 1.2M ICU admissions. External validation across 14 global health systems (n=284,000). Primary metric: AUC for 48-hour progression to severe sepsis or MODS. Results: AUC 0.94 vs 0.78 for traditional SOFA/qSOFA. Time-to-alert: 6.2 hours earlier than clinical recognition. Calibration remained stable across demographics. Implementation reduced 30-day mortality by 14% in prospective rollout. Conclusion: Multi-omics AI models significantly improve early sepsis stratification, enabling precision critical care interventions.