Partial Response and Stable Disease Correlate with Positive Outcomes in Atezolizumab-treated Patients with Advanced Urinary Tract Carcinoma

Background: The value of a complete response to immune checkpoint inhibitor treatment for urothelial cancer is well recognised, but less is known about long-term outcomes in patients with a partial response or the benefit of achieving disease stabilisation. Objective: To determine clinical outcomes in patients with a partial response or stable disease on atezolizumab therapy for advanced urinary tract carcinoma (UTC). Design, setting, and participants: Data were extracted from three prospective trials (IMvigor210 cohort 2, SAUL, and IMvigor211) evaluating single-agent atezolizumab therapy for platinum-pretreated advanced UTC. The analysis population included 604 atezolizumab-treated and 208 chemotherapy-treated patients (229 achieving a partial response and 583 achieving stable disease). Intervention: Atezolizumab 1200 mg every 3 wk until progression or unacceptable toxicity or single-agent chemotherapy for patients in the control arm of IMvigor211. Outcome measurements and statistical analysis: Baseline characteristics, treatment exposure, overall survival, duration of disease control. Partial response and stable disease populations were analysed separately. Results and limitations: The population of patients with a partial response included more patients with programmed cell death ligand 1 (PD-L1) expression on ≥5% of tumour-infiltrating immune cells than the stable disease population. The median time to best response was 2.1 mo across trials and treatments, regardless of the type of response. Atezolizumab-treated patients with a partial response had sustained disease control (median overall survival not reached); durations of disease control and overall survival were longer with atezolizumab than with chemotherapy. In patients with stable disease, median overall survival was numerically longer with atezolizumab (exceeding 1 yr) than with chemotherapy. Irrespective of treatment, durations of disease control and survival were shorter in patients with stable disease than in those achieving a partial response. These analyses are limited by their post hoc exploratory nature and relatively short follow-up. Conclusions: Stable disease and partial response are meaningful clinical outcomes in atezolizumab-treated patients with advanced UTC. Patient summary: In this report, we looked at the outcomes in patients whose tumours responded to treatment to some extent, but the tumour did not disappear completely. We aimed to understand whether a modest response to treatment was associated with meaningful long-term outcomes for patients. We found that on average, life expectancy was >1 yr in patients whose disease was stabilised and even longer in those whose tumours showed some shrinkage in response to treatment.