OUTCOMES AND HEALTHCARE RESOURCE UTILIZATION OF PATIENTS RECEIVING DEFINITIVE VERSUS NON-DEFINITIVE TREATMENT FOR MUSCLE-INVASIVE BLADDER CANCER: A REAL-WORLD ANALYSIS WITHIN THE VETERANS AFFAIRS SYSTEM

Introduction

Approximately 25% of bladder cancer patients (pts) are diagnosed with muscle-invasive disease (MIBC). Current definitive treatments (DT) for MIBC with curative intent include radical cystectomy (RC) and chemoradiation (CR). Despite established survival benefits and strong recommendations for DT in guidelines, approximately 50% of MIBC pts only receive non-definitive treatments (NDT). The natural histories and outcomes of MIBC pts receiving NDT are insufficiently described. We analyzed real-world data to better characterize MIBC pts undergoing NDT by evaluating their demographic and clinical characteristics, survival, treatment patterns, health care resource utilization (HCRU) and associated costs, and compare them to pts receiving DT within the context of an equal-access health care system, the US Department of Veterans Affairs (VA).

Methods

Pts diagnosed with MIBC within VA system from 2010-2019 were identified by a natural language processing model and supplementary chart review, then categorized into one of two DT groups (RC or CR) or the NDT group. Baseline characteristics were summarized and compared by omnibus tests. Overall survival (OS) was visualized by the unadjusted Kaplan-Meier plot and compared by the log-rank test. Multivariable Cox proportional hazard models with time-dependent covariates tested the impact of treatment groups on OS and bladder cancer-specific mortality (BCSM), adjusting for baseline characteristics. Competing risk models accounted for death from other causes when analyzing BCSM. All-cause HCRU and associated costs were summarized on a per-pt per-year (PPPY) basis, and differences across treatment groups were assessed by omnibus tests.

Results

We identified 1,524 pts with MIBC: 650 received NDT, 740 received RC, and 134 received CR. NDT pts were older at MIBC diagnosis (median: 77[NDT], 68[RC], 73[CR] years; p<0.001) and had higher Charlson Comorbidity Index scores (mean: 1.9[NDT], 1.3[RC], 1.7[CR]; p<0.001). No difference was found for socioeconomic status among treatments. An OS advantage was observed for pts receiving DT vs NDT in both the unadjusted Kaplan-Meier plot (log-rank p<0.001) and the Cox model (adjusted hazard ratios (HR): 0.49[RC], 0.65[CR]; both, p<0.001) (Figure 1). Additionally, pts receiving DT had lower BCSM compared to those receiving NDT (adjusted HR: 0.45[RC], 0.57[CR]; both, p<0.001). In all-cause HCRU and costs, inpatient costs were highest for RC, outpatient costs were highest for CR, and the number of emergency room (ER) visits was highest for NDT (Table 1; all, p<0.001).

Conclusions

This real-world analysis in a large equal-access health care setting suggests that pts with MIBC receiving NDT were older, had more comorbidities, similar socioeconomic status, and worse survival compared to pts receiving DT. Although NDT was associated with lower inpatient and outpatient costs than DT, the overall inpatient and outpatient costs for NDT pts remain substantial despite not receiving guideline-recommended treatment. High number of ER visits further increases the burden of NDT pts. Although we measured HCRU and associated costs, further studies are needed to evaluate other impacts of DT and NDT on pts including productive life-years lost, and the physical and emotional burden on pts and caregivers, especially for NDT pts as the latter have not been well studied. Taken together, these findings highlight unmet needs among MIBC pts receiving NDT where novel, more efficacious bladder-sparing treatments are required.