Baseline characteristics and clinical outcomes of prostate cancer patients on delayed palliative management: a PIONEER analysis based on big data.

Abstract

Introduction: Delayed palliative management (DPM) is an alternative for prostate cancer (PCa) patients with poor performance status, or those who received radical treatment but progressed and no longer meet the criteria for curative treatment. PIONEER is a large network of federated data analytic platforms in PCa that aims to improve its care through the application of big data analytics. The objective of this study was to describe clinical baseline characteristics and outcomes of PCa patients receiving DPM using big data.

Material and methods: Descriptive study of patients on DPM from four databases in PIONEER (Pharmetrics Plus, Optum Clinformatics, Marketscan and Columbia University Irving Medical Center (CUIMC)). Baseline characteristics, including comorbidities (hypertension, type 2 diabetes (T2DM), asthma/chronic obstructive pulmonary disease (COPD) and obesity), were stratified by age. Outcomes of interest were annual emergency department (ED) visits, hospitalization and symptomatic progression. Additional outcomes were time to death, hospitalization and time to symptomatic progression for CUIMC.

Results: We included 13246 men with a median age of 68-75 and Charlson Comorbidity index of 6-8. The three most common comorbidities were hypertension (80-93% [>80 years] vs 69-80% [55-80 years] vs 59-64% [<55 years]), T2DM (29-41% [>80 years] vs 26-38% [55-80 years] vs 23-26% [<55 years] and asthma/COPD (28-37% [>80 years] vs 19-30% [55-80 years] vs 16-19% [<55 years]). ED visits and hospitalizations were highest in the first year of follow-up (19-33% and 21-48% respectively). The median time to death was 548 days (IQR 1265 days) and to symptomatic progression was 408 days (IQR 1125 days) in CUIMC.

Conclusions: Men on DPM were in their mid-seventies, with the three most common comorbidities being hypertension, T2DM and asthma/COPD, regardless of age groups. This study reflects the potential of PIONEER as a federated network of databases that may be used to harness big data in PCa research.