JMIR Publications

ABSTRACT

Background:

Progression-free survival (PFS) is a crucial endpoint in cancer drug research. Real-world evidence (RWE) is increasingly accepted to augment traditional clinical trial findings to better understand the effectiveness of oncological interventions. RWE can be leveraged to improve novel therapy development programs and provide better post-market surveillance of approved therapies.

Objective:

To develop and validate a novel semi-automated workflow that estimates real-world progression-free survival (rwPFS) in patients with metastatic breast cancer (mBC).

Methods:

This study analyzes de-identified EHR data using the nference® nSights platform. The overall cohort included 316 patients with HR-positive, and HER2-negative mBC, who were started on Palbociclib and Letrozole combination therapy between January 1, 2015, and December 31, 2021. We curated ground-truth datasets to evaluate the workflow's performance at sentence and patient levels. NLP-captured-progression or a change in the line of therapy were considered outcome events. Death, loss to follow-up, and end of study period were considered censoring events for computing rwPFS.

Results:

At the sentence-level, progression events were captured from clinical notes and radiology reports with a sensitivity of 99.8%, specificity of 96.7%, and accuracy of 98.2%. At the patient-level, initial progression was correctly captured within a window of ±30 days with a sensitivity of 92.5%, specificity of 83.0%, and accuracy of 88.0%. In a sample of 100 patients, the median rwPFS was determined to be 22 months (95% CI; 15-35 months) by the computational workflow and 25 months (95% CI; 15-35 months) by manual curation.

Conclusions:

A semi-automated workflow enabled rapid and reliable determination of rwPFS in mBC patients receiving a combination therapy.