The impact of cancer-related diarrhea on changes in cancer therapy.

PURPOSE

The impact of cancer-related diarrhea (CRD) on changes in cancer therapy remains poorly characterized despite its prevalence.

METHODS

We performed a longitudinal observational study using IQVIA PharMetrics Plus claims data. Patients included adults with CRD identified by diagnosis codes or pharmacy claims and compared their outcomes to matched (1:1) patients without CRD. Treatment parameters (discontinuation, persistence, augmentation, dose titration, adherence) were evaluated and stratified for the first cancer therapy (chemotherapy vs. targeted therapy vs. both). A multivariate Cox proportional hazards model was used to estimate the difference in risk of each treatment parameter between cohorts, adjusting for cancer type, therapy, and comorbidities.

RESULTS

We identified 104,135 matched pairs of patients with solid (n = 94,411) or hematologic cancers (n = 9,724) receiving chemotherapy (n = 47,220), targeted therapy (n = 2,427), or both (n = 5,313). Patients with CRD discontinued therapy more frequently than those without CRD (chemotherapy [81.5% vs. 62.3%], targeted therapy [69.2% vs. 64.3%], both [96.0% vs. 85.5%], p < 0.0001). The overall proportion of discontinuation was higher (82.4% vs. 64.6%, p < 0.0001), including a higher risk of discontinuation (HR = 1.40, p < 0.001) for patients with CRD. The mean time to discontinuation (59.6 ± 54.1 vs. 68.3 ± 76.6 days), switch (72.0 ± 48.6 vs. 96.9 ± 84.0 days), persistence (95.1 ± 98.1 vs. 154.3 ± 142.7 days), and adherence (25.5% ± 37.2 vs. 47.9 ± 41%) were all lower (p < 0.0001) among patients with CRD.

CONCLUSION

Patients who develop CRD undergo significant and clinically impactful index treatment discontinuation, treatment switching, and have lower adherence and persistence of anticancer therapy compared to patients without CRD. Strategies to control CRD to optimize cancer therapy are urgently needed.