Factors Associated with Early Termination of Palliative Radiotherapy in a Hospitalized Patient Cohort

Abstract

Patients receiving inpatient palliative radiotherapy (RT) for symptomatic metastases may not complete their treatment course, often due to declining functional status. Established tools estimating life expectancy have variable success in assessing whether patients are likely to complete a course of palliative RT, especially in a hospitalized setting where patients have other acute medical concerns. We report on a retrospective analysis of hospitalized patients initiated on palliative RT to determine predictive factors associated with early treatment termination. A retrospective analysis of 76 consecutive inpatient palliative RT treatments from August 2017 to September 2019 at a single academic institution was performed. Baseline characteristics including age, sex, Eastern Cooperative Oncology Group (ECOG) performance status, cancer diagnosis, and RT site were recorded. Treatment features reviewed included hospitalized days prior to consultation, time from consultation to treatment delivery, and delays (defined as 1+ hour from planned simulation/treatment). Univariable logistic regression was used to determine whether any variables were associated with incomplete treatment delivery. The Benjamini-Hochberg correction method was used to control for the false discovery rate (FDR) with an adjusted p-value threshold of 0.20 to address the multiple testing issue. Median age of the cohort was 61 years (interquartile range (IQR) 45-71). 74% and 26% of patients were ECOG 1-2 and 3-4, respectively. 80% completed treatment, while 20% did not complete palliative RT. Of those who terminated RT early, a median of 10 (IQR 10-12) fractions were planned, but a median of 4 (IQR 2-4) fractions were ultimately delivered. 9 of 15 (60%) patients who terminated RT early either rapidly declined in performance status or electively stopped to transition to hospice. Of the factors evaluated, only increased time from consultation to first treatment (odds ratio 1.01, unadjusted p = 0.02, adjusted p = 0.16) was significantly associated with early RT termination. Median time from consultation to first treatment was 2.9 (IQR 1.2 – 6.0) days. Patients who terminated RT early died after median of 23 (IQR 13–33) days from initial consultation. A fifth of hospitalized patients were unable to complete palliative RT, the majority of whom rapidly declined in ECOG status and/or transitioned to hospice. Increased time from consultation to initial treatment was related to early RT cessation, potentially due to ongoing deliberations about whether to initiate RT, or due to acute complications impacting medical stability for treatment. Implementing earlier goals-of-care discussions and prioritizing shorter fractionation schema would be helpful at reducing the number of initiated RT courses which are not completed, as these treatments have low therapeutic potential.