Outcomes with initiation of inpatient immunotherapy.

Abstract

7 Background: Immune checkpoint inhibitors (ICIs) have transformed cancer care; however, studies establishing their benefit have been primarily conducted in robust patients (pts) with good performance status. Administration of ICI to hospitalized pts remains a subject of debate. While there is often a desire to reverse clinical course in patients who are acutely ill, previous studies show minimal benefit from anticancer therapy near the end of life with the potential for toxicity. While ICI are generally better tolerated than traditional cancer therapies, the median time to response ranges from 2-6 months and pts with a poor PS may not derive benefit. ICIs also bring significant financial burdens to pts and healthcare systems. The role of ICIs in the inpatient setting remains unclear. Methods: This study retrospectively examined pts who received PD-(L)1 inhibitors while IP between 2016 and 2023 at MedStar Georgetown University Hospital. Data were collected manually from institutional EMR. Descriptive statistics were used to depict the pt population and show their outcomes. The Kaplan-Meier method estimated survival probabilities from treatment initiation to follow-up or death, with 95% confidence intervals (α = 0.05). The Log rank non-parametric test compared survival distributions among ICIs. Results: Of the 38 pts who initiated ICI as an inpatient, 16 were male (42.1%) and 22 were female (57.9%). The median age was 60 years (31–88 years). Gastrointestinal malignancies were most common indication (31.6%), followed by thoracic/head and neck (21.1%) and hematologic (18.4%) cancers. Among the cohort, 26.3% received nivolumab, 68.4% received pembrolizumab, and 5.3% received a combination of nivolumab and ipilimumab. PD-L1 status was unknown for the majority (79.0%) of pts at time of ICI; 15.8 % were known to be PD-L1 high. Only 14 of 35 (40.0%) patients went on to receive any additional immunotherapy after discharge. The median days to death after discharge was 16 days (0-376 days), with a mean of 36.4 days. The median overall survival was 137 days from IP treatment initiation. The survival probability at 91 days was 65.5% (95% CI: 45.4-79.7) and 31.0% (95% CI: 15.6-47.9) at 182 days. No significant differences were observed between the survival distributions by ICI type (p=0.966) or primary malignancy (p=0.488). Conclusions: Our findings underscore the complexities of initiating inpatient ICI treatment and question its appropriateness. Delaying ICI to outpatient settings may enhance pt tolerance, optimize outcomes, and potentially mitigate unnecessary costs. This approach may allow thorough consideration of performance status outside the hospital, possibly improving treatment effectiveness while reducing futile acute care costs.