The Inpatient Immunotherapy Outcomes study: A multicenter retrospective study of patients treated with immune checkpoint inhibitors in the inpatient setting.

300 Background: Immune checkpoint inhibitors (ICIs) have revolutionized the care of patients (pts) with cancer. However, median time to response is 2-6 months and many pts derive no benefit from ICIs. Pts are often admitted to the hospital with complications of advanced disease or its treatment, and many have prognoses limited to months. Several recent studies have demonstrated a limited benefit of anticancer therapy at end of life. ICIs are also associated with significant financial toxicity for both pts and the health care system. The role of ICI therapy in the inpatient (IP) setting is unclear. To begin to address this gap in knowledge, we conducted the Inpatient Immunotherapy Outcomes Study (IIOS) to describe characteristics and outcomes of pts who received IP ICIs. Methods: IIOS is a multicenter, retrospective study of pts treated with PD-(L)1 and CTLA-4 inhibitors during IP hospitalization between 2012-2021 at 4 large academic institutions: Mount Sinai Hospital, Yale-New Haven Hospital, University of Pennsylvania, and Georgetown University Hospital. Manual data collection was performed using each institution’s EMR and was approved by each institution’s IRB. Descriptive statistics were used to characterize the population and demonstrate pt outcomes. Results: A total of 159 pts received IPI ICI. Median age was 61 years. 54.7% of pts were white and 17.6% were Black; 12.6% were Hispanic. Thoracic/head and neck malignancies were most common (26.4%), followed by gastrointestinal (19.5%) and hematologic malignancies (17.6%). Most pts (73%) initiated ICIs in the IP setting while 27% continued an outpatient ICI regimen. 129 pts (81.1%) had stage IV solid malignancies at the time of ICI initiation in any setting; median prior lines of systemic therapy was 1 (range, 0-11). The most commonly administered IP ICI was pembrolizumab (49.1%) followed by nivolumab (34.0%), with ICIs administered with non-curative intent in 91.8% of pts. In 44.7% of pts, the ICI given did not have an FDA approval for that cancer type and stage at the time of administration. PD-L1 expression was available on tumors from 60 pts, 32 (53.3%) of whom had expression of 50% or higher. 112 pts (70.4%) had no documented clinical or imaging-based response to ICI therapy. Discharge disposition included home (47.2%), IP death (27%), rehabilitation centers (15.1%), and hospice (10.1%). Median days between first IP ICI dose and death was 47 (95% CI, 33-68). Conclusions: This is the largest multi-institutional effort to understand pt outcomes following IP ICI administration. Preliminary data, as outlined above, is concerning for poor clinical outcomes which should give clinicians pause when considering IP ICI use. Further analysis is ongoing to determine predictors of overall survival and discharge to home.