Effect of timing of palliative care consultation on survival and end-of-life care outcomes in mNSCLC.

Abstract

247 Background: Early specialist palliative care (PC) involvement in metastatic non-small cell lung cancer (mNSCLC) is associated with improved quality of life, less aggressive end of life (EoL) care, and longer survival. However, PC utilization remains low, and treatment paradigms have evolved. We examined how the timing and extent of PC involvement impacted outcomes in patients with mNSCLC in the era of targeted- and immune-therapy. Methods: This retrospective review analyzed patients with mNSCLC who initiated first-line treatment with chemotherapy, immunotherapy, or combined chemoimmunotherapy at Duke University between March 2015 and July 2019 and had at least one PC visit. Data were abstracted through November 2022. Patient outcomes data, including the NCCN Distress Thermometer (DT), survival, and EoL care, were analyzed using descriptive statistics. Aggressive EoL care was defined as any cancer-directed therapy, >1 emergency department visit, >1 hospital admission, or >14 days spent in the hospital in the last 30 days of life; or hospice entry within the last 3 days of life or in-hospital death. To examine associations between PC utilization and timing of PC referral with actionable distress we fit generalized estimating equations. Results: The 60 patients were stratified by time to first PC encounter and total number of PC visits. Average age was 63.6 years, 53% were female, and 60% Caucasian. 32% were seen within 2 months of diagnosis (early), 37% between 2-6 months (moderate), and 32% were seen by PC after 6 months (late). 27% of patients saw outpatient PC only once, 43% between 2-4 times, and 30% saw PC 5 or greater times. There was no evidence of an association between PC utilization and actionable distress (NCCN DT score of 4 or greater) nor with timing of PC referral and actionable distress. Median survival was 15.3 months in the early PC group, 6.2 months in the moderate group, and 17.0 months in the late group, with a median survival of 12.5 months. Patients who received early PC spent more time on hospice (45.5 days) compared to those with moderate (29.6 days) and late (22.5 days) PC. Receipt of aggressive EoL care was high overall; 32% of patients with early PC visits received aggressive EoL care compared to 50% in moderate and 63% in late PC groups. Of the 51 decedents, only 30 (59%) enrolled in hospice. 11% of patients in the early PC group died in the hospital, compared to 14% in the moderate and 21% in the late groups. Conclusions: This real-world study reveals that referrals to PC still occur late in the disease course of mNSCLC, despite proven benefits of early integration. Early PC resulted in longer time on hospice, lower frequency of aggressive EoL care, and lower rates of in-hospital death, with no evidence of decreased survival. These findings suggest that early specialist PC improves outcomes in the era of novel mNSCLC therapies, but that PC remains underutilized, and many unmet needs persist in this population.