Nausea as a mediator of olanzapine’s effect on quality-of-life improvement: Insights from a phase III NCORP RCT.

352 Background: Refractory nausea remains a significant and distressing symptom for cancer patients despite guideline-based antiemetic regimens, often severely impacting quality of life (QOL). This analysis determined the extent to which reductions in nausea from olanzapine (OLZ) contribute to improvements in QOL, using moderated mediation analysis to quantify this relationship. Methods: The URCC NCORP Research Base, in collaboration with 21 NCORP practices, screened 1,363 chemotherapy-naïve breast cancer patients starting a moderate to highly emetogenic chemotherapy regimen for this Phase III RCT (NCT03367572), with all patients receiving ASCO-recommended antiemetics. Three timepoints were defined: baseline at time 0; screening during the first chemotherapy cycle at time 1; and randomization at the second cycle (time 2) for patients meeting eligibility criteria. Patients reporting moderate nausea (≥3 on a 1-7 scale) during Cycle 1 and willing to continue were randomized (1:1:1) into three arms each receiving the ASCO regimen (n = 310): 1 OLZ, 2 prochlorperazine (PC), or 3 placebo. The analysis focused on OLZ as it had the largest effect size on QOL. Nausea was evaluated using a four-day home diary (4x daily); QOL was measured with the FACT-G. Moderated mediation analysis using a structural equation model estimated total and indirect effects, and proportion meditated by nausea between OLZ and QOL. Analyses were done separately for FACT-G and its subscales. Changes of ~5 points in FACT-G and 2 points in subscales were considered clinically meaningful. Time moderation was evaluated with interactions between time and arm. A Bonferroni correction (α = 0.00625) adjusted for multiple testing. Results: Nausea significantly mediated the relationship between OLZ and QOL, with 50.5% of the total effect at time 2 attributable to decreased nausea (p = 0.001). The total effect of OLZ on FACT-G was an increase of 4.67 points (p = 0.002), a clinically meaningful improvement, with 2.36 points mediated by decreased nausea (p = 0.001). For the PWB subscale, the total effect was an increase of 2.53 points (p = 0.001), with 1.50 points mediated by decreased nausea (p < 0.001), representing 59.4% of the total effect at time 2 (p < 0.001). A significant time-by-arm interaction was observed in both the FACT-G and PWB models (p < 0.001), highlighting a time-dependent relationship between nausea and QOL improvements. Conclusions: Effective nausea control is a key mediator of OLZ’s impact on QOL, with over 50% of the improvement in FACT-G and nearly 60% in PWB attributable to decreased nausea —both statistically significant, PWB with clinical significance, and FACT-G approaching clinical significance. These findings highlight the critical role of effective nausea management in enhancing QOL and support the integration of OLZ into routine clinical practice as a valuable tool for improving patient outcomes. Clinical trial information: NCT03367572 .