Dietary links with symptoms in early-phase immune checkpoint combinations trials.

341 Background: We examined diet and its association with symptoms at first tumor restaging/end of cycle 2 (EOC2) among patients enrolled in early-phase trials of combination treatment including an immune checkpoint inhibitor. Methods: Patients (ECOG 0–2) scheduled to begin an early-phase immune checkpoint inhibitor-included combination trial were recruited into a prospective longitudinal study. At first tumor restaging/ EOC2, patients completed validated instruments: Multifactor Dietary Screener (↑ scores = more frequent consumption of food during past month), Subjective Global Assessment Scale (usual amount of food intake/d during past month compared to pre-trial = unchanged, less, or more), MD Anderson Symptom Inventory (0–10 scale, ↑ scores = worse symptom severity), Functional Assessment of Chronic Illness Therapy “bothered by side effects of treatment” item (0–3 scale, ↑ scores = worse bother). Statistical tests included Spearman’s rho, analysis of variance (ANOVA) and post-hoc Scheffe. Results: Patients (N=150) were 53% female, 77% non-Hispanic White; median age was 59y. The most frequently consumed foods were fruit, other vegetables (>5 times/wk each), green salad (2.8), whole grain breads (2.3), nuts/seeds (2.2), red meat (1.7), and 100% fruit juice (1.6). From pre-trial to restaging, 56% of patients ate the same total amount of food/d, while 32% had decreased and 13% had increased their food intake/d. Symptoms and food choices by change in total food intake from pre-trial to restaging are shown below (table 1). Patients whose food intake decreased from pre-trial to restaging reported worse lack of appetite and nausea (P<.001 each) at restaging than patients whose food intake increased or remained the same. Increased bother at restaging from side effects of treatment was linked to more frequent fruit juice and yogurt intakes and less frequent milk intake (P<.041 each). Appetite loss was associated with more frequent cereal and yogurt intakes/wk (P<.005 each). Worse fatigue correlated with more frequent intakes/wk of yogurt (P=.004) and fruit juice (P=.005). Worse dry mouth was linked with more frequent yogurt and ice cream intakes/wk (P<.041 each). Conclusions: Patients’ symptoms during treatment in early-phase clinical trials of combination therapies that include immune checkpoint blockade may be associated with dietary variance. Fuller understanding of the relation between symptom presentation and dietary patterns in these patients is needed. Less than usual food amount ‡ Unchanged food amount ‡ More than usual food amount ‡ ANOVA P value Symptoms (severity) ↓ Appetite 2.5 a, b 1.4 a 1.3 b <.001 Nausea 2.2 a, b 0.4 a 0.4 b <.001 Fatigue 3.9 a 2.4 a 2.7 .002 Distress 2.1 a 1.0 a 1.3 .008 Food choices (times/wk) Cookies, brownies 1.1 0.9 a 1.8 a .015 Doughnuts 0.7 0.4 0.8 .043 Fruit juice 2.3 1.2 2.1 .053 ‡ At first tumor restaging compared to pre-trial. Similar superscripts across rows = P<0.05 by Scheffe.