Taste Disturbance Evaluation Is Feasible and Higher Melphalan Pharmacokinetics Are Associated with Poorer Nutrition and Symptom Burden after Autologous HCT for Multiple Myeloma

Abstract

Introduction Dysgeusia commonly influences nutritional intake and quality of life after HCT, yet its pathobiology is poorly understood and methods to study it are not established. We report interim results of a prospective feasibility study of chemical gustometry (CG) to measure dysgeusia in pts with multiple myeloma (MM) undergoing high-dose melphalan and autologous HCT (mel-AHCT). We evaluated associations between dysgeusia and symptoms, nutrition, mel pharmacokinetics (PK), and oral microbiota composition. We hypothesized that mel penetration into saliva contributes to dysgeusia. Methods CG was performed by the Henkin method pre-AHCT and on days (d) -1, +7, +14, +30 testing 5 flavors (sweet, sour, salty, bitter, umami) to calculate a total taste score (max value 30). CG was considered feasible if > 60% of pts could complete > 60% of all evaluations. Caloric intake (kcal) was calculated with 24-hr recall using the USDA 5-Step Multiple-Pass Method, and pt-reported symptoms were assessed using the MD Anderson Symptom Inventory (MDASI) at the above times and on d +3 and +10. Salivary microbiome diversity was assessed longitudinally by 16S sequencing and the inverse Simpson index. 6-point serum mel concentrations were used to calculate an area under the curve (AUC) and peak salivary mel was obtained 75 min post-infusion. Spearman rank-sum correlation was used to assess associations. Results 22 mel-AHCT MM pts have follow-up through d +30. 18 pts (82%) completed > 60% of all assessments. Median age was 62 (range 41-75). 11 pts (50%) were male. Mel 200 mg/m2 and 140 mg/m2 was given in 19 (86%) and 3 (14%) pts, respectively. Median serum mel AUC was 12.6 mg*h/L (range 5.8-17.5); median salivary mel was 119 ng/ml (range 25-662). Higher serum mel AUC correlated with higher salivary mel levels, p=0.004, and associated with poorer caloric intake from pre-ASCT to d +30, p=0.04. Higher serum mel AUC associated with higher MDASI scores (worse symptom burden) from pre-AHCT to d +30, p=0.005, but not with taste scores. Higher salivary mel associated with poorer caloric intake from pre-AHCT to d +30, p=0.02. MDASI scores peaked d +10, coincident with nadirs in taste scores and caloric intake on d +7 and +10, respectively. Taste scores partially recovered by d +30, but often did not return to baseline (Fig 1A-C). Oral microbiota diversity was stable early after AHCT but increased by d+100. D+7 diversity did not correlate with serum mel AUC, p=0.4, or salivary mel, p=0.84 (Fig 2A-C). Conclusion CG is a feasible method to evaluate dysgeusia after AHCT. Dysgeusia peaked by d +7 and often persisted through d +30. Oral microbiota diversity was stable early after AHCT. Higher serum and salivary mel levels did not correlate with microbiota diversity but were associated with poorer caloric intake and higher MDASI scores suggesting that using targeted mel PK dosing may improve caloric intake and limit symptom burden after AHCT.