Hypoalbuminemia as a Predictive Biomarker of Dyspnea Response to Dexamethasone and Placebo in Patients With Cancer: Secondary Analysis of the ABCD Randomized Clinical Trial.

Abstract

We previously found that baseline cytokine levels predicted dyspnea response to dexamethasone and placebo in patients with cancer. However, cytokine immunoassays are impractical for clinical applications, highlighting the need for alternative inflammatory biomarkers. We sought to examine the predictive value of albumin, neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) for dyspnea response to dexamethasone or placebo in patients with cancer. This is a secondary analysis of the Alleviating Breathlessness in Cancer Patients with Dexamethasone randomized clinical trial in which patients with cancer and moderate-to-high dyspnea received either dexamethasone or matching placebo for 14 days. We used generalized additive models to examine associations between baseline albumin, NLR, and PLR and dyspnea numeric rating scale (NRS) response to dexamethasone or placebo. Clinical bloodwork was available for 63 patients. Both the dexamethasone and placebo groups had a significant decrease in dyspnea by day 14 (mean change -1.8 and -1.9, respectively), with no significant between-group differences (-0.1 [95% CI, -1.1 to 1.3]). In the dexamethasone group, lower baseline albumin was significantly associated with a greater reduction in dyspnea NRS (day 14 slope: +1.2, <i>P</i> = .05); in contrast, higher baseline albumin was significantly associated with a greater reduction of dyspnea NRS in the placebo group (day 14 slope: -3.0, <i>P</i> = .004). No associations were found between baseline NLR or PLR levels and dyspnea response to either dexamethasone or placebo. Hypoalbuminemia may be a potential differential biomarker of dyspnea response to dexamethasone and placebo. On further validation, these findings may facilitate future personalized clinical trial design and tailoring of dyspnea interventions.