Anti-PD-L1 Augmentation of IFN-γ ELISpot Improves Detection of Culprit Drugs in Nonimmediate Drug Hypersensitivity: A...

The IFN-γ enzyme-linked immunospot (ELISpot) assay identifies culprit drugs in nonimmediate drug hypersensitivity reactions (DHRs), but sensitivity is reduced in certain clinical contexts, particularly with systemic corticosteroid use.

To evaluate whether anti-programmed death-ligand 1 (anti-PD-L1) supplementation increases detectable drug-specific T-cell responses in patients at a higher risk of false-negative conventional ELISpot results.

In this retrospective cohort study, 223 patients (450 drug-level tests) with severe cutaneous adverse reactions and other nonimmediate DHRs underwent paired ELISpot testing under conventional and anti-PD-L1-supplemented conditions. Anti-PD-L1 was applied selectively to patients with corticosteroid use, remote reaction history, or drug reaction with eosinophilia and systemic symptoms (DRESS) phenotype. Positivity was defined as ≥20 spot-forming units (SFU) per 10⁶ peripheral blood mononuclear cells. Paired positivity was compared by McNemar's test and SFU magnitude by the Wilcoxon signed-rank test, with prespecified subgroup analyses by corticosteroid exposure, phenotype, and drug class.

Anti-PD-L1 supplementation increased ELISpot positivity from 14.4% (65 of 450) to 35.3% (159 of 450), an absolute increase of 20.9% (94 additional positive tests; P < .001). Conventional positivity was numerically lower in corticosteroid-treated patients (11.6% vs 18.5%, P = .059), whereas anti-PD-L1-modified positivity was similar between groups (35.4% vs 34.7%, P = .950). Augmented responses were consistent across all phenotypes and correlated with established causality scores. Among 114 patients tested against more than 1 distinct drug, positivity remained predominantly limited to a single drug per patient under both testing conditions.

Anti-PD-L1 supplementation substantially increased detectable ELISpot reactivity. Although conventional positivity was numerically lower in corticosteroid-treated patients, positivity was similar between groups after anti-PD-L1 supplementation. Although single-drug positivity predominated, some patients showed reactivity to 2 or more drugs under anti-PD-L1-supplemented conditions. Prospective multicenter validation is warranted.