The use of patch tests in the diagnosis of delayed hypersensitivity drug eruptions.

Abstract

Adverse drug reactions (ADRs) are common. In the absence of a sufficiently sensitive and specific laboratory test, identification of the culprit drug remains a diagnostic challenge. Patch tests have recently been advocated as a means of detecting drug sensitivity. To further elucidate the usefulness of patch tests in revealing the causative drugs of cutaneous ADRs (CADRs). We conducted a non-blinded, prospective, controlled clinical trial. Twenty-five patients with a history of CADRs were patch tested, and 25 healthy subjects who had never experienced CADRs served as controls. A morbilliform eruption was the most frequent skin reaction. Patch tests were positive in eight of the 25 patients with CADR (32%). Specifically, five of the 13 patients with morbilliform drug eruption (38.4%) tested positive, as did one of the four patients with erythema multiforme/Stevens-Johnson syndrome (25%), and one of the two patients with the drug reaction with eosinophilia and systemic symptoms syndrome. Antibiotics and anticonvulsants resulted in positive patch tests most often. Patch test sensitivity was 32%, specificity was 92%, and negative and positive predictive values were 57.5% and 80%, respectively. Significant correlation was found between the patch test result and the clinical probability of a CADR according to the imputability score of the drug. Patch testing for drugs causing ADRs shows high specificity rates even though the sensitivity is low. Such tests may therefore be useful in supporting the diagnosis of delayed-type CADRs, particularly when antibiotics or anticonvulsants are involved and the cutaneous reaction is a morbilliform rash.