Negative Predictive Value of First-line Anti-tuberculosis Drug Patch Test in Delayed-type Hypersensitivity with Cutaneous Manifestations: The Role of Application Timing.

BACKGROUND: Tuberculosis (TB) remains a significant global health burden, particularly in Indonesia, which ranks second worldwide for TB cases. Hypersensitivity reactions to first-line anti-TB drugs (ATDs) often lead to treatment interruption and increased morbidity. While the drug provocation test (DPT) is the gold standard for diagnosis, it carries significant risks, especially in patients with severe cutaneous adverse reactions (SCARs). Patch testing offers a safer in vivo alternative, yet data regarding its negative predictive value (NPV) and the optimal application time interval in the Indonesian population remain limited. This study aimed to evaluate the positivity rate and NPV of patch testing for first-line ATDs (rifampicin, isoniazid [INH], pyrazinamide, and ethambutol) and to analyze the role of the application time interval in the accuracy of diagnostic results.

METHODS: A cross-sectional study was conducted using data from Dr. Cipto Mangunkusumo Hospital (RSCM), Jakarta, from January 2021 to November 2025. Subjects with a history of delayed-type hypersensitivity to ATDs manifesting as cutaneous reactions were recruited. Subjects underwent patch testing with 30% allergen concentration in petrolatum. Patients with negative patch test (PT) results proceeded to DPT to assess negative concordance. The NPV was calculated, and the relationship between the PT application time interval (days from symptom resolution) and the negative concordance was analyzed using the Mann-Whitney test.

RESULTS: The study included 88 adult patients. The majority were female (67%), with a median age of 43 years. Maculopapular exanthema was the most common phenotype (77.3%), followed by SCARs (22.7%). The PT positivity rates were highest for INH (19/85, 22.4%) and ethambutol (18/85, 21.2%), and lowest for rifampicin (15/83, 18.1%). Based on DPT confirmation, the NPVs were as follows: INH 87.9% (95% confidence interval [CI]: 79.5%-96.3%), ethambutol 80.0% (95% CI: 69.9%-90.1%), pyrazinamide 66.7% (95% CI: 54.1%-79.3%), and rifampicin 60.6% (95% CI: 48.3%-72.9%). The median time interval from symptom resolution to patch testing was 22.5 days. Statistical analysis revealed no significant difference in the median application time between concordant and discordant groups for all drugs ( P > 0.05).

CONCLUSIONS: Patch testing for first-line ATDs demonstrates a high NPV for INH and ethambutol, making it a reliable tool for excluding delayed-type hypersensitivity. However, negative results for rifampicin and pyrazinamide require cautious interpretation due to lower NPVs. Variations in application timing (median: 22.5 days) did not significantly influence the accuracy of negative results.