Previously reported skin reactions related to INH were late-type reactions such as maculopapular rash, dermatitis, erythema multiforme, drug reaction with eosinophilia and systemic symptoms (DRESS) syndrome, Stevens-Johnson syndrome, and toxic epidermal necrolysis.5,6 The concerning prediagnosis of psoriatic exacerbation in our patient was ruled out by the absence of typical skin lesions such as well-defined, erythematous plaques and pustules and atypical localization such as the dorsal hands and feet rather than the knees, elbows, lumbosacral region, scalp, and abdomen, which is typical of psoriasis. DRESS syndrome was unlikely with the absence of fever, lymphadenopathy, hypereosinophilia, leukocytosis, and renal and hepatic dysfunction.7 There were no widespread blisters, epidermal detachment, or mucosal involvement on the trunk or face typically associated with Stevens-Johnson syndrome and toxic epidermal necrolysis.7,8 A possible diagnosis of contact dermatitis was suspected with likely skin lesions as exfoliation and chapping, typical localization on the hands and feet, and positive patch test that supported sensitization to the drug. However, the patient’s skin lesions were not eczematous (characterized by erythema, vesiculation, exudation, or bullous edema in the acute phase), and were not localized to areas of irritant exposure.3 In our patient, erythematoedematous lesions in an acral distribution with no mucosal involvement and systemic exposure to INH was compatible with erythema multiforme, whereas the absence of target appearance, positive patch test, and late appearance were incompatible with erythema multiforme.8
Because the clinical picture did not fit contact dermatitis or erythema multiforme, a diagnosis of erythema multiforme–like noneczematous dermatitis was suggested. Noneczematous dermatitis has subtypes that include purpuric, lichenoid, pustular, lymphomatoid, dyshidrosiform, and pigmented, as well as erythema multiforme–like contact eruptions.9 These clinical entities are not associated with contact exposure, but are related to systemic exposure, as seen in our patient.10 The patch test positivity and skin biopsy report also supported the diagnosis of erythema multiforme–like dermatitis. Erythema multiforme–like dermatitis is thought to be caused by medications or infections inducing immunocomplexes and lymphocytic infiltration in the dermis and subepidermis. Nevertheless, the prognosis was self-limiting in both.8 The clinical polymorphism caused by INH in this patient was suggested to be related with individual susceptibility, variability of contact-activating modalities, and the targeted cutaneous structures. Furthermore, among the risk factors for cutaneous ADRs—HIV, polypharmacy, older age, and preexisting renal and liver impairment—the only notable factor in this patient was psoriasis as an autoimmune disorder.
Patients with skin diseases such as psoriasis should be followed up by closer monitoring during INH use. Withdrawal of the drug and symptomatic treatment of the lesions with corticosteroid and antihistamine are the first steps of drug intolerance. After complete recovery and termination of antiallergic drugs, diagnostic tests are recommended if the drug reaction was not life-threatening. Skin prick and intradermal tests are useful in early-type drug reactions, whereas patch testing and late evaluation of an intradermal test may be helpful in the diagnosis of delayed-type reactions. The full dose of INH is avoided in an intradermal test against irritation. A patch test with INH was performed by diluting a 100-mg tablet with 1 mL of distilled water, and used as 1/100, 1/10, and 1/1 dilutions.8 Patch testing with INH also was done in 5 healthy control patients to exclude the irritation effect in this case. The rechallenge of INH was done in a controlled manner in our patient to rule out psoriasis activation since it was a localized skin reaction with no serious ADR. An oral provocation test with the culprit drug is the gold standard of drug allergy diagnosis that should be done in a tertiary hospital with an intensive care unit.
This case of erythema multiforme–like dermatitis due to INH is interesting due to systemic intake of INH, which resulted in dermatitis with localized involvement similar to erythema multiforme but with no immunologic processes or prior sensitization. With the increasing use of anti–TNF-α treatment, INH use will be more prevalent than in the past for the treatment of latent TB. Even though the skin-restricted ADRs of INH are rare and minor, particular attention should be paid to patients with dermatologic diseases. In our case, diagnostic drug allergy evaluation was performed to optimize the second-line treatment of TB infection, in addition to early withdrawal of the culprit drug.