Lichenoid drug reaction to antituberculosis drugs treated through with topical steroids and phototherapy

Sir, Tuberculosis-associated adverse drug reactions (ADRs) result in an interruption and change of treatment, both impacting on treatment failure, the development of drug resistance, relapse and the transmission of disease.1 The threat of transmission makes prompt resumption of therapy necessary, but the limited number of effective antituberculosis drugs complicates management. The spectrum of severe tuberculosis-associated cutaneous ADRs is wide and includes Stevens–Johnson syndrome, drug hypersensitivity syndrome and lichenoid drug reactions (LDRs).2 In all, with the exception of LDRs, acute clinical or laboratory markers enable early detection and withdrawal of the offending drug following drug rechallenge.3 We present a case of severe LDR related to antituberculosis drugs in which the offending drug could not be definitively identified and therapy was continued successfully under the cover of topical steroids and phototherapy. A 39-year-old HIV-infected man presented with erythroderma 3 months after starting isoniazid, rifampicin, pyrazinamide and ethambutol for pulmonary tuberculosis and co-trimoxazole for Pneumocystis jiroveci prophylaxis. He had a CD4 count of 8 cells/mm3 and was taking efavirenz, lamivudine and stavudine, all of which were started after the onset of the rash. A skin biopsy showed a lichenoid reaction with eosinophils and apoptotic keratinocytes compatible with LDR. Antituberculosis drugs and co-trimoxazole were stopped and he was started on topical clobetasol (Dovate™) daily. The rash settled with residual pigmentation and ofloxacin, streptomycin and ethionamide were initiated.1 Rifampicin was then reintroduced with escalating doses over 4 days without a clinically evident reaction. Ethambutol, pyrazinamide and isoniazid were not reintroduced on the assumption that they were the likeliest offenders. He was discharged from hospital on the four antituberculosis drugs. Six weeks later he presented again with a 3 week history of the same rash. In the meantime his sputum reverted to culture-negative, and his tuberculosis symptoms and radiological features had improved. We thought rifampicin was the likeliest offender, but decided not to replace it with another second-line drug or rechallenge with the three first-line drugs initially omitted. This was informed by the patient's unwillingness to restart the rechallenge process, the clinical improvement and the delay in clinical manifestation of LDR, possibly due to topical steroids. We continued treatment under cover of clobetasol. The rash improved and he was discharged from hospital. He presented again 3 weeks later with worsening hyperpigmentation, depigmentation and diffuse alopecia, despite topical steroids. We introduced thrice weekly bath psoralen and ultraviolet A (PUVA). His skin gradually improved; however, 1 month later he developed photo-distributed rash with painful fissures, which we attributed to excessive sun exposure (see Figure 1). PUVA was suspended for 1 week and he was advised to minimize sun exposure. Sunscreen was prescribed, and he was advised to wear a hat and gloves on the day he received psoralen, a potent photosensitizer. He completed 12 months of treatment, and continued steroids and PUVA for 3 months thereafter. His skin has clinically improved with only residual hyperpigmentation and leucoderma (see Figure 1). Figure 1. (a and b) Lichenoid drug reaction showing depigmentation, hyperpigmentation and fissuring. (c) Clinically improved skin 1 week after completing antituberculosis therapy. LDR presents as purple itchy papules becoming confluent and hyperpigmented with continuing exposure to the offending drug. The interval between initiating the drug and the rash ranges from days to years, with most cases occurring within months. On withdrawing the drug, the lesions resolve with persistent hyperpigmentation, often lasting for many years.4 The lack of acute markers, insidious onset of the rash, and varying intervals between drug initiation and a clinically detectable rash make it difficult to establish a temporal relationship with the drug and ascribe causality in LDR. This is more so in patients receiving multiple drugs. The limited number of effective antituberculosis drugs, the cessation of which is associated with a higher mortality, increases risk of drug resistance, longer duration of therapy and public health concerns, make it necessary to balance the interruption of therapy against treating through the ADRs.1,5 PUVA is used to treat a wide variety of inflammatory dermatological conditions, such as scleroderma, lichen planus, psoriasis, vitiligo, cutaneous T cell lymphoma and atopic dermatitis. It is also effective for pruritus associated with systemic disease. Psoralen is administered orally or topically by soaking in bathwater, followed by irradiation. The first-line therapy generally used for LDRs is high-potency topical steroids, although there are no clinical trials supporting their use. There are better data supporting the use of acitretin, systemic steroids and other immunosuppressants.6 However, considering our patient's marked immunosuppression and concurrent tuberculosis, we decided to use only topical steroids. This case illustrates practical dilemmas in managing tuberculosis-associated ADRs. Should suboptimal treatment be used or should a potentially life-threatening therapy be continued despite serious side effects? We decided to continue treatment and manage the side effects, as the alternative would likely result in a poorer outcome. Phototherapy, in combination with topical steroids, is an option in the management of LDRs when the offending drug cannot be identified. Written permission was given by the patient to publish the case and images.