As more patients are being treated with imatinib (formerly STI 571), cutaneous reactions1,2 are being recognized as very common among the side effects of the drug, ranging from mild macular erythema to widespread papulous lesions, that sometimes are pruriginous and cause discomfort to the patients. Some rare cases of Stevens-Johnson syndrome have also been described.3 In some patients, the cutaneous side effects of imatinib led to the discontinuation of the treatment, while in others the drug could be reinitiated progressively after a course of corticosteroids.2
Two of our patients treated with imatinib developed widespread cutaneous reactions characterized as symmetric pruriginous maculopapular exanthema that was confluent in some areas. The skin biopsy showed in both cases an inflammatory infiltrate in the superficial dermis, mainly of perivascular lymphocytes. The physiopathology of this type of cutaneous lesion is unknown, as it is not known whether it is immunologically mediated. We tried not to give corticosteroids, because high doses of prednisone are sometimes required and could be harmful, but we continued to give imatinib while looking for an alternative solution.
We searched for a drug that could inhibit the cellular mechanisms presumably implicated in this type of skin lesion, and as a result we decided to try cyclosporin A4 because of its known immunosuppressive action. However, the systemic action of cyclosporin A given orally seemed to us an excessive treatment, so we decided to try an ointment of cyclosporin A cream, with the aim of determining its efficacy and its absorption through the skin. The cream was prepared in the British Hospital Pharmacy Department because it is not commercially available. Briefly, a 1% cyclosporin A cream was prepared as follows: 1000 mg cyclosporin A (10 mL of Sandimmune neoral solution; Novartis, Basel, Switzerland) was mixed with urea 1%, 100 000 IU vitamin A palmitate, vitamin E (as tocoferol acetate 2%), 10 drops lemon essence, and hydrosoluble cream in an amount to equal 100 grams for the entire solution.
Patient 1 was a 50-year-old female with chronic myeloid leukemia (CML), with an interval of 16 months since diagnosis. She started imatinib (400 mg daily) because of interferon resistance and intolerance. After 3 weeks on imatinib, she presented with a pruriginous maculopapular exanthema in the abdomen and both legs. She was treated with the 1% cyclosporin A cream, 3 to 4 times per day. After 48 hours, an initial response was seen, with less erythema and pruritus. After continuous treatment in the way described, the cutaneous lesions completely disappeared in 30 days.
Patient 2, a 63-year-old male with chronic-phase CML diagnosed 48 months earlier, was switched to imatinib therapy (400 mg daily) because of cytogenetic resistance to interferon. After the first week, he presented with a pruriginous maculopapular exanthema that affected arms, legs, and abdomen. He was treated with cyclosporin A cream in the same schedule as the previous patient, with complete resolution of the cutaneous lesions after 2 months. As mentioned, none of the patients required corticosteroids, and the imatinib administration was maintained at the same dosage during the time of the skin treatment. Serial blood determinations of cyclosporin A disclosed undetectable levels, so we conclude that the absorption of cyclosporin A through the skin was minimal or null while it exerts its pharmacologic action locally.
After the complete disappearance of the cutaneous reactions induced by imatinib, both patients stopped the treatment with the cyclosporin A cream while maintaining the imatinib administration. No new cutaneous reactions have been observed in either patient since that time, with a follow-up period of 24 months.
The exact mechanism of action of cyclosporin A in the skin is controversial. Some report a direct effect on Langerhans cells,5 while other experimental papers report an inhibition on T cells as measured in the in vitro mixed skin-cell lymphocyte reaction.6 We think that this small experience could be expanded in the future, because more patients are being treated with imatinib in different indications (such as first-line treatment in CML) and more cutaneous reactions to the drug are expected. In this way, perhaps this cyclosporin A-based local treatment could be compared with other different type of creams, like the commercially available corticosteroid-containing creams. This approach could also be tested for the treatment of cutaneous reactions associated with other drugs (such as antibiotics) or of the localized liquenoid cutaneous lesions in patients with chronic graft-versus-host disease.
