The optimal dose of radiation in the treatment of AIDS-associated Kaposi's sarcoma has been controversial based on previous nonrandomized retrospective studies.

Seventy-one cutaneous AIDS-associated Kaposi's sarcoma lesions were randomly assigned to 1 of 3 radiation dose regimens--8 Gy in 1 fraction, 20 Gy in 10 fractions, and 40 Gy in 20 fractions. Lesions were measured prior to and following treatment. Complete resolution of palpable tumor was considered a complete response, regardless of residual purple pigmentation. Reduction in palpable tumor to less than 50% of pretreatment area was considered an objective response. Less than 50% reduction in tumor size was considered a nonresponse.

Complete response was higher (p = .04) with 40 Gy (83%) and 20 Gy (79%) than with 8 Gy (50%). Absence of residual purple pigmentation was greater (p = .005) with 40 Gy (43%) than with 20 Gy (8%) or 8 Gy (8%). Lesion failure was lower (p = .03) with 40 Gy (52%) than with 20 Gy (67%) or 8 Gy (88%). Median time to failure was 43 weeks with 40 Gy, 26 weeks with 20 Gy, and 13 weeks with 8 Gy (p = .003).

Fractionated radiotherapy to higher total doses resulted in improved response and control of cutaneous Kaposi's sarcoma. This dose-dependence should be considered in determining the optimal radiotherapeutic regimen for individual patients treated for epidemic Kaposi's sarcoma.