Hydroxyurea Treatment in HIV-Infected Patients

To the Editor: Hydroxyurea (HU) is a free-radical quencher that is able to inhibit the cellular enzyme ribonucleotide reductase, decreasing the intracellular pool of desoxyribonucleotides (dNTP) (1). A number of reports published arca 1993-94 proposed that HU, by itself or together with DDI, could play a positive role in the reduction of HIV replication in vitro. Lori et al. demonstrated that as HU decreased the size of the intracellular dNTP, it also inhibited the activity of HIV-reverse transcriptase in HIV-infected peripheral blood mononuclear cells (PMBC) in vitro as well as in macrophages and PMBC infected in vivo (2-5). Based on these data, in the fall of 1993 we undertook a pilot study to evaluate the antiretroviral activity of hydroxyurea in HIV-infected patients who are unresponsive to standard antiretroviral therapy. Patients were eligible for enrollment in the study if they met all the following criteria: a documented history of HIV-antibody seropositivity (both ELISA and Western blot), failure to respond to zidovudine or didanosine therapy, neutrophil granulocyte count >500/mm3 and/or platelets >20,000/mm3, and absence of active opportunistic infections. Previous AIDS diagnosis was not a criterion for exclusion. Hydroxyurea was given orally 500 mg tid. If WBC count dropped below 1000/mm3, a weekly alternating regimen was initiated of 1500 mg/day for the first week and 1000 mg/day for the subsequent week; however, treatment was interrupted if WBC count decreased below 1000/mm3 and/or opportunistic infections occurred. If the patient recovered completely from leukopenia and/or opportunistic infection, the regimen was restarted. Patients were considered evaluable for response when the duration of therapy was ≥10 weeks; they were considered evaluable for toxicity after they were treated for ≥2 weeks. Patients were evaluated for clinical status and hematologic, immunological (CD4-CD8 count), and virological (p24 plasma level, HIV viremia) parameters every 15 days until the 6th week, then every 30 days thereafter. In three patients (Table 1), plasma viremia was noted until the 8th week in two cases and the 18th week in the thrid case. Briefly, plasma viremia was evaluated using an RT-PCR competitive approach involving co-reverse transcription and subsequent co-amplification of the sample with in vitro transcripted competitor RNA that had the same recognition size of the most common HIV-1 strains and similar length and base pairs composition (6). In all patients, the clinical stage was determined according to CDC classification 1993 (7). For the AIDS diagnosis, following WHO guidelines, only CDC clinical criteria were adopted (8). Toxicity was evaluated according to WHO criteria (9). The following response criteria were monitored for all patients: increase of CD4 count/mm3 >50% of baseline values and a significant decrease (>0.7 log) of HIV viremia (noted in seven of our patients). This clinical protocol was approved by the ethics committee of our institution, and all the patients gave oral informed consent. We enrolled 16 consecutive patients. Eleven were men and five were women. Their median age was 37.5 years (range, 30-66 years). Six patients were intravenous drug users, eight were homosexual men, and two were heterosexuals. Two patients had been previously treated with zidovudine alone; 11 patients with zidovudine and then didanosine, two with zidovudine alone and then with zidovudine plus interferon-α, one with concomitant zidovudine and didanosine. At study entry, two were asymptomatic (A3), five had HIV-related disease (one B2, four B3) and nine had a previous diagnosis of AIDS (one C2, eight C3). The median baseline CD4 count was 24.5 cells/mm3 (range, 2-365), corresponding to a median percentage of 5.8% (range, 0.3-19%). The median duration of therapy was 10 weeks (range, 2-43 weeks). All 16 patients were evaluable for toxicity. At study entry, the median WBC count (cells/mm3) was 3260 (range, 1660-8820), the median platelets count (ptl/mm3) was 156,000 (range, 53,000-202,000), the median Hb value (g/dl) was 11.75 (range, 7.2-14.9). The therapy nadir median values were 1905 (range, 680-3640), 107,500 (range, 49,000-157,000), and 10.85 (range, 8.1-13.6), respectively. It is worth noting that the median WBC values were <4000/mm3 and that seven patients had WBC counts <3000/mm3 at study entry. These data could explain the high rate of hematologic toxicity that we observed during treatment: two G1 (12%), five G2 (31%), eight G3 (50%), one G4 (6%). Six patients were withdrawn before the 10th week because of severe leukopenia and opportunistic infections. Overall, 10 patients were evaluable for response. Before therapy, the median CD4 count of the patients evaluable for activity was 51 cells/mm3 (range, 2-365) corresponding to a median percentage of 6.35% (range, 0.3-19). At the end of therapy, the median values were 14 cells/mm3 (range, 1-191) and 4% (range, 0.3-19.4), respectively. No patient had an increase in the absolute number of CD4 cells at the end of therapy (Table 1). A steady-state level of plasma viremia with a maximum five-fold positive or negative fluctuation among the several determinations was observed in seven patients, and none of them showed a significant decrease in plasma viremia at the end of therapy (Table 1, Fig. 1). Our data demonstrated that hydroxyurea is not an effective antiretroviral drug in heavily pretreated and immunologically compromised HIV-infected patients. In fact, it does not provide any improvement of immunological status and does not significantly decrease the HIV-RNA plasma titer. Nevertheless, we believe that further investigations are needed to evaluate hydroxyurea in patients with less advanced HIV disease and lower immunological impairment, particularly when HU is administered in association with other antiretroviral drugs. Acknowledgment: Dr. Manola Comar was supported by a fellowship from IRCCS Burlo Garofolo, Trieste, Italy. Further support was received from ISS grants (U.T.), VIIIth AIDS project. We also thank Mrs. Daniela Furlan for her help with the manuscript. Cecilia Simonelli; Guglielmo Nasti; Emanuela Vaccher; Umberto Tirelli Division of Medical Oncology and AIDS Stefania Zanussi; Paolo De Paoli Microbiology-Immunology-Virology Division; Centro di Riferimento Oncologico; Aviano, Italy Manola Comar; Mauro Giacca International Centre for Genetic Engineering and Biotechnology; Trieste, ItalyFIG. 1: . Plasmatic viremia was determined by RT-competitive PCR. The X-axis shows the number of weeks after beginning HU therapy. The Y-axis shows the number of HIV-RNA molecules/ml of plasma (logarithmic scale). Patients symbols: □ no. 1, ♦ no. 3, ⋄ no. 4, ▵ no. 6, • no. 7, ○ no. 10. White rectangles at the bottom of X-axis show the number of weeks of discontinuous therapy for patient 4 and patient 10.