Low risk of pancreatitis in HIV-infected patients on hydroxyurea plus didanosine.

Abstract

The perspective of more than 5 years of highly active antiretroviral therapy (HAART) shows that the dramatic clinical benefit of therapy should be counterbalanced by the impairment in the quality of life seen in patients on treatment. Drug toxicities and high pill burden often compromise the life-long benefit of therapy, and are among the main reasons for poor treatment adherence and, as a consequence, for treatment failure. Major attention has been paid in the past few years to the prescription of simpler regimens, mainly based on the substitution of protease inhibitors by either nucleosides or non-nucleoside analogues. Overall, these so-called simplified regimens are able to sustain virus suppression with fewer drug-related disturbances [1]. Structured treatment interruptions represent an alternative approach, but the control of virus replication by eliciting a more robust anti-HIV cellular response is rarely obtained in patients with chronic HIV infection, even after long periods of complete virus suppression on HAART [2]. Between the previous two models of simplification therapy (one using three drugs and another with intermittent drug holidays), we have shown that the combination of hydroxyurea plus didanosine, as maintenance therapy after complete virus suppression on HAART, may provide long-term control of virus replication (less than 5000 HIV-RNA copies/ml) in approximately half of the patients [3]. This regimen is remarkably simple (three pills a day) and has little impact on quality of life. Although in our initial report we noted that didanosine plus hydroxyruea was generally well tolerated, two subsequent studies have stressed the risk of pancreatitis in patients exposed to this combination [4,5]. One study [4] claimed that pancreatitis may appear in 6.25% person-years, and another reported two cases of fatal pancreatitis [5], although an unacceptable delay in diagnosis and stopping therapy after patients’ complaints occurred in these cases. Here we report the rate of pancreatitis in a group of 114 HIV-positive individuals who began didanosine plus hydroxyurea after maintaining an HIV-RNA level below 50 copies per millilitre under a triple combination for at least one year. The replacement of HAART by the simplest didanosine plus hydroxyurea regimen was intended to provide an easier and less toxic treatment. As is well known, hydroxyurea enhances the antiviral activity of didanosine and provides some favourable immune effect [6]. Data collected after one year on didanosine plus hydroxyurea were analysed. Most subjects were men (83%). Their mean baseline CD4 cell count was 585 ± 134 cells/μl. The mean triglyceride level was 231 ± 207 mg/dl and the mean amylase value was 63 ± 26 IU/l, with two subjects having asymptomatic hyperamylasaemia (normal upper limit 123 IU/l). Eight subjects underwent treatment discontinuation before completing one year on didanosine plus hydroxyurea: three developed skin side-effects related to hydroxyurea, one a distal symmetric neuropathy, another thrombocytopenia, and three complained of abdominal discomfort, only one of which could be attributed to mild pancreatitis (amylase peak of 271 IU/l). Another 10 subjects showed mild elevations in amylase levels (average 157 ± 42 IU/l) without symptoms. These results place the incidence of pancreatitis in subjects receiving didanosine plus hydroxyurea at 0.87% person-years, which is in the range of what is expected for didanosine monotherapy [7]. One possible explanation for such a discrepant result may be related to the higher baseline CD4 cell count among our patients in comparison with those in the study by Moore et al. [4] (mean CD4 cell count 275 cells/μl). We found a significant correlation between the risk of hyperamylasaemia and the baseline CD4 cell count. Alternatively, it should be borne in mind that a past history of pancreatitis was considered an exclusion criteria in our simplification trial. The clinical significance of low amylase levels is unknown, but we noted a rate of nine cases per 100 person-years of asymptomatic hyperamylasaemia. On the other hand, a significant reduction in the rate of hypertriglyceridaemia was recorded in our patients switching to didanosine plus hydroxyurea (from 40% at baseline to 18% one year thereafter;P < 0.05). Hypertriglyceridaemia is a risk factor for pancreatitis, and its frequency in patients taking didanosine plus hydroxyurea as part of a triple combination as used in the above-mentioned trials [4,5] might be higher than the rate recorded in our cohort of patients on single dual didanosine plus hydroxyurea. In conclusion, the association of didanosine plus hydroxyurea seems to be safe as a maintenance regimen in individuals with good immunological status, although the close monitoring of amylase levels and abdominal symptoms is warranted. However, this combination should be avoided in patients with low CD4 cell counts because of the well-known cytostatic effect of hydroxyurea, which may worsen the reduced CD4 cell count, and the potential higher incidence of pancreatitis. Pablo Barreiro Vincent Soriano Eulalia Valencia Beatriz Díaz Juan González-Lahoz