Treatment of tuberculosis in HIV-infected patients: safety and antiretroviral efficacy of the concomitant use of ritonavir and rifampin

Shortly after the introduction of protease inhibitors (PI), some experts called attention to the difficulties of treating tuberculosis in HIV-infected patients who were receiving these new drugs ^[1]. PI and the rifamycins share the CYP450 metabolic pathway, leading to significant pharmacological interactions that preclude the co-administration of the two classes of drugs. Early recommendations formally contraindicated the use of rifampin together with any of the PI, because rifampin decreases by more than 80% the blood plasma levels of most of these drugs ^[2].

The same recommendations suggested rifabutin as an adequate alternative to treat tuberculosis in HIV-infected patients taking PI ^[2]. This was as a result of the observation that rifabutin decreases by approximately only 32% the blood plasma levels of indinavir and nelfinavir ^[3,4]. Surprisingly, the level of interaction between rifampin and ritonavir is similar to that of rifabutin and indinavir or nelfinavir ^[5], but no consideration was given to the co-administration of the two drugs.

This is a pilot, non-randomized, open-label study in which the co-administration of rifampin and ritonavir in HIV-infected patients with tuberculosis was evaluated. Patients were recruited at seven Spanish hospitals from May 1998 to July 1999. Subjects included were HIV-1-seropositive adults with active tuberculosis who had received antituberculosis treatment for less than 2 months. Eligible patients were PI naive. After the diagnosis of tuberculosis, patients were treated with only antituberculosis drugs (rifampin 600 mg/day, isoniazid 300 mg/day, and pyrazinamide 25 mg/kg/day) during the first 2 months. Then, patients received antituberculosis and antiretroviral therapies concomitantly. As antiretroviral treatment, patients received two nucleoside analogues plus ritonavir (600 mg twice a day).

After baseline evaluation, patients were seen at weeks 4, 8, 12, 20, 28, and 40. At each visit, complete history and physical examination were performed. Blood samples were drawn for routine haematology and chemistry determinations, CD4 cell count, and HIV-RNA plasma levels. The trough (C_min) plasma levels of ritonavir and rifampin were measured at each visit by high performance liquid chromatography and reverse phase high performance liquid chromatography, respectively. A chest X-ray and sputum examination for mycobacteria were performed at week 12, or when clinically indicated.

Eighteen patients were eligible to participate in the study. Seventeen patients had AIDS before the development of tuberculosis. The median CD4 cell count and HIV-RNA plasma levels were 79 cells/mm³ and 6.04 log₁₀, respectively. Of the 18 patients, 13 were naive for antiretroviral drugs. Tuberculosis was disseminated in 11 patients, and 14 had Mycobacterium tuberculosis isolated from culture.

Ten patients left the study before week 8: six developed toxicity (in all cases, intolerance to the oral solution of ritonavir), one was lost to follow-up, two withdrew consent, and one patient died after the development of toxoplasma encephalitis. For the purposes of analysis, all 18 patients will be included in the analysis of safety, whereas only the eight patients who continued in the study after week 8 will be considered for efficacy analysis.

There was a significant reduction in plasma HIV-1 RNA in all the patients. Viral load decreased a median of 3.72 log₁₀, from 6.01 log₁₀ at baseline to undetectable levels (< 200 copies/ml) at week 28. All patients showed an increase in the CD4 cell count. The CD4 cell count increased a median of 173 cells/mm³, from 85.5 to 286 cells/mm³ at week 28, and a median of 225 cells/mm³ at week 40. Tuberculosis was adequately controlled in all eight patients who completed therapy. Clinical evaluation and chest X-rays showed a complete resolution of the disease between weeks 12 and 20 in all the patients.

The C_min of ritonavir had wide variations from 0.35 to 9.67 μg/ml (median 2.22 μg/ml), but it was well over the IC₉₀ for ritonavir at most time-points (Fig. 1). Rifampin levels were within normal limits in all patients. Trough levels of the drug ranged from 0.77 to 21.01 μg/ml, with values greater than 1.58 μg/ml in 75% of the measurements.

Fig. 1

There were no unexpected side-effects, nor serious adverse events that could be related to ritonavir or rifampin. Five patients had an increase in transaminase levels. The increase was no greater than grade 2 (less than a threefold increase of the normal values), and was transient in all cases. There was an apparent association between concomitant chronic hepatitis C and increased transaminase levels. All four patients with hepatitis C developed hypertransaminasaemia, whereas it developed in only one of the four patients without hepatitis C.

As could be predicted from previous pharmacokinetic studies ^[5], our results show that the C_min of ritonavir are high enough to warrant therapeutic levels, and that the levels of rifampin are not as high as to increase toxicity. This was reflected in an adequate virological response, a significant increase in CD4 cell counts, and the lack of significant toxicity attributable to the study drugs. The efficacy results are similar to those observed in other studies in advanced HIV-infected patients treated with full doses of ritonavir.

We were concerned with toxicity issues, particularly liver toxicity. Both ritonavir and the antituberculous drugs (rifampin, isoniazid) have the potential to cause severe liver toxicity that could potentiate each other. We observed only a mild, transient increase in the transaminase levels in five of the eight patients who completed the study, mainly in patients with chronic hepatitis C. As a consequence, no precautions seem to be necessary in patients who receive ritonavir and rifampin concomitantly, in addition to the usual monitoring of liver function tests.

Many patients who develop tuberculosis will have to take PI. The updated guidelines have included the combination of rifampin and ritonavir as one of the alternatives for the treatment of tuberculosis among HIV-infected patients ^[6], although no new clinical data have been originated since the previous recommendations. The results of this study may thus be important, especially after some recent clinical studies have shown significant limitations for the use of rifabutin in combination with indinavir or nelfinavir that could preclude their concomitant use ^[7]. New studies with different combinations of antituberculous and antiretroviral drugs should clarify the optimal regimens to be used.

Santiago Morenoa

Daniel Podzamczerb

Rosa Blázquezc

José A. Iribarrend

Elena Ferrerb

Jesús Reparaze

José M. Peñaf

Esther Cabrerog

Luis Usáng