The treatment of tuberculosis in patients infected with HIV can be challenging, as a result of the overlapping toxicities of medications, the large pill burden, and drug–drug interactions. When patients are receiving highly active antiretroviral therapy (HAART), the use of rifabutin instead of rifampin is recommended to minimize these interactions [1]. A dosage adjustment of rifabutin is often necessary, depending on the specific antiretroviral agents being used [1]. However, even when these dosing adjustments are followed, unpredictable drug interactions can still occur. We report the case of an HIV-positive patient with persistent tuberculous adenitis who had sub-therapeutic rifabutin levels despite megadoses of the drug because of a drug interaction with efavirenz. A 39-year-old HIV-positive Kenyan man, whose CD4 cell count was 120 cells/mm3 and viral load was 117 000 copies/ml had a positive purified protein derivative test. Chest radiograph was normal and he was given 9 months of isoniazid. The patient left the country and returned one year later, and confirmed that he had completed a full 9 months of isoniazid. The patient was clinically well, without diarrhea, and his CD4 cell count and viral load were essentially unchanged. Once daily treatment with efavirenz 600 mg, tenofivir 300 mg, and lamivudine 300 mg was started. Two weeks later a new left-sided cervical node appeared. A Gram stain of aspirated material from the node was negative for acid-fast bacilli, but mycobacterial cultures revealed pan-sensitive Mycobacterium tuberculosis. Daily treatment with isoniazid 300 mg, rifabutin 450 mg, ethambutol 1200 mg and pyrazinamide 1500 mg was started, as well as prednisone 40 mg. Prednisone was tapered over time as the node stabilized or lessened in size, and increased when the lymph node enlarged. Over the next 6 months, the node persisted and a new right-sided node appeared. Drug levels of isoniazid and rifabutin, drawn at 1.5 and 2 h after ingestion, respectively, were subtherapeutic (rifabutin < 0.1 μg/ml, isoniazid < 0.5 μg/ml ). The dosages of rifabutin and isoniazid were increased incrementally, and by month 11 the patient was receiving daily rifabutin 1350 mg, isoniazid 750 mg, ethambutol 900 mg, pyridoxine 50 mg, and prednisone 15/10 mg. Rifabutin levels were still low (0.1 μg/ml), whereas isoniazid levels were considered to be therapeutic (3.5 μg/ml). Three stool samples obtained for ova and parasites were negative. Because of persistent scrofula and a suspected drug–drug interaction, efavirenz was stopped and nevarapine was started. Fifteen days later the patient developed iritis, with higher levels of rifabutin (0.6 μg/ml). The dosages of rifabutin and isoniazid were decreased, and within two additional months the adenopathy had completely resolved. His CD4 cell count at that time was 128 cells/mm3 and his viral load was less than 75 copies/ml. Daily prednisone 10 mg was slowly tapered off. Efavirenz is used commonly in many HAART regimens because of its potency and convenience. It is known to induce the cytochrome P450 system, in particular the CYP3A4 isoenzyme [2]. This interaction was recognized by the Centers for Disease Control and Prevention and prompted the recommendation to increase rifabutin dosages from 300 to 450–600 mg/day [1], as was done initially with our patient. Even when the rifabutin dosage was increased to 1350 mg/day, however, low drug levels still occurred. Although the patient was receiving other antiretroviral agents, the putative effects of efavirenz was well demonstrated when the substitution of this agent by nevarapine resulted in a sixfold increase in rifabutin levels. In addition to the effects of efavirenz on rifabutin, a drug interaction between efavirenz and isoniazid may also have occurred in our patient. Alternatively, decreased isoniazid may have occurred, as patients with advanced HIV may have decreased drug absorption and low drug levels [3]. Poor drug absorption could also explain the previous failure of isoniazid in our patient when he was not receiving HAART, although this issue was not addressed in the laboratory. Because of the issues of tuberculosis drug interactions and poor absorption, there is growing interest in therapeutic drug monitoring (TDM) in HIV-positive patients, whether or not they are receiving HAART. However, TDM is not routinely performed in the community setting. The issue of TDM is further confounded by the fact that it is unclear at exactly what level isoniazid and rifabutin are effective or ineffective [4]. Furthermore, most HIV-positive patients treated for tuberculosis have a complete response to tuberculosis treatment without TDM. On the basis of our case report and the experience with an additional patient not reported above, we believe that efavirenz should be used cautiously when treating patients with tuberculosis. When efavirenz cannot be avoided, TDM should be performed on most or all patients treated with HAART, with special attention to isoniazid and rifabutin levels.
Read full abstract
May 30, 2024
Ercan Yenilmez + 26
Open Access
