Rates of TB treatment interruption unaffected by HIV infection or HAART

HIV-positive patients receiving treatment for tuberculosis (TB) do not interrupt TB treatment more frequently than HIV-negative patients, although highly active antiretroviral therapy (HAART) does increase the incidence of side-effects, according to a case note review of patients at a London hospital. These findings were presented this week at the Seventh International Congress on Drug Therapy in HIV Infection in Glasgow.

TB is the most common opportunistic infection seen in HIV-positive patients worldwide. Recent studies estimate that 8% of TB cases occur in HIV-positive people, but that 13% of deaths caused by TB are in HIV-positive patients, and 11% of deaths in people with AIDS are due to TB.

Standard treatment for TB includes anti-bacterial drugs such as the rifamycins, rifampicin (Rifadin / Rimactane) or rifabutin (Mycobutin). However, these drugs can interact with anti-HIV medications, causing elevations in drug levels and side-effects. This study was carried out to assess the impact of HIV infection and HAART on the incidence of drug-related adverse events.

“Anti-TB medication has a similar adverse event profile in HIV-positive and HIV-negative individuals,” state the investigators. “HAART is associated with an increase in such events but these do not lead to more treatment interruptions.”

The investigators carried out a retrospective review of patients’ notes at the Royal Free Hospital in London between February 1997 and July 2003. They compared the incidence of adverse events between 115 patients with active TB and HIV and 114 HIV-negative TB patients treated over the same time period. Fifty-seven per cent of the patients had pulmonary TB, and 8% were resistant to isoniazid. Baseline demographic characteristics were similar in the two groups, with a median age of 34, 52% of the patients being female and 72% being black African.

Rifamycin-based therapy was used in 98% of the patients in both groups, for a median of 6 months. However, 5% of the HIV-positive patients were receiving more than five anti-TB drugs, compared to none in the HIV-negative group. Of the HIV-positive patients, 73% received HAART during their TB treatment, starting a median of 2 months after the initiation of TB therapy. The majority of these were on a non-nucleoside reverse transcriptase inhibitor (NNRTI)-based regimen (54%), with 31% on a protease inhibitor-based regimen and the remainder taking triple nucleoside therapy. Twenty-one per cent of the HIV-positive patients were receiving HAART before TB treatment was begun.

The rate of TB treatment interruption was similar in the two groups of patients (HIV-positive: 11%; HIV-negative: 9%), as was the rate of completion of TB therapy (96 vs. 87%).

Seven per cent of patients stopped HAART, although none of these were due to side-effects. However, HAART combinations were altered in 15 patients (18%), due to pill burden (four patients), peripheral neuropathy (three patients), abacavir hypersensitivity (three patients) and AZT (zidovudine, Retrovir)-mediated anaemia (two patients).

Although TB treatment interruption was similar in HIV-positive and -negative patients, the rates of grade 3 or 4 side-effects tended to be higher in the HIV-positive group. Peripheral neuropathy was seen in 14% of the HIV-positive patients and 4% of the HIV-negative. This was judged to be primarily due to an interaction between isoniazid and d4T (stavudine, Zerit) or ddI (didanosine, Videx / VidexEC) in patients on HAART, leading Dr Ronan Breen, presenting, to recommend avoidance of these drug combinations.

Rash was also more common in HIV-positive patients (9 vs. 3%), as was persistent vomiting (9 vs. 4%). However, elevated liver enzymes were seen in 7% of patients in both groups, with liver toxicity being responsible for comparable rates of TB treatment interruption in HIV-positive and –negative patients (62 vs. 80%).