Drug sensitive tuberculosis (TB) is generally treated with a rifamycin-containing regimen, typically rifampicin (RMP) . This potent inducer of the hepatic cytochrome P450 (CYP) enzyme has significant drug interactions with combination antiretroviral therapy (ART) such as protease inhibitors (PIs) and non-nucleoside reverse transcriptase inhibitors (NNRTIs). An alternative rifamycin, rifabutin (RFB), has less effect on CYP and can be substituted for RMP when ART is co-prescribed .
RFB has been used successfully to treat non-HIV-infected individuals . However, there is little long-term outcome data on HIV coinfection . Systematic reviews suggest that despite its high cost, RFB may be cost saving in patients taking PIs .
When RFB is used with ritonavir-boosted PIs, it is generally given in low dose, intermittent regimen to compensate for both the increased drug concentration of RFB associated with concomitant PI use and the drug's long half-life. However, this has resulted in anti-TB treatment failure and relapse with acquired rifamycin resistance . A pharmacokinetic study demonstrated subtherapeutic concentrations of RFB and its active metabolite when combined with ritonavir-boosted PIs and given at current recommended dosages in HIV-infected subjects, with subsequent acquired rifamycin resistance .
Our centre has been using RFB as part of its anti-TB treatment strategy for several years. In view of these reports, we undertook a review of outcomes of our TB/HIV coinfected population who received rifamycin during their treatment for active TB.
We identified all adult HIV-infected individuals receiving a rifamycin between January 1997 and December 2008 at our HIV centre. TB was diagnosed if the patient was culture positive for Mycobacterium tuberculosis, or culture negative but nucleic acid amplification assay positive with clinico-radiological features and treatment response consistent with TB, or had histological findings and treatment response consistent with TB. Patients with known rifamycin resistance were excluded from assessment.
The duration of anti-TB treatment, whether the senior supervising clinician considered treatment was successfully completed, the occurrence of severe (grade III/IV) adverse drug reactions causing functional disability or requiring medical intervention  and the frequency of TB recurrence was identified using retrospective case note review.
Prescription of anti-TB medication and ART was at the discretion of individual physician using local treatment protocols. Drug dosage was in line with current British HIV Association guidance ; prescribed as self-administered treatment using dosette boxes and pill counts when required. RMP was used at a weight-based dosage of 450–600 mg once daily with concurrent NNRTIs and single agent ritonavir. RFB was prescribed at 450 mg once daily with efavirenz and 300 mg once daily with nevirapine NNRTIs and was given at a standard dosage of 150 mg thrice weekly with boosted PIs. Other anti-TB drugs were taken daily. As this was a clinical service evaluation, ethical approval was not required.
Subgroup analysis of population characteristics by rifamycin and ART use was undertaken using χ2-test and Fisher's exact test.
One hundred and forty-one TB/HIV coinfected patients received rifamycin-based anti-TB treatment. Their median age was 36 years [interquartile range (IQR) 32–43], 52% were women and 72% black African. One hundred and three (72%) had positive culture results for M. tuberculosis (four with isoniazid drug resistance). Pulmonary TB was demonstrable in 58 (41%) patients. Median blood CD4 cell count prior to anti-TB therapy was 141 cells/μl (IQR 51–272).
One hundred and six (75%) used ART at any time during anti-TB therapy. Forty-one (29%) were already receiving ART, whereas the median ART start time for the others once established on anti-TB treatment was 2 months (IQR 1–3). RMP was given to 86 patients (60% received concomitant ART) and RFB to 79 (99% received concomitant ART). Twenty-four patients initiated treatment with RMP and then switched to RFB (in two-thirds due to the introduction of a PI-based ART regimen).
Six different ART regimens were used in 106 patients: 68 (64%) were PI-based (including one patient on PI monotherapy), 36 (34%) were NNRTI-based, 1 (1%) was NRTI-based and in one (1%) the specific ART regimen was unknown.
When RMP was used as the sole rifamycin during anti-TB treatment, 21 of 28 (75%) were prescribed NNRTI-based ART and seven of 28 (25%) PI-based ART. In contrast, 44 of 54 (81%) patients prescribed RFB used a PI-based regimen and 10 of 54 (19%) NNRTI-based ART, P < 0.001. When patients switched from RMP to RFB, 17 of 24 (71%) were prescribed PI-based, five (21%) NNRTI-based and one (4%) triple NRTI-based ART.
Patients on ART + RMP or who switched rifamycin had evidence of a trend towards a longer median duration of anti-TB treatment, P = 0.06 (Table 1).
Severe (grade III/IV) adverse events occurred in 39 of 106 (37%) patients who did and 13 of 35 (43%) did not receive ART during anti-TB treatment, P = 0.21 (Table 1). There was no significant difference in the overall number of patients experiencing at least one severe adverse event according to type of rifamycin and concurrent ART usage, P = 0.58.
The frequency of individual severe adverse events was similar between rifamycins for all but arthralgia [present in seven of 13, 54%, receiving RMP without concurrent ART compared with RMP or RFB with ART (1/13, 8% and 2/13, 15%, respectively), P = 0.002] (Table 1). No patient had to modify therapy, though 11 of 141 (8%) briefly interrupted treatment due to hepatotoxicity. This was unrelated to rifamycin type or concurrent ART usage.
There was no significant difference in the proportion completing anti-TB therapy using different combinations of rifamycin and ART, P = 0.53 (Table 1). Six of 141 (4%) patients were lost to follow-up and two (1%) died during anti-TB therapy. The remaining 133 (94%) patients were considered cured of TB by their treating physician.
After a median follow-up from TB treatment completion of 64 months (IQR 29–98), four of 133 (3%) patients relapsed – all with a drug-sensitive organism. Of these, only one had received RFB as part of anti-TB therapy (Table 1).
In our retrospective cohort study, at the current recommended dosing, when compared with RMP, RFB does not appear to be associated with an excess of severe adverse events or an increased risk of relapse or subsequent acquired rifamycin resistance. Other data have suggested that acquired rifamycin resistance is more likely with intermittent RFB dosing. However, Burman et al. used a variety of treatment regimens, including a less frequent rifamycin schedule than reported here, and, in fact, found the use of ART to be protective. The carefully performed study by Boulanger et al.  in predominantly African–Americans, noted that RFB, and its metabolite, levels were low using intermittent dosing. Of the 10 patients, one (weight 118 kg, BMI 34.3 kg/m2) developed acquired rifamycin resistance. Although the numbers were small and follow-up was not reported (all were regarded as cured at treatment completion), it is possible that therapeutic failure reflected the patient's high BMI and would be less likely in average-sized individuals. Also, other anti-TB drugs appear to have been given thrice weekly, rather than daily, which is our practice.
In our study, almost all patients using RFB were on ART (predominantly PI). Yet, severe adverse events were no more frequent – implying that these agents can be safely combined in routine clinical practice.
Although our dataset is retrospective, reasonably small in size, and drug levels were not measured, we believe that it lends support to the wider use of RFB in TB/HIV-coinfected patients. It is hoped that its recent price reduction will improve access to this promising therapeutic option in resource-poor areas with high rates of co-existent TB and HIV (http://www.clintonhealthaccess.org/news-and-information/arv-price-reductions-faqs-august-2009).
R.S. and M.C.L. had the original idea. R.S., N.M., C.J.R., R.A.B. and L.S. contributed to data collection. R.S., N.M., C.J.S. and C.J.R. contributed to data analysis. R.S., N.M., C.J.S., C.J.R., R.A.B., S.B., I.C., S.H., L.S., M.A.J. and M.C.L. contributed to manuscript preparation and writing.
Conflicts of interest
N.M. – financial relationships for payment for lectures (Janssen, London School of Pharmacy, London Middlesex University, BMS) and travel/accommodations/meeting expenses unrelated to activities listed (Janssen, BMS, Boehringer, Abbott).
C.J.S. – grants/grants pending (Bristol Myers Squibb).
S.B. – financial relationship for payment for lectures (BMS, MSD, Gilead).
L.S. – payment other (Bristol Myers Squibb, Gilead, VIIV, Janssen, Abbott).
M.A.J. – financial relationships for consultancy, payment for lectures (BMS, Abbott, VIIV, Gilead, MSD) and payment for development of educational presentations (BMS, Abbott, Janssen).
R.S., C.J.R., R.A.M., I.C., S.H. and M.C.L. – none.
1. National Institute for Health and Clinical Excellence. Tuberculosis. Clinical diagnosis and management of tuberculosis and measures for its prevention and control. CG117
. London: National Institute for Health and Clinical Excellence; 2011.
2. Pozniak AL, Coyne KM, Miller RF, Lipman MC, Freedman AR, Ormerod LP, et al. British HIV Association guidelines for the treatment of TB/HIV coinfection 2011
. HIV Med
3. McGregor MM, Olliaro P, Wolmarans L, Mabuza B, Bredell M, Felten MK, et al. Efficacy and safety of rifabutin in the treatment of patients with newly diagnosed pulmonary tuberculosis
. Am J Respir Crit Care Med
4. Davies GR, Cerri S, Richeldi L. Rifabutin for treating pulmonary tuberculosis
. Cochrane Database Syst Rev
5. Loeliger A, Suthar AB, Ripin D, Glaziou P, O’Brien M, Renaud-Thery F, et al. Protease inhibitor-containing antiretroviral treatment and tuberculosis: can rifabutin fill the breach?
. Int J Tuberc Lung Dis
6. Burman W, Benator D, Vernon A, Khan A, Jones B, Silva C, et al. for the Tuberculosis Trials ConsortiumAcquired rifamycin resistance with twice-weekly treatment of HIV-related tuberculosis
. Am J Respir Crit Care Med
7. Boulanger C, Hollender E, Farrell K, Stambaugh JJ, Maasen D, Ashkin D, et al. Pharmokinetic evaluation of rifabutin in combination with lopinavir-ritonavir in patients with HIV infection and active tuberculosis
. Clin Infect Dis
8. AIDS Clinical Trials Group. Table for grading severity of adult adverse experiences
. Rockville, MD: Division of AIDS, National Institute of Allergy and Infectious Diseases; 1996.