The treatment of tuberculosis in HIV-infected persons

Introduction Tuberculosis (TB) is one of the most important infections affecting HIV-positive patients in the world. Rates of HIV-related TB have risen in countries in Europe, United States and South America [1,2], and the rates have increased so rapidly in India and the rest of Asia that they may equal those in sub-Saharan Africa by the year 2000 [1,3,4]. One in 11 cases of TB are attributed to HIV globally and will rise to one in seven by the year 2000 [5]. As therapy for HIV disease becomes more available, physicians need to know how to treat these two diseases effectively while minimizing the risk of drug interactions and maintaining the shortest possible duration of treatment for TB. Problems arise both when patients are contemplating starting antiretroviral therapy and when they are already on such therapy and TB is diagnosed. Some guidelines regarding the treatment of TB and HIV already exist [6] and usually require modification to either TB therapy or antiviral drugs. These interactions are problematic because rifampicin-based regimens are the gold standard for short-course TB therapy and should be used wherever possible. Major drug-drug interactions can occur especially between the rifamycins and protease inhibitors and non-nucleoside reverse transcriptase inhibitors (NNRTI). Unfortunately, there are little or no data on what are the optimal regimens when TB and HIV are being treated concomitantly, but priority should be given to treating and notifying TB [7]. This review will examine the evidence from clinical trials for the efficacy of rifamycin-based (rifampicin, rifabutin and rifapentine) short-course regimens for TB when rifamycin is used for the whole treatment period or only in the intensive phase. It will examine the drug-drug interactions between anti-TB and antiretroviral drugs and discuss possible TB regimens that might be used in HIV-positive patients on or starting antiretroviral therapy, and suggests best practice treatment strategies. TB chemotherapy Each of the main anti-TB drugs varies in its capacity to kill bacteria, prevent the emergence of drug resistance, and sterilize lesions. Isoniazid is the most potent bactericidal drug and kills more than 90% of bacilli within 7 days by acting on metabolically active bacilli. It is also quite effective at preventing the emergence of drug resistance. Rifampicin is also a good bactericidal drug, with a potent sterilizing effect, and the ability to prevent the emergence of drug resistance. In addition to acting on rapidly dividing bacilli, it kills so-called ‚persisters‚, which remain inactive for long periods but have intermittent periods of metabolism, with only short drug exposure. This is crucial to its sterilizing ability, and also confirms experimental studies of TB of both early bactericidal activity [8,9] and sterilizing activity [10-12]. Pyrazinamide, although bactericidal, is mainly used for its sterilizing effect. It is particularly effective at killing intracellular bacilli sequestered inside macrophages in an acid environment [12]. Ethambutol and streptomycin are less potent drugs, ethambutol probably only being bactericidal at high concentrations. They are less effective at preventing emergence of resistance to rifampicin and isoniazid. A fourth drug (such as ethambutol or streptomycin) can play a role in patients with an increased risk of drug resistance, such as in HIV-positive individuals, and must be included in the initial phase of treatment of such patients [7,13-15]. There is a close correlation between the power of a regimen to convert a sputum smear and culture-positive case to sputum culture-negative at 2 months and its overall sterilizing efficacy (i.e., the higher the culture-negative rate at 2 months, the higher the cure rate, or conversely, the lower the relapse rate after 6 months of total treatment). Six-month short-course chemotherapy is now the gold standard against which regimens have to be compared, and has the key elements of pyrazinamide for 2 months in the initial phase, and rifampicin in both the initial and continuation phases, whether the dosing schedule is daily throughout, daily for the initial phase and intermittently in the continuation phase, or intermittently throughout. Rifampicin Initial and continuation phases Table 1 shows some of the data on the efficacy of 6-month regimens of rifampicin. These data are for fully sensitive organisms. Any reduction below the 6-month total treatment gives an unacceptably high relapse rate [16,17]. The relapse rate is also significantly higher for isoniazid or streptomycin and isoniazid-resistant organisms [18].Table 1: Six-month short-course regimens.Initial phase only Table 2 shows data on studies where rifampicin was omitted from the continuation phase mainly on grounds of cost. Six-month regimens had inadequate sterilization and higher relapse rates even for sensitive organisms. To obtain the results quoted, some also required the first 2 months of the rifampicin-containing initial phase to be given in hospital. Some of these regimens would not be appropriate for HIV-positive individuals because they contain thiacetazone, but have been shown to be effective in International Union Against Tuberculosis and Lung Disease-assisted programme conditions [19]. There is however the problem of greatly increased toxicity to thiacetazone in HIV-positive patients in Africa [13], where such regimens have been used with great success. To overcome this, ethambutol may have to be substituted for thiacetazone [19]. When rifampicin is not used in the continuation phase there is also a significantly increased failure rate if the organism is found to have initial isoniazid resistance [18].Table 2: Rifampicin in initial phase only.Rifabutin instead of rifampicin in standard regimens Rifabutin, a different rifamycin from rifampicin, has been used for TB therapy. Experimental work suggests that rifabutin has a lower early bactericidal activity, by a factor of 2.73 compared with rifampicin [8]. The limited data for the use of rifabutin in controlled trials for treatment of TB are shown in Table 3. A further uncontrolled study of 50 HIV-positive cases in Uganda also showed similar efficacy [20]. Although superficially rifabutin seems to perform as well as rifampicin, not all patients in the trials have reached 24 months of evaluation after cessation of treatment, and there has been no information on how such regimens perform in the presence of initial isoniazid resistance, having only been tested on fully susceptible organisms [21,22].Table 3: Rifabutin/rifapentine for treatment of pulmonary tuberculosis.Rifapentine instead of rifampicin in standard regimens Rifapentine has recently been approved as part of combination treatment of pulmonary TB [23] when given weekly with isoniazid in the continuation phase (after 2 months of four-drug treatment with isoniazid, rifampicin, streptomycin and pyrazinamide). The potential advantages over rifampicin were its use once weekly in the continuation phase, and a better adverse events profile. The disadvantage was the higher bacteriological relapse rate (Table 3). Regimens containing neither rifampicin nor isoniazid Such regimens should really only be used for treating multidrug-resistant (MDR)-TB. Because by definition there is resistance to both rifampicin and isoniazid, both the main bactericidal drug (isoniazid) and the main sterilizing drug (rifampicin) have been lost, such regimens have to include multiple drugs, often second-line agents. Additional drug resistance often accompanies those of rifampicin and isoniazid. The treatment of such cases is prolonged, expensive, and success rates of only just over 55% have been reported [24]. Improved outcome has been reported in both HIV-negative and HIV-positive patients with MDR-TB when appropriate therapy was started promptly and maintained [25-27]. Although there have been no controlled studies of treatment for MDR-TB, several principles of treatment can be derived [28-30]. First, a single drug should never be added to a failing regimen, because to do so is likely to promote further drug resistance. Second, a drug regimen should include four and preferably five drugs to which the patient‚s organism has been shown to be susceptible on in vitro testing, and ideally which have not been used on the patient before, and should also include an injectable medication. Finally, the therapy of all patients with MDR-TB, both as an inpatient and as an outpatient, should be directly observed. To ensure that these criteria are met, it is best that only physicians with substantial experience of drug-resistant TB manage cases of suspected or proven MDR-TB. The role of rifabutin in the management of MDR-TB is controversial. First, there is the major problem that only one-third of rifampicin-resistant organisms retain susceptibility to rifabutin because of cross-resistance [31,32]. One open study showed a bacteriological response in up to 60% at dosages of 450-600 mg daily, but this may have been to other concurrent agents, particularly fluoroquinolones [33]. In a controlled study [34], rifabutin at 450-600 mg once daily was no better than including rifampicin, and no significant difference in outcome was noted. Length of treatment There is debate on how long patients with HIV disease with TB should be treated, and guidelines therefore vary. The British Thoracic Society recommend standard short-course TB treatment regimens [7], but one study [35] supported a 12-month total treatment period, which appears to be the basis of some recommendations. The apparently better results with 12-month regimens may be because, in the short term (e.g., at 18 months), reinfection is less likely in patients treated for 12 versus 6 months. If physicians wish to treat for prolonged periods, then the continuation phases should be extended accordingly for all the short-course regimens. Highly active antiretroviral therapy Several clinical endpoint studies have shown the benefit of highly active antiretroviral therapy (HAART) on morbidity and mortality from HIV disease [36-38]. Most HAART will result in HIV viral loads below the limit of detectability and substantial rises in CD4 cell count; these surrogate marker responses will, in most patients, translate into clinical benefit with decreased morbidity and mortality. Because TB can occur at any stage of immune suppression, patients may already be on antiviral therapy when TB treatment is started. Others may request or be advised to start concomitant anti-HIV therapy on the basis of the risk of developing further opportunistic infections or overall poor prognosis, and many guidelines indicate when antiviral therapy should be started [39,40]. According to the latest guidelines [41], potential choices of HAART regimens might be a protease inhibitor and two nucleoside reverse transcriptase inhibitors (NRTI), an NNRTI and two NRTI, or two protease inhibitors with or without NRTI. In clinical practice, many other HIV regimens are used, such as protease inhibitors and NNRTI, or triple NRTI. In the short-term, triple NRTI regimens using abacavir have also been shown to have marked effects on surrogate markers. It should be noted that for HIV-infected patients with TB, the viral load and CD4 cell count may be adversely affected by the TB itself and may not be an accurate guide to the patients‚ viral or CD4 cell count to therapy to many drugs used for and treatment of opportunistic infections occur with increased HIV-infected patients of the adverse to anti-TB therapy in patients with HIV the and use of rates of adverse in therapy to anti-TB therapy have been Several studies have compared both HIV-infected and patients with TB. These have shown higher rates of adverse in HIV-infected In a study by rates of adverse were in HIV-infected patients and in In a study by rates of adverse to anti-TB therapy were in HIV-infected patients and in those by In a study by of patients with TB with HIV adverse to only of those with TB were not with HIV adverse In all these studies the of adverse within the first 2 months of starting therapy, and rifampicin was for of adverse events in the study from used in has a significant adverse In the to isoniazid with in those versus in those than 50 It is also more likely in those with and and those are with rifampicin. have an increased of from the of which is by into may be at even risk of which a of that results in the of which may itself be in a patient an adverse to anti-TB therapy are to the therapy with close clinical and to therapy if the is and to once the has or to the In patients of anti-TB therapy, may be This is for HIV-infected patients are of There are of to rifampicin, ethambutol and isoniazid in of drugs, including rifampicin, isoniazid, and has been reported in with and and This has been to in drug and treatment In this the of drug-resistant TB has been reported The of should be in HIV-infected with or without or other clinical of to to therapy good or directly therapy. of of is interactions other than of the drugs used to treat have or of and with other drugs by with HIV These interactions have been in of rifampicin is by up to by of including and all of which an of In may which are less and less well and may rifampicin, rifabutin and are by and of results in of other drugs this of The effects of may be within a days starting a but more than 7 days to important drug interactions occur between rifamycins and several (Table for drug the of the rifamycin is to its the by its should be if any of these are used in combination with a and the of should be should also be given to drug in failing to on therapy if they are also a rifamycin as significant interactions between rifamycins and other by HIV-infected in the with Rifampicin the of by more than and the of by In the of rifampicin, the of this is by of rifampicin and the other may of rifabutin If rifampicin is given to an for of there is a reduction in this drug The of is by 90% if rifampicin is is by if rifabutin is In because itself it may in and the of rifabutin by up to Rifampicin the of by and its by Rifampicin also the of by Some patients, so-called poor have marked of when this drug is with rifampicin. The of this is not but it is to be a of rifampicin the of is usually with and may in rifabutin by of may do the The of these interactions is by may significant in the of and Isoniazid the of and its isoniazid may also have a similar on Ethambutol is both and by the from the is by with If given these drugs should be by 2 should be if other drugs are with such as with or There is a important between pyrazinamide and the of by its in the of should be if pyrazinamide is with because increased toxicity may be from the therapy The study of drug interactions between TB drugs and is a of As antiretroviral are and clinical evaluation in trials and clinical practice, so adverse have been noted and drug interactions have been In to of interactions between and other drugs that are in experimental or clinical many of drug-drug interactions are or are on In this significant drug interactions occur as a result of both and of As the rifamycins (rifampicin, rifabutin and rifapentine) are all by In vitro studies using that rifampicin is a more potent of than and both have a significantly higher potential than rifabutin In rifabutin has less than rifampicin that the or the of drugs used as HAART may be important in the of viral resistance. and The interactions between the rifamycins and antiretroviral therapy are shown in Table Rifampicin the and the of the of rifampicin increased of This is not significant and is not In rifabutin not to the of between rifampicin, rifabutin and antiretroviral the presence of the and in the or the which is with and in to this, the and of rifampicin is To the of the between and rifampicin, these two drugs should be at 2 and protease inhibitors inhibitors are both and inhibitors of is a inhibitor of and has the potential to of drugs this of of rifampicin or rifabutin results in of Rifampicin by and rifabutin by There are no data on the between rifamycins and and concomitant use be at has a similar drug to but is a more potent The result is that of drugs by to a than and also of are to a by such as rifamycins when compared with on If rifabutin is given with there are in of and in the of rifabutin In clinical practice, if these two drugs are increased of are (i.e., daily and mg once daily than mg once daily of rifabutin is Rifampicin is not used with because of its potential for and increased of has a on of and also and the potential for drug interactions A combination of is with rifamycins because of the significant in of rifabutin by of its by Such in of rifabutin are with a marked in adverse including and The combination of and rifampicin is because this would result in marked in rifampicin and marked in In be used in combination with rifabutin (i.e., more data are available, this combination be for clinical Rifampicin a in of and these drugs should not be used by In rifabutin by If the of should be increased to 1 daily and the of rifabutin to other protease is both a and inhibitor of It is that the of rifabutin is if it is with In studies in of in in rifabutin of between and In rifabutin in of of When was with rifampicin, the of was by these drugs should not be used If rifampicin is to be started in a patient already protease should be given to a 2 or days is after the protease or if the rifampicin is (e.g., in a the protease inhibitor is and rifampicin is at to after 1 In if including a protease is to be in a patient is already rifampicin and the rifampicin is to be to a should be for reduction of the activity of rifampicin to and NNRTI are also by and to a by is significantly increased by rifampicin, in a in of There is no in rifampicin and the It might be possible to overcome this by the of by to mg daily There are no data for at this There are also no data for but this may not have significant effects in clinical are by rifabutin and rifampicin, both of which of rifabutin there is a and with rifampicin a in In the of rifabutin is increased by if it is with the of adverse is increased in study in which was with of to than mg daily were required to similar effects to the of mg daily given to were not rifabutin in conditions were mg rifampicin for 7 is important to that the ability of rifampicin may than 7 this there was a in of and a in the of in rifampicin was not affected in was when with rifampicin, but there are no data on this The data on rifabutin showed no significant of rifabutin on but a in rifabutin by The rifabutin may need to be increased by if the drugs are to be used with no in interactions NRTI The NRTI and may all and there is potential for toxicity if isoniazid is The by which toxicity are but in clinical practice it may be to which drug is for should be given to patients isoniazid [7]. of isoniazid is by and also the rate of of isoniazid There is a between abacavir and isoniazid. drugs may be by Isoniazid may both as a for and is by both and in of isoniazid and a in abacavir may occur if they are because abacavir can be this is probably of little of protease inhibitors with isoniazid may result in potential for as a result of of isoniazid by the protease TB and HIV In patients with HIV have TB, there is no that the priority is to treat TB, especially in cases where there is an important patients with TB can have antiretroviral therapy concomitant with anti-TB therapy, but this to be The role of drug of both antiviral and anti-TB drugs in patient management to be should be and by physicians with experience in TB and should work with have the experience in using antiretroviral therapy. The best practice and most rapidly treatment for TB is short-course chemotherapy with rifampicin throughout. This also has the potential of the effects of TB on HIV by in vitro and in studies therapy also has the best and of evidence to it the gold for treatment regimens of TB from best to would be as standard short-course therapy with rifampicin (Table for 6 standard initial phase isoniazid, rifabutin in continuation phase for use rifabutin instead of rifampicin (Table for 6 standard initial phase and use continuation phase for 6 months (Table total of regimen regimen used, but for 18 months. The for and have that for HIV-infected patients protease inhibitors treatment for months with isoniazid and pyrazinamide be as an to regimens using rifamycins Because there have been data regarding drug-drug have for antiretroviral therapy concomitant with TB regimens into four possible of is on is to best on is from of The fourth with little data and be of these and are by the use of or weekly regimens for TB. Table Tuberculosis treatment regimens and antiviral patient is HIV-positive is on no antiretroviral therapy If the risk of of the HIV disease is not to be significant within the 6 months, then the patient should be given a rifampicin-based regimen If are required after this there should be a period of after anti-TB therapy. If the patient is to require antiretroviral therapy for HIV disease because the risk of within the months is TB regimen be given with a triple NRTI combination with a of antiviral therapy if after 6 months. a regimen for TB be given and the patient then be an or antiretroviral regimen with two NRTI. Finally, a NNRTI two NRTI in be given if rifabutin was used regimen 3). The antiretroviral regimen be after the TB treatment was A might be that the risk of of HIV disease was no given after 2 months and then started after 2 months, that the period was regimen a regimen be given for 6 months after this initial 2 months Finally, if it was to antiretroviral therapy that with rifampicin and then a regimen would have to be used throughout, the period of treatment to 18 months this be already on protease inhibitors TB The patient the antiretroviral therapy for 6 months and a rifampicin-based regimen given regimen or the antiretroviral therapy be for 2 months and the rifampicin-based regimen given for the intensive phase, but the continuation phase be one in which rifabutin or a regimen be given and protease inhibitor regimen 2 or would have to into this of especially the of the TB, viral CD4 cell count and the of these surrogate If the patient is on or they be treated with a regimen regimen 3). A patient on or should protease inhibitors to one of the patients on a protease inhibitor regimen be to a triple NRTI regimen and given TB regimen 1 or In they be given a regimen and rifabutin given as the basis of the anti-TB therapy. If they on or then TB regimen would have to be already on NNRTI The antiviral drugs might be or to triple NRTI for 6 months and TB regimen 1 They be for the first 2 months and regimen If rifabutin were used would have to be and to or or If the patient was on then this might also be but in be protease inhibitor or protease These regimens have been used for so-called months therapy would have to be given if the patient was on a protease inhibitor combination of or or and to on that therapy This be as good practice and any remain for should be of these those such as and require drug-drug studies to be TB can be There are no data for interactions with or These would probably be little affected or by rifampicin. The treatment of TB and HIV is problematic and of anti-TB and antiretroviral therapy. interactions are and more is required to therapy can be given to patients with TB, but the priority is to treat TB This more than the appropriate drug regimen, and and directly therapy, for are the of this but may well to the use of both TB and antiviral therapy drugs. The role of drug of anti-TB and protease NNRTI should be because this may be a in the management of these Because are being used more and especially in developing it is important that a treatment is that is on evidence than or The at for on the and to for in the