Article Text

Download PDFPDF

HIV-related TB and adverse drug events
  1. Chi Chiu Leung1,
  2. Chi Kuen Chan1,
  3. Cheuk Ming Tam1
  1. 1Tuberculosis and Chest Service, Centre for Health Protection, Department of Health, Hong Kong SAR, China
  1. R A M Breen2,
  2. R F Miller2,
  3. M C I Lipman2
  1. 2Royal Free Hospital, Pond Street, London NW3 2QG, UK; rambreen{at}

Statistics from

Breen and coworkers1 showed that, in the era of effective antiretroviral therapy, discontinuation of anti-tuberculosis (TB) treatment occurred with a similar frequency in HIV-infected and HIV-uninfected individuals despite a greater rate of serious (grade III/IV) adverse events among HIV-infected individuals.

According to the Division of AIDS table for grading the severity of adult and pediatric adverse events (, grade III adverse events are likely to cause inability to perform usual social and functional activities while grade IV adverse events are potentially life-threatening. However, among HIV-infected patients with grade III/IV adverse events in the above study,1 treatment was interrupted only in a minority of patients, except for those with hepatotoxicity, and no mention was made regarding any modification of treatment regimens. With the retrospective study design, it might be difficult to exclude some degree of subjective bias in symptom reporting/grading/interpretation, especially among HIV-infected individuals, despite the use of a standardised grading scheme.

In contrast with previous studies,2,3 anti-TB drug-related hepatotoxicity was observed at a similar rate in HIV-infected and HIV-negative patients.1 Differing abilities to control sociodemographic and clinical profounders—such as malnutrition, alcohol use, drug abuse, hepatitis B/C, anti-retroviral drugs—could account for the difference, especially with the limited sample sizes of these studies.1,3 In this regard, it is interesting to note that use of rifampin plus pyrazinamide in the treatment of latent TB infection was associated with apparently higher prevalences of hepatotoxicity in clinical trials conducted among HIV-negative subjects4 than those conducted among HIV-infected individuals.5 As hepatotoxicity is a major factor leading to interruption of anti-TB treatment,1 the similar incidence of hepatotoxicity in HIV-infected and HIV-negative patients is perhaps reassuring.

However, while the attending clinicians might be unwilling to interrupt the anti-TB treatment among HIV-infected subjects even in the face of severe vomiting and peripheral neuropathy,1 patient cooperation could be jeopardised and drug adherence would then be difficult to ensure outside the setting of directly observed therapy. Non-adherence, frequent regimen modifications and treatment interruptions certainly increase the risk of treatment failure and relapse with acquired resistances. With the recent report of highly fatal cases of drug resistant TB among HIV-infected patients,6 there remains a need for heightened awareness of possible adverse drug events, as well as vigilance in the prevention, detection and management of such events.


Author’s reply

We thank Dr Leung and colleagues for their interest in our data1 and their insights into the management of individuals with HIV and tuberculosis (TB) co-infection. We agree that, as with any retrospective study, it is possible that some events may have been misclassified despite our best efforts. However, we would be surprised if this were the case for hepatotoxicity, which is measured by objective blood test results, or treatment interruption.

Highly active anti-retroviral therapy (HAART) has radically altered the management of HIV and TB co-infection. Our overall aim was to describe the occurrence of adverse events and treatment interruption in this era. Data drawn from before this may not serve as a valid comparator. We looked carefully for differences according to anti-retroviral usage but none were observed. As highlighted in our discussion, there do exist a number of other factors that are difficult to control and may account for differing results between studies. Our data suggest a role for ethnicity, which might explain the divergent results seen between populations. As Dr Leung mentions, differences in event rates according to HIV infection with rifampicin and pyrazinamide combination are intriguing. It has been postulated that this reflects immune function.2 We found no evidence for this when we analysed our event rates according to either baseline blood CD4 count or changes in this value at 2 months.

We agree that maintaining patient cooperation and adherence with appropriate drug regimens is vital to outcome for both TB and HIV, especially when the management of both conditions may be complicated by the development of drug resistance. However, in our cohort we observed reassuringly high levels of TB treatment completion and low rates of TB recurrence (regardless of HIV infection), as well as excellent virological responses to HAART.3


View Abstract

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.