University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Impact of tuberculosis cotreatment on viral suppression rates among HIV-positive children initiating HAART
Global Health Sciences Literature Digest
Published January 18, 2011
Journal Article

Zanoni BC, Phungula T, Zanoni HM, France H, Feeney ME. Impact of tuberculosis cotreatment on viral suppression rates among HIV-positive children initiating HAART. AIDS. 2011, Jan 2;25(1):49-55.

In Context

Treating patients with HIV and tuberculosis (TB) co-infection can be challenging because of overlapping toxicities associated with HIV and TB therapies.(1) Despite recommendations not to use ritonavir and lopinavir/ritonavir in patients receiving anti-TB treatment, the limited number of antiretroviral therapy (ART) options in South Africa has resulted in national treatment recommendations to use lopinavir/ritonavir as first-line ART for children under three years.(2)


To evaluate the association between HIV virologic failure and treatment of HIV-TB co-infection in children.


Two clinical sites that provide treatment for HIV and TB in children located in KwaZulu Natal, South Africa.

Study Design

Retrospective cohort.


HIV-infected pediatric patients (<18 years) who initiated ART at one of the clinical sites between May 2004 and December 2008.


HIV suppression (plasma RNA PCR <400 copies/ml at six and 12 months after treatment initiation).


Data were abstracted from paper and electronic medical records. The standard of care calls for CD4, viral load, and weight monitoring every six months. ART is started for children with WHO clinical disease stage three or four and/or CD4 percentage less than 15 in children under 18 months of age.(3) Because TB therapy can reduce serum concentrations of lopinavir/ritonavir, children who received both ART and anti-TB medications had their dose of lopinavir/ritonavir increased.(2) Children over three years and over 10 kg were started on an NNRTI-based regimen (efavirenz, stavudine, and lamivudine). Due to a change in documentation procedures, the medical records from one of the sites prior to 2007 were not included.

Patients who were receiving anti-TB treatment at the time of ART initiation or who initiated anti-TB treatment within six months of ART initiation were defined as "co-treated'. Children with a history of TB in the past or who had never been diagnosed with TB were defined as "not co-treated'. Additional variables included in the analysis were age at beginning ART, gender, baseline viral load (prior to starting ART), baseline CD4 count and percentage, ART regimen, and weight for age z scores.

Children with pulmonary TB receive six months of rifampin, isoniazid, and pyrazinamide. Children with TB meningitis, military or extrapulmonary TB receive nine months of these three drugs plus ethionamide.(2)


A total of 1029 children under 18 years of age received ART at the study sites. The median age at ART initiation was 5.1 years. Two thirds of the patients were started on an NNRTI-based regimen and the others received a protease inhibitor (PI) based regimen. The median CD4 cell count was 283 and median percentage was nine percent at the time ART was started.

Complete baseline data and viral load results at six months were available for 719 (70%) of the 1029 patients and were included in the analysis. The reasons for exclusion were death within six months of initiating ART: 36 (3.5%); loss to follow-up: 24 (2.3%), change of service provider: 47 (4.6%); less than 6 months on ART: 8(0.7%); transferred to another site: 61 (5.9%); no viral load after starting ART: 134 (13%). At 12 months, 555 (53.9%) of patients had data to evaluate. The reasons patients were excluded at 12 months were death: 46 (4.5%) deaths; loss to follow-up: 34 (3.3%), change of service provider: 72 (7%; on treatment for less than 12 months: 195 (19%); and did not have a viral load after initiating ART 127 (12.3%).

There were 609 (59%) children with a history of TB of whom 373 (36%) were co-treated. Of the TB cases, 55 (14.6) had mycobacteria identified. Viral suppression occurred among 85% of children at six months and 87% at 12 months. The rate of viral suppression was significantly higher among children who received a NNRTI-based regimen compared to a PI-based regimen (91% and 71% respectively, p<0.0001). At 12 months there was a non statistically significant difference in the rate of viral suppression between children on the NNRTI-based regimen (89%) and children on the PI-based regimen (81%, p=0.38).

Viral suppression was significantly lower among TB-HIV co-treated patients than non co-treated patients at six months (79% and 88% respectively, p=0.003) but not at 12 months (86% and 90%, p=0.16). When stratified by treatment regimen, co-treated patients on the NNRTI-based regimen had similar rates of viral suppression as the non co-treated patients on the NNRTI-based regimen (88% and 92% respectively, p=0.14 at 6 months, 87% and 90% at 12 months, p=0.4). In contrast, viral suppression was significantly lower among co-treated patients on the PI-based regimens compared to non co-treated patients on PI-based regimens at both six and 12 months (63% and 78% respectively at six months, p=0.02 and 78% and 91% respectively at 12 months, p=0.03).


Viral suppression is reduced in patients co-treated for HIV and TB when using a PI-based regimen but not with NNRTI-based regimens.

Quality Rating

This was a poor study for several reasons. It is not evident how representative the subjects are. The proportion of subjects excluded was high. Also, as noted by the authors, ascertainment of TB disease is not entirely evident from clinic records. Lastly, the descriptions of the analyses were not always clear; this was particularly problematic for the multivariable analysis.

Programmatic Implications

The findings from this study support current guidelines that call for using a non-PI-based regimen among persons co-treated for TB. Countries that do not follow these guidelines should develop methods to ensure that patients are treated according to standard recommendations.


  1. Meyers T, Moultrie H, Naidoo K, Cotton M, Eley B, Sherman G. Challenges to pediatric HIV care and treatment in South Africa. J Infect Dis 2007; 196 (Suppl 3):S474-S481.
  2. Guidelines for the management of HIV-infected children. National Department of Health, South Africa; 2008.
  3. National Department of Health SA. National antiretroviral treatment guidelines. Health NDo, ed. Johannesburg: Minuteman Press; 2004