A detectable CMV viral load could signal a high risk of mortality in older individuals hospitalised with HIV-associated TB

Theo Smart
Published: 09 August 2016

A study conducted in a cohort of hospitalised adults with HIV-associated tuberculosis (TB) co-infection from Khayelitsha found that having a detectable cytomegalovirus (CMV) viral load was associated with higher mortality within the first 12 weeks on TB treatment, according to Dr Amy Ward of the University of Cape Town who presented the findings at the 21st International AIDS Conference (AIDS 2016) last month in Durban, South Africa

However, there was no clear evidence that CMV was the direct cause of increased mortality; and after adjustment for other risk factors, the strength of association with mortality was no longer statistically significant for the overall cohort. Nevertheless, among older adults (those who were 36 years old or older), both the likelihood of early mortality and CMV viraemia were increased,“perhaps reflecting premature aging of the immune system,” Ward suggested.


People who are hospitalised with advanced HIV-TB have an extremely high rate of mortality ranging between 15 and 50%, despite antiretroviral therapy (ART) and TB treatment. The reasons for this are unclear; but one idea is that active CMV infection may play a role. Autopsy studies indicate that CMV organ disease is often present in individuals dying of HIV-associated TB leading to the hypothesis that CMV could negatively impact the immune response to either TB or the virus (for instance, perhaps by heightening inflammation).

Of note, there could be a precedent for this in other severe immune suppressed conditions, such as in people who have had organ transplants, where CMV viraemia is directly related to increased mortality. In that context, CMV is treated in order to reduce the risk of progression to illness and death – suggesting that if CMV plays a similar role in individuals with HIV-associated TB, it may be worth conducting a trial to see whether CMV treatment might reduce the mortality.

Ward and colleagues decided to first conduct a prospective study to better characterise the association of CMV with mortality in people hospitalised with HIV-associated TB.

The study

Adult HIV-positive inpatients with a CD4 of 350 or less and a new diagnosis of microbiologically proven or clinical TB were recruited from January 2014 to June 2015 at Khayelitsha Hospital, a large district hospital in Khayelitsha township near Cape Town. Blood tests, including a CMV viral load test were performed at baseline. In addition, each subject had a clinical assessment (a full history, physical examination and tests to confirm TB) and a fundoscopy to look for CMV retinitis. Follow-up was performed daily while in the hospital, and then by phone or in person at weeks 4 and 12. The primary endpoint was 12-week mortality after the initiation of TB treatment.

The study included 256 participants with a diagnosis of TB and CMV viral load data. Five individuals were lost to follow-up. Over half the participants were women and the median age was 36. At the time of enrolment, only 35% were on ART, 44% had previously been on ART but were not at the time of hospitalisation, and 21% were previously untreated. The median CD4 cell count was 64 cells/mm3, and the median HIV viral load was 5.2 log (~158,000 copies per ml). 40.6% had a positive blood culture for TB and 7.4% were rifampicin-resistant.

Almost a third (30.9%) had a detectable CMV viral load, with a median value of 1250 copies/ml. Fifty-two per cent of those with detectable CMV had CMV loads of 1000 copies per ml or more. None of the participants had CMV retinitis.

Mortality and analysis

At 12 weeks, 23% of the participants had died. Thirty-eight per cent of those with a detectable CMV viral load had died by week 12 compared to 17.8% of those without detectable CMV.

In a univariate Cox regression analysis, having a detectable CMV viral load was associated with twice the risk of 12-week mortality. Other factors associated with increased mortality were increasing age, having a lower CD4 count, and lower albumin levels. When these variables, plus HIV viral load, TB blood culture and sex were included in a multivariate analysis, having detectable CMV was still associated with a 70% increase in mortality, though the p-value (0.077) did not quite reach significance. Only the association between increasing age and mortality remained a statistically significant in the multivariate model.

There was no association between mortality and having a higher CMV viral load.

In an analysis stratified by the median age, participants aged 36 years and older had a significantly higher mortality (32.8% compared to 14.1%). This did not appear to be significantly associated with having a CD4 cell count below 50, however, it was significantly associated with having detectable CMV viral loads but only in those aged 36 or older.

“This could indicate that independent of CD4 count there are other factors in this older group that are causing immune suppression resulting in an increase in both CMV viraemia and mortality,” Ward postulated – suggesting that CMV viraemia could be a marker of premature ageing of the immune system.

Even so, she concluded that, “larger studies are required to address this question as we did note a trend towards increased mortality in those with CMV viraemia even after controlling for relevant co-variates.”


Ward A et al. CMV viraemia and 12-week mortality among hospitalised adults with HIV-associated TB in Khayelitsha Hospital, South Africa: A prospective cohort study. 21st International AIDS Conference, Durban, abstract WEAB0203, 2016.

View the abstract on the conference website.

Download the presentation slides from the conference website.

Watch the webcast of this presentation on YouTube.

NAM’s news coverage of the International AIDS Conference has been made possible thanks to support from Merck & Co. NAM's wider conference news reporting services have been supported by Gilead, Janssen and ViiV Healthcare.

Community Consensus Statement on Access to HIV Treatment and its Use for Prevention

Together, we can make it happen

We can end HIV soon if people have equal access to HIV drugs as treatment and as PrEP, and have free choice over whether to take them.

Launched today, the Community Consensus Statement is a basic set of principles aimed at making sure that happens.

The Community Consensus Statement is a joint initiative of AVAC, EATG, MSMGF, GNP+, HIV i-Base, the International HIV/AIDS Alliance, ITPC and NAM/aidsmap

This content was checked for accuracy at the time it was written. It may have been superseded by more recent developments. NAM recommends checking whether this is the most current information when making decisions that may affect your health.

NAM’s information is intended to support, rather than replace, consultation with a healthcare professional. Talk to your doctor or another member of your healthcare team for advice tailored to your situation.