CD4 cell counts may not be accurate marker of treatment failure for resource-poor HIV care

This article is more than 15 years old. Click here for more recent articles on this topic

Clinicians treating people with HIV in resource-limited settings should be cautious when making treatment switches based solely on CD4 cell counts, say a group of researchers including International AIDS Society president, Julio Montaner. Writing in the December 15th issue of the Journal of Acquired Immune Deficiency Syndromes, the team reports that the practice, recommended in large parts of the world where viral load testing is not available, may not accurately identify treatment failure, leading clinicians to delay switching from a failing regimen or causing them to switch away from a still effective regimen.

Antiretroviral therapy suppresses viral replication in people living with HIV, slowing disease progression and maintaining health. In resource-rich settings, the effectiveness of antiretroviral therapy is monitored by viral load tests, which measure the level of HIV in the blood. WHO guidelines suggest that when viral load tests are not available, CD4 cell counts should be used to evaluate whether or not antiretroviral therapy is effectively suppressing viral replication. This is common in many resource-limited parts of the world where HIV is prevalent.

Antiretroviral regimens are changed if monitoring reveals that the virus is not suppressed. With viral load tests, a detectable viral load is taken as a sign of treatment failure. In the absence of viral load tests, the WHO guidelines recommend that a person change their drug regimen if either their CD4 cell count returns to levels seen before starting antiretrovirals or their CD4 cell count falls to less than 50% of its peak value.


treatment failure

Inability of a medical therapy to achieve the desired results. 

detectable viral load

When viral load is detectable, this indicates that HIV is replicating in the body. If the person is taking HIV treatment but their viral load is detectable, the treatment is not working properly. There may still be a risk of HIV transmission to sexual partners.

positive predictive value

When using a diagnostic test, the percentage of those testing positive who are correctly diagnosed. This will vary according the prevalence in the local population.

disease progression

The worsening of a disease.


The process of viral multiplication or reproduction. Viruses cannot replicate without the machinery and metabolism of cells (human cells, in the case of HIV), which is why viruses infect cells.

There is limited clinical evidence in resource-limited settings that CD4 cell counts are accurate indicators of viral suppression on treatment. To address this question, a team of US and Canadian researchers undertook an analysis of the relationship between immunologic and virologic markers among a group of people initiating treatment in the African country of Uganda.

The Home-Based AIDS Care Project, based in eastern Uganda, is studying the impact of different monitoring strategies on disease progression. In this analysis, participants with a CD4 cell count below 250 cells/mm3 or signs of advanced disease were offered the fixed regimen of 3TC (lamivudine), d4T (stavudine) and nevirapine. Blood samples were collected every three months and tests performed to determine viral load and CD4 cell counts.

Using time points up to 24 months, investigators grouped data according to whether viral load was detectable or not (using a range of cut offs: 50, 500, 1000, 5000 copies/ml) and then compared three factors between the two groups: absolute CD4 cell count; median change in CD4 cell count from baseline; and percentage showing no change or a decline in CD4 cell counts.

The proportion of participants failing at each time with each definition varied between 1.5% and 16.4%. For each time point, the absolute CD4 cell count and change in CD4 cell count was lower in patients with a detectable viral load, and differences were larger when a more stringent definition of undetectable was used. However, not all differences were statistically significant, indicating that some of the time CD4 cell counts do not distinguish between successful and unsuccessful treatment.

To better quantify the ability of CD4 cell counts to detect treatment failure accurately, investigators calculated the sensitivity and positive predictive value of several possible CD4 cell count measures. Sensitivity is the proportion of people with a condition who are identified by the test, while the positive predictive value is the proportion of people with a positive test who actually have the condition.

For a definition of treatment failure similar to that in the WHO guidelines, that is: “no increase in CD4 cell count or an at least 50% drop in CD4 cell count at twelve months”, the calculated sensitivity was 0.08, meaning only eight of 100 people failing treatment would actually be identified. The investigators point out that not identifying all people failing treatment could increase the risk of disease progression and drug resistance within a population.

The positive predictive value for this definition was 0.11, meaning of 100 people identified as failing treatment according to CD4 cell counts, only eleven would actually be failing and 89 would have suppressed virus. “If the current WHO guidelines were applied to these patients,” the investigators write, “they would have been mistakenly identified as failing ART and prematurely switched from their primary antiretroviral regimen which was effectively controlling viral replication.”

The investigators acknowledge that the value of their analyses is limited by the small number of participants with treatment failure in their study (between 2% and 16%) and may underestimate the predictive power of CD4 cell counts. In defence, they point to other data that support their finding, and say that even in a scenario where 30% of people were failing treatment, the positive predictive value would be only 0.64.

While CD4 cell counts may not be an accurate marker of treatment failure, the investigators remark that both CD4 counts and viral load are independently associated with disease progression. CD4 cell counts, they propose, have other benefits in monitoring patient health, including the need for prophylaxis against opportunistic infections. Also, “CD4 counts could be used as a screening test to identify those persons who require VL testing, potentially reducing the demand for viral load testing in situations where it is available,” they propose.

In a strongly worded conclusion the investigators say, “immunologic parameters do not seem to accurately identify individuals receiving antiretrovirals with unsuppressed viral loads or virologic failure. Guidelines for monitoring individuals on antiretroviral therapy in resource-limited settings should be adapted to recognize this limitation.” They caution, “In the interim, clinicians should be cautious in initiating therapy changes based solely on CD4 cell count criteria.”


Moore DM et al. CD4+ T-cell count monitoring does not accurately identify HIV-infected adults with virologic failure receiving antiretroviral therapy. J Acquir Immune Defic Syndr 49: 477 – 484, 2008.