CROI: Is there a ceiling to CD4 recovery on virologically effective therapy?

This article is more than 17 years old.

People who respond well to antiretroviral treatment in virologic terms, with HIV viral load dropping to undetectable levels and remaining suppressed, can have differing immunologic responses. Several research teams at the recent Fourteenth Conference on Retroviruses and Opportunistic Infections in Los Angeles presented data on trends in CD4 cell recovery after treatment initiation, and characteristics which may lead to differences in CD4 cell recovery rates.

The extent of CD4 cell recovery on antiretroviral therapy has recently sparked renewed interest following the publication of findings from the Johns Hopkins University cohort in the United States, which suggested that CD4 cell counts are unlikely to reach normal levels in treated individuals unless treatment begins before the CD4 cell count falls below 350 cells/mm3, and that no further gains could be expected after four years of treatment.

Data presented at CROI did not altogether confirm these findings; the EuroSIDA collaboration could not find evidence for a plateau effect, and the Swiss HIV Cohort could not confirm that people with CD4 counts above 200 cells/mm3 experienced greater CD4 cell gains, suggesting the need for much larger, pooled analyses of cohort studies to answer this critical question.


Data from the large, observational EuroSIDA study were presented in a poster. The 1,517 participants included in this analysis were previously antiretroviral-naïve individuals who began combination antiretroviral therapy between July 1997 and January 2003, and had at least two viral load measurements under 50 copies/ml. The researchers calculated annual changes in CD4 cell count in people with sustained viral load suppression; “baseline” was the time of the first viral load measurement below 50 copies/ml.



In HIV, an individual who is ‘treatment naive’ has never taken anti-HIV treatment before.

statistical significance

Statistical tests are used to judge whether the results of a study could be due to chance and would not be confirmed if the study was repeated. If result is probably not due to chance, the results are ‘statistically significant’. 


The result of a statistical test which tells us whether the results of a study are likely to be due to chance and would not be confirmed if the study was repeated. All p-values are between 0 and 1; the most reliable studies have p-values very close to 0. A p-value of 0.001 means that there is a 1 in 1000 probability that the results are due to chance and do not reflect a real difference. A p-value of 0.05 means there is a 1 in 20 probability that the results are due to chance. When a p-value is 0.05 or below, the result is considered to be ‘statistically significant’. Confidence intervals give similar information to p-values but are easier to interpret. 


A precursor to a building block of DNA or RNA. Nucleosides must be chemically changed into nucleotides before they can be used to make DNA or RNA. 


A person who has never taken treatment for a condition.

The study group average age was 39 years; average CD4 cell counts were 210 cells/mm3 before starting therapy and 410 cells/mm3 at baseline. A total of 341 (23%) had had a previous AIDS diagnosis; 1,202 (79%) were men. HIV risk categories were varied. Results were based on a total of 12,137 consecutive pairs of (undetectable) viral load measurements; 5,614 of these (46%) were in people taking a protease inhibitor (PI) -based regimen, 4,923 (41%) in people on an NNRTI-based regimen, and 1600 (13%) in those on triple-nucleoside regimens based upon abacavir.

Annual increases in CD4 cell count were seen in all participants except those who had been on antiretroviral therapy for more than five years, and whose CD4 cell counts had already reached levels over 500 cells/mm3. (The actual median CD4 cell count in the “over 500” group was 701 cells/mm3.) Otherwise, CD4 cell counts increased at varying rates: the most rapid increases (approximately 100 cells/mm3/year) happened in the first year after starting treatment, regardless of initial CD4 cell count; after the first year, increases occurred more slowly. After five years on therapy, modest but statistically significant increases (approximately 50 cells/mm3/year) were still occurring in participants with CD4 cell counts less than 500/mm3.

The researchers concluded that there was “little evidence of a plateau effect of CD4 rise except at near-normal CD4 levels … The majority of patients continued to experience significant rises in CD4 cell count, even after five years ... These results suggest that near-normal CD4 counts could be reached in all patients with maximum virologic suppression for a sufficient length of time.”

The Swiss HIV Cohort Study

Researchers at the Basel Institute for Clinical Epidemiology also reported on treatment-naïve individuals beginning combination antiretroviral therapy for the first time. Out of 2,860 who began treatment, 1,816 (63%) achieved viral load suppression (below 50 copies/ml) on at least two measurements within twelve months of starting therapy. CD4 cell counts in this group were followed for up to five years, or until viral load rebounded to above 400 copies/ml. In general, median CD4 cell counts increases were quicker at first and then slowed down, with increases of 87, 52, and 19 cells/mm3/year in the first year, years two to three, and years four to five, respectively. (At year five, 573 participants remained in the cohort.)

Greater average CD4 cell count increases were seen in:

  • Women, with increases between 15 and 28 cells/mm3/year more than men (p
  • Younger people – especially in the first year, where average CD4 cell count increases were 16 cells/mm3/year lower with each decade of age (p
  • People who started treatment with high viral loads – in the first year, average increases dropped off by 15 cells/mm3/year for each log10 decrease in initial viral load (p = 0.002).

Lower CD4 cell count increases were seen in:

  • People with CD4 cell counts of 650 cells/mm3 or more at the start of treatment (average 15 to 37 cells/mm3/year less, p = 0.014) (The EuroSIDA cohort, which indicated no difference in response by initial CD4 count, was composed of people with relatively low CD4 counts and would have contained few at this high level.)
  • Those taking NNRTI-based treatment (average six to 22 cells/mm3/year less compared to boosted protease inhibitors, p = 0.043).
  • Those on any combination containing tenofovir (Viread) (average 39 to 44 cells/mm3/year less compared to stavudine (Zerit), p = 0.015). This lowered response was seen for tenofovir overall, not just for the combination of tenofovir and didanosine (Videx), which has been shown to blunt CD4 responses in other studies.

Patients without hepatitis B (p=0.031) or with a CD4 cell counts initially above 200 cells/mm3 (p=0.017) also had higher CD4 increases but this was “not confirmed in all sensitivity analyses.” There were also trends toward lower increases with triple nucleoside therapy or zidovudine/lamivudine (AZT/3TC), but these were not statistically significant (p = 0.077 and 0.0504, respectively).

The Basel team concluded that the “clinical relevance of these findings must be confirmed in large collaborative cohort projects but could influence guidelines in older patients and those starting cART [combination antiretroviral therapy] at low CD4 cell levels.”

Western India

Both the preceding studies were done on populations in resource-rich countries; another presentation described immunologic outcomes of treatment in western India. This retrospective chart review looked at 515 previously treatment-naïve people on virologically successful treatment (minimum of one viral load result below 400 copies/mL). The median age was 36 years, 113 (80%) were male; the median CD4 cell count was 110 cells/mm3before treatment.

Treatment was an NNRTI-based triple combination, consisting of lamivudine (3TC, Epivir), d4T (stavudine, Zerit) or zidovudine (AZT, zidovudine, Retrovir), and nevirapine (Viramune) or efavirenz (Sustiva). Only patients with above 90% adherence (according to four-day self-recall) were included.

A total of 392 participants were followed for at least six months; follow-up is available to 54 months in eleven. An average overall increase of 232 cells/mm3 was seen after the first year and 335 cells/mm3 at four years.

All participants with pre-therapy CD4 counts below 200/mm3 attained CD4 cell counts above 200 cells/mm3 over the course of study follow-up; 50% attained CD4 cell counts above 350/mm3, and 16% achieved CD4 cell counts above 500/mm3. As in the Swiss cohort, poorer CD4 recovery was seen in men; the west India study also saw poorer recovery in those with baseline CD4 cell counts below 50 cells/mm3.

The investigators stated that “[patterns] of immunologic recovery (very rapid initial rise followed by slower increase) and risk factors associated with poor CD4 recovery among virologically suppressed HIV-infected patients on first-line HAART in western India is similar to those documented in the developed world. However the proportion of patients with normal CD4 counts (≥500), even at four years, is minimal. Immunologic recovery is better if HAART is initiated before CD4 drops to 3.”


Wolbers M et al. Predictors for CD4 cell count increase for patients with sustained viral load suppression within 1 year after starting cART: The Swiss HIV Cohort Study. Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, abstract 518, 2007.

Mocroft A et al. Normalization of CD4 counts in patients taking combination ART with maximum virologic suppression. Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, abstract 519, 2007.

Pujari S et al. Immunologic recovery among virologically suppressed HIV-infected patients in western India. Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, abstract 521, 2007.