Other immunological markers

In resource-limited regions, less expensive and more readily accessible prognostic markers for HIV disease besides CD4 counts and viral load are needed. Researchers from the Multicenter AIDS Cohort Study examined how total lymphocyte counts (TLC) and plasma haemoglobin levels may be used for predicting AIDS.

Using data from their cohort between 1983 and 1996, they identified the first time TLC declined by more than 33% per year and the first time haemoglobin declined by more than 11.6% per year. These values were previously identified as maximising sensitivity and specificity for predicting AIDS.

They found that a TLC <1200 cells/mm³, which is the WHO recommendation for initiation of therapy when CD4 counts are not available, predicted progression to AIDS (risk hazard [RH] = 6.14; 95% confidence interval [CI]: 4.33, 8.71). A rapid decline in either TLC or haemoglobin was also significantly associated with progression to AIDS (RH = 4.70 [3.23, 6.86] and 5.55 [3.69, 8.36], respectively).

Furthermore, when present along with TLC <1200 cells/mm³, a rapid decline in either TLC or haemoglobin was strongly associated with progression to AIDS (RH = 11.30 [7.39, 17.31] and 21.19 [12.43, 36.12], respectively). These results support the usefulness of these markers for monitoring HIV-infected people in resource-limited regions.¹

CD8 cell counts have some prognostic significance. Early in HIV infection, the higher the total CD8 count, the higher the risk of eventually reaching a CD4 count below 50 cells/mm³. This may be because high CD8 cell counts are related to excessive immune activation. Later in HIV infection, a higher CD8 count may indicate a reduced risk of developing AIDS in the short term.

CD8 T-cells are a mixed population of suppressor and cytotoxic cells. Several studies found higher counts of a subset of CD8 cells, called CD8/CD38, to predict a worse risk of developing AIDS. Conversely, a higher count of the CD8/CD28 subset may predict a lower risk of progression.

Overall, the total CD8 cell count is less helpful than the total CD4 cell count as a prognostic marker. Subsets of CD8 T-cells also seem to be related to measures of viral load.

The sustained immune activation (inflammation) that results from chronic HIV infection is now recognised as playing a key role in HIV disease progression, including HIV-associated metabolic disorders. Researchers are investigating several markers of immune activation that may help identify HIV-positive people most at risk for rapid progression.

In a case-control analysis of patients from the SMART study, two markers – one inflammatory and one linked to blood coagulation – very strongly predicted an increased risk of death. High levels of interleukin-6 (IL-6), a marker of inflammation, were associated with a nearly twelvefold increase in mortality risk (adjusted odds ratio [AOR], 11.8). High levels of blood coagulation marker D-dimer predicted mortality risk even more strongly (AOR, 26.5). This was the case whether levels were high at baseline or whether they increased after antiretroviral therapy was interrupted. Levels of IL-6 and D-dimer were also higher in SMART participants who interrupted their therapy than in those who remained on continuous therapy, and levels were higher in patients with higher viral loads.

Smaller increases in mortality risk were linked to higher levels of two other markers of inflammation: hsCRP (high sensitivity C-reactive protein) and amyloid P. Elevated hsCRP and IL-6 levels were also predictors of developing an opportunistic disease.^{2 3}

In another study using SMART data,^4,5 levels of C-reactive protein (CRP), IL-6, D-dimer and cystatin C were measured in 781 HIV-positive SMART participants and in participants in two other large US studies of cardiovascular disease, the US Multi-Ethnic Study of Atherosclerosis (MESA) and Coronary Artery Risk Development in Young Adults (CARDIA). SMART study patients with HIV had 42% higher CRP levels and 59% higher IL-6 levels than CARDIA participants, and significantly higher levels of CRP, IL-6 and D-dimer than MESA participants. CRP levels were higher in those on non-nucleoside reverse transcriptase inhibitors (NNRTIs) than on protease inhibitors (PIs), and in those on abacavir compared to other nucleoside reverse transcriptase inhibitors (NRTIs). While SMART participants were more likely to smoke and have higher cholesterol than the other patients, the differences in immune markers remained when these factors were controlled for.

Investigators from the FRAM study found that CRP and fibrinogen levels both independently predicted mortality risk. Risk of death was four times higher for FRAM participants with baseline CRP levels greater than 3 mg/dl than for those with levels less than 1 mg/dl. In participants with the highest fibrinogen levels (greater than 406 mg/dl), mortality was three times higher than in those with the lowest levels (less than 319 mg/ml). Higher fibrinogen levels increased the risk of death within each category of CD4 cell counts.⁶

Beta-2 microglobulin and neopterin are naturally occurring proteins released from activated lymphocytes. Levels of beta-2 microglobulin above 5 micrograms per litre and of neopterin above 15 nanograms per millilitre suggest a higher relative risk of disease progression. Beta-2 microglobulin levels after seroconversion are predictive of the risk of disease progression within ten years.

Several proteins produced by CD4 and CD8 cells have been shown to correlate with disease progression. High levels of CD127, an indicator of CD8 memory cell survival, correlate with better prognoses. Conversely, high levels of Ki-67, a marker of immune activation in CD4 and CD8 cells, are associated with higher risk of AIDS and mortality.⁷

Abnormal blood cytokine levels in HIV infection are probably related to the stage of HIV disease and CD4 cell counts and may be predictive of disease progression. Possible useful markers are levels of IL-2 (interleukin-2) and interferon gamma, which both decrease in HIV infection. Assays measuring cytokines are expensive, difficult to use, and may not add to the value of knowing the CD4 count.