Survival rates after the initiation of antiretroviral therapy stratified by CD4 cell counts in two cohorts in Canada and the United States

In order to characterize patterns of survival among individuals initiating double and triple combination therapy in Canada and the United States, we conducted a population-based analysis of antiretroviral therapy in HIV-infected individuals enrolled in the British Columbia Drug Treatment Program (BC). We then compared the Canadian data with the recent report of the Centers for Disease Control and Prevention Adult and Adolescent Spectrum of Disease Project in the USA (CDC) [¹], a medical record review surveillance project conducted in 11 US cities. Eligible patients in the two observational studies were treatment naive and initiated antiretroviral therapy with double or triple drug combinations in 1994 or later (n = 1219 in BC; n = 5110 in CDC). Rates of progression from the initiation of triple drug antiretroviral therapy to death were stratified by CD4 cell counts at the time when antiretroviral therapy was initiated. Using the Cox proportional hazards method, we computed the hazard ratios of death, using those who started treatment at a CD4 cell count of 500 cells/mm³ or greater as reference (Table 1).

Table 1

Both cohorts showed a statistically significantly greater risk of death for those who started treatment with a CD4 cell count below 200 cells/mm³. There was no statistically significant difference in the risk of death when treatment was started at a CD4 cell count of 200-500 and 500 cells /mm³ or greater in both cohorts.

We then compared the Kaplan-Meier estimates of survival between the BC and CDC studies. The overall 2 year survival rate was 93.4% [95% confidence interval (CI) 92.3-94.5%] and 80.9%, respectively (data were not available to calculate the confidence intervals of the CDC estimates). A lower survival rate was apparent in the CDC cohort at low baseline CD4 cell counts. In the CD4 cell count strata of 0-49 and 50-99 cells/mm³, the lower 95% CI limits of the BC survival rates were 73.8 and 84.4%, which were higher than the CDC point estimates of 64.8 and 78.1%, respectively (see Fig. 1).

Fig. 1

These results are consistent with recently published data regarding uniformly low rates of disease progression and death among patients initiating therapy with CD4 cell counts of 200 cells/mm³ or greater [²]. However, the reported rates of death for patients initiating therapy with CD4 cell counts below 200 cells/mm³ appear to be consistently lower at any CD4 cell count threshold in the BC than in the CDC cohort.

Clearly, reasons other than antiretroviral treatment are needed to explain the differences in survival rates at low baseline CD4 cell counts between the BC and CDC cohorts. Baseline differences in the population mix could have been contributory, as age and possibly ethnicity have been associated with survival, despite similar levels of CD4 cell count [³]. In the BC study, we identified a previous AIDS diagnosis, CD4 cell count at treatment initiation and baseline plasma HIV-RNA levels as significantly associated with death in univariate analysis. There was no difference in mortality by either age or sex. Of note is the fact that in multivariate analyses, only the CD4 cell count remained significantly associated with death [²]. Non-adherence is a strong determinant of treatment success, and a predictor of survival [⁴]. However, we do not believe that this is the likely explanation of our findings, as there is no evidence that adherence is systematically decreased in the USA versus Canada.

We feel that the differences in survival rates at low CD4 cell counts between the BC and CDC cohorts are worthy of further examination. Differences in important baseline characteristics not captured in this report, such as sex balance, dual versus triple regimen use, socioeconomic status, risk factors for HIV infection, and co-morbidities are some of the possible variables that may account for these differences. Also, a possible contribution of heahealthcare delivery would need to be explored. Whereas all patients in the BC cohort have free access to all licensed antiretroviral therapy, plus laboratory and medical monitoring as part of the Canadian medical system, this is not necessarily the case in the United States.

Kenny C. W. Chana

Benita Yipb

Robert S. Hoggb

Julio S. G. Montanerb

Michael V. O'Shaughnessyb