CD4 Count Before Starting HIV Treatment Predicts Mortality for First 5 Years of Treatment

Pretreatment CD4 count strongly predicted mortality in the first years of antiretroviral therapy (ART), according to analysis of 18 cohorts in Europe and North America. But among people who survived five or more years on ART, pretreatment CD4 count no longer affected mortality.

The Antiretroviral Therapy Cohort Collaboration (ART-CC) has collected data in Europe and North America since the year 2000. Analysis of three- and five-year survival on ART published in 2002 and 2007 identified pre-ART CD4 count as the strongest predictor of short-term morality. To extend that analysis to patients treated for up to 15 years, the ART-CC team conducted a new study.

Related: Starting HIV Treatment Soon After Diagnosis Trims Time to Viral Suppression

The analysis involved ART-CC members who started their first combination ART regimen between 1996 and 2001 and had a CD4 count in the three months before starting ART. Follow-up continued through July 2013, so all participants had at least 12 years of potential follow-up. All ART-CC members are at least 16 years old. The investigators divided pre-ART CD4 counts into six groups (0-49, 50-99, 100-199, 200-349, 350-499 and ≥500 cells/mm3), pre-ART viral loads into three groups (0-9999, 10,000-99,999 and ≥100,000 copies/μL), baseline age into five groups (16-29, 30-39, 40-49, 50-59 and ≥60 years) and year starting ART into three groups (1996-1997, 1998-1999 and 2000-2001).

The ART-CC team used Poisson models to estimate crude mortality rate (MR) by pre-ART CD4 count. They used Cox models stratified by cohort to estimate crude mortality rate ratio (MRR) and adjusted MRR by pre-ART CD4 group (with 200-349 cells/mm3 as the reference) overall and separately according to time on ART. Adjusted MRR considered age, sex, transmission risk group, AIDS at baseline, baseline viral load and year starting ART.

The study involved 37,496 people starting ART: 8304 in 1996-1997, 15,599 in 1998-1999 and 13,593 in 2000-2001. Most cohort members (78%) were men. Median pre-ART CD4 count measured 221 cells/mm3, median viral load 75,000 copies/mL and median age 37 years. Median follow-up stood at 11.3 years, with 77% of participants having at least five years of follow-up and 58% having more than 10 years. During 359,219 years of follow-up, 6344 people died.

Pre-ART CD4 count proved highly prognostic for unadjusted cumulative mortality in an analysis of all patients: The lower the CD4 count, the higher the mortality. An adjusted estimate of cumulative mortality in a typical patient group (men who have sex with men in their 30s starting ART in 1998-1999 in France) showed mortality clearly increasing as pre-ART CD4 count decreased. Cumulative 15-year mortality measured 7.1% in men starting ART with a CD4 count above 500 cells/mm3 versus 10.7% for those starting with a CD4 count below 50 cells/mm3.

Overall MR fell from 33 per 1000 person-years (95% confidence interval [CI] 30 to 36) during the first six months of ART to 14 per 1000 person-years (95% CI 13 to 15) in patients who survived 10 years on ART, even though the population aged during this period. MR was strongly inversely associated with pre-ART CD4 count at shorter times since starting ART, but the magnitude of this association waned as time since starting ART increased. After 10 years of ART, the analysis showed little evidence of differences in crude MR by baseline CD4 count.

In an analysis adjusted for other prognostic factors, the MRR for starting ART below 50 CD4 cells/mm3 versus 200-349 cells/mm3 was 2.81 (95% CI 2.12 to 3.71) during the first six months of ART. That MRR dropped to 1.59 (95% CI 1.31 to 1.92) after three to 4.9 years of ART. After five years of ART, MRRs changed little by pre-ART CD4 count: Between five and 9.9 years of ART, the adjusted MRR per CD4 category stood at 0.97 (95% CI 0.94 to 1.00, P = .054), and after 10 or more years of ART, the adjusted MRR per CD4 category stood at 1.02 (95% CI 0.98 to 1.07, P = .32). In other words, baseline CD4 count stopped predicting mortality after five years of ART.

The ART-CC team concludes that "[e]ven patients who started ART with very low (<50 cells/μL) CD4 counts may experience convergence of their mortality risk to that of patients with intermediate (200-349 cells/μL) or high (≥500 cells/μL) baseline CD4 count, from 5 years after the start of treatment." The researchers suggest "this is an important message for patients and physicians alike, as the moving from a higher- to a lower-risk group through long-term treatment adherence may be a powerful motivator."

Of course, this convergence depends on surviving the first five years of ART, and this study confirmed that the proportion of people who died dwindled as pre-ART CD4 count rose: 25% with 0-49 cells/mm3 died, versus 22% with 50-99 cells/mm3, 19% with 100-199 cells/mm3, 14% with 200-349 cells/mm3, 12% with 350-499 cells/mm3 and 11% with ≥500 cells/mm3. The ART-CC investigators note that "the observed convergence of survival rates after 5 years may be partly explained by the early death of patients with low to very low baseline CD4 count[.]" But the authors point out that 85.7% of cohort members starting ART with a CD4 count below 200 cells/mm3 did survive five years.