Long-term Mortality in HIV-Positive Individuals Virally Suppressed for >3 Years With Incomplete CD4 Recovery
Download the PDF here
Response to antiretroviral therapy: improved survival associated with CD4 above 500 cells/μl
AIDS: 17 July 2012
"Patients with CD4 cell counts greater than 500 cells/μl had better survival than those with levels between 350 and 500 cells/μl [adjusted hazard ratio (aHR) 1.70; 95% CI 1.26-2.30] and with CD4 cell counts between 200 and 349 cells/μl (aHR 2.56; 95% CI 1.93-3.38)."
Mortality in Patients with Well Controlled HIV and High CD4 Counts in the Continuous ART Arms of the SMART and ESPRIT Trials Compared to the General Population......they find evidence that 350-500 has greater mortality risk than >500
CD4 Count Under 700 Linked to All-Cause Mortality in Veterans Cohort
Fewer CD4s at HIV Diagnosis Tied to Higher Risk of non-AIDS Deaths
Clin Infect Dis. January 22, 2014
"The cumulative probability of survival stratified by CD4 count is presented in Figure 2. The estimated 5-year cumulative mortality (with 95% CIs) was 11.8% (8.9%-15.2%) in patients with a CD4 count <200 cells/μL at the end of the suppressed period, compared with 4.1% (3.1%-5.3%), 2.2% (1.4%-3.4%), and 2.2% (1.2%-3.7%) in patients with a CD4 count of 201-350, 351-500, and >500 cells/μL, respectively, at the end of the suppressed period. Compared with individuals with CD4 count >500 cells/μL at the end of the suppressed period, adjusted HRs in individuals with a CD4 count of 351-500, 201-350, and ≤200 CD4 cells/μL were 0.62 (95% CI, .32-1.19), 1.28 (95% CI, .74-2.23), and 2.62 (95% CI, 1.47-4.67), respectively."
Long-term Mortality in HIV-Positive Individuals Virally Suppressed for >3 Years With Incomplete CD4 Recovery
Clin Infect Dis. January 22, 2014
Frederik N. Engsig,1,2 Robert Zangerle,3 Olga Katsarou,4 Francois Dabis,5 Peter Reiss,6,7 John Gill,8 Kholoud Porter,9 Caroline Sabin,10 Andrew Riordan,11 Gerd Fatkenheuer,12 Felix Gutierrez,13 Francois Raffi,14 Ole Kirk,1,15 Murielle Mary-Krause,16 Christoph Stephan,17 Patricia Garcia de Olalla,18 Jodie Guest,19 Hasina Samji,20 Antonella Castagna,21 Antonella d'Arminio Monforte,22 Adriane Skaletz-Rorowski,23 Jose Ramos,24 Giuseppe Lapadula,25 Cristina Mussini,26 Lluis Force,27 Laurence Meyer,28 Fiona Lampe,29 Faroudy Boufassa,30 Heiner C. Bucher,31 Stephane De Wit,32 Greer A. Burkholder,33 Ramon Teira,34 Amy C. Justice,35,36 Tim R. Sterling,37 Heidi M. Crane,38 Jan Gerstoft,1 Jesper Grarup,15 Margaret May,39 Genevieve Chene,5 Suzanne M. Ingle,39 Jonathan Sterne,39 and Niels Obel1; for the Antiretroviral Therapy Cohort Collaboration (ART-CC) and the Collaboration of Observational HIV Epidemiological Research Europe (COHERE) in EuroCoord
Background. Some human immunodeficiency virus (HIV)-infected individuals initiating combination antiretroviral therapy (cART) with low CD4 counts achieve viral suppression but not CD4 cell recovery. We aimed to identify (1) risk factors for failure to achieve CD4 count >200 cells/μL after 3 years of sustained viral suppression and (2) the association of the achieved CD4 count with subsequent mortality.
Methods. We included treated HIV-infected adults from 2 large international HIV cohorts, who had viral suppression (≤500 HIV type 1 RNA copies/mL) for >3 years with CD4 count ≤200 cells/μL at start of the suppressed period. Logistic regression was used to identify risk factors for incomplete CD4 recovery (≤200 cells/μL) and Cox regression to identify associations with mortality.
Results. Of 5550 eligible individuals, 835 (15%) did not reach a CD4 count >200 cells/μL after 3 years of suppression. Increasing age, lower initial CD4 count, male heterosexual and injection drug use transmission, cART initiation after 1998, and longer time from initiation of cART to start of the virally suppressed period were risk factors for not achieving a CD4 count >200 cells/μL. Individuals with CD4 ≤200 cells/μL after 3 years of viral suppression had substantially increased mortality (adjusted hazard ratio, 2.60; 95% confidence interval, 1.86-3.61) compared with those who achieved CD4 count >200 cells/μL. The increased mortality was seen across different patient groups and for all causes of death.
Conclusions. Virally suppressed HIV-positive individuals on cART who do not achieve a CD4 count >200 cells/μL have substantially increased long-term mortality.
The introduction of combination antiretroviral therapy (cART) has decreased morbidity and mortality in individuals infected with human immunodeficiency virus (HIV) due to viral suppression and CD4 cell recovery [1-3]. However, some individuals treated with cART achieve viral suppression but not CD4 cell recovery [4-6]. Several studies have shown that individuals with successful virological response to cART and incomplete CD4 recovery have increased mortality [4, 5, 7-9]. However, the only previous study exclusively examining long-term mortality in individuals started late on cART with sustained viral load (VL) suppression and low CD4 count is limited by small sample size .
By combining data from 2 large international collaborations of HIV cohorts, the Antiretroviral Therapy Cohort Collaboration (ART-CC) and the Collaboration of Observational HIV Epidemiological Research Europe (COHERE), we examined risk factors for failure to achieve CD4 recovery among treated individuals who were virally suppressed for >3 years, and compared mortality rates after 3 years of viral suppression according to the CD4 cell count reached at the end of the virally suppressed period.
Setting and Participants
The ART-CC (Error! Hyperlink reference not valid. is an international collaboration of 18 cohort studies of HIV type 1 (HIV-1)-positive individuals from Europe and North America that was established in 2000 to examine the prognosis of HIV-1-positive, treatment-naive individuals initiating cART . COHERE (Error! Hyperlink reference not valid. was established in 2005 and is an international collaboration of 35 cohorts from 29 European countries. The COHERE data were pooled within the EuroCoord network (www.eurocoord.net). Each collaboration focuses on scientific questions requiring large sample sizes, which the contributing cohorts cannot answer individually .
Study Population and Design
We identified all HIV-1-positive individuals who (1) were >16 years old at start of the suppressed period; (2) were on cART continuously (defined in ART-CC as treatment with at least 3 drugs from 2 different classes and in COHERE as the concomitant use of at least 3 antiretroviral drugs) for at least 3 years; (3) after start of cART had a period with suppressed VL of at least 3 years (all VL ≤500 HIV-1 RNA copies/mL and never a time span of >7 months between VL measurements); and (4) had a CD4 cell count ≤200 cells/μL at the start of the virally suppressed period (Figure 1). Individuals enrolled in >1 cohort were identified and only 1 record per individual was included. A VL cutoff of ≤500 copies/mL was chosen to overcome the heterogeneity of the assay detection limits used during the study period.
Risk Factors for Failure to Achieve a CD4 Count >200 Cells/μL After 3 Years of Viral Suppression
We used logistic regression to identify risk factors for not achieving CD4 count >200 cells/μL after 3 years of sustained viral suppression. We assessed the effect of CD4 cell count at the start of the suppressed period as a categorical variable (0-25, 26-50, 51-100, 101-150, and 151-200 cells/μL). In a sensitivity analysis, we fit separate models for each of these CD4 strata. The following variables were included in all models: age at start of virally suppressed period (<30, 30-39, 40-49, ≥50 years), probable route of infection (men who have sex with men [MSM], male heterosexual sex, female heterosexual sex, injection drug use [IDU], other/unknown [the number of male and female injection drug user and other/unknowns did not allow for further classification according to sex]), pre-cART VL (last VL available within 3 months before start of cART or first VL within a month after start of cART if the former was not available; VL <100 000 copies/mL, ≥100 000 copies/mL, and missing), year of cART initiation (1996-1997, 1998-2000, 2001 onward), time from cART initiation to start of suppressed period (<12 months, ≥12 months), AIDS (no AIDS event, ≥1 AIDS event before start of the virally suppressed period). We tested for interactions between pairs of variables.
Risk Factors for Mortality After 3 Years of Viral Suppression
Person-years at risk were calculated from the date of the first CD4 count within 3 months after 3 years of viral suppression to the earlier of time of death, loss to follow-up, or end of observation. We estimated mortality rates and Kaplan-Meier plots according to CD4 count at the end of the 3-year virally suppressed period (≤200, 201-350, 351-500, or >500 cells/μL) and used Cox regression to estimate hazard ratios (HRs) for death according to these CD4 count groups. All analyses were adjusted for the covariates listed above (AIDS events in this analysis were up to the end of the virally suppressed period) and stratified by cohort. We used Cox regression to compare individuals with CD4 counts ≤200 cells/μL and >200 cells/μL at the end of the suppressed period within strata defined by age at start of virally suppressed period, route of infection, AIDS status, and CD4 at start of suppressed period. In sensitivity analyses, we used a cutoff value of 50 copies/mL to define viral suppression and included only study subjects reporting sexual route of transmission and no positive test for hepatitis C virus (HCV) coinfection (HCV was defined as a positive test for HCV RNA or a positive test for HCV immunoglobulin G antibody at any time during follow-up).
Analysis of Causes of Death
COHERE does not collect data on causes of death, so these analyses were restricted to individuals included in ART-CC. Supplementary Data 1 describes how the causes of death data in this study were assigned and categorized. Causes of death were further categorized as AIDS defining, non-AIDS defining, unnatural (accident/violent/suicide/drug abuse), and unknown . Non-AIDS causes of death were further divided into hepatitis, cancers, and other. Because analyses of different causes of death can pose the problem of competing risks, we estimated both subdistribution HRs (using the Fine and Gray approach) and standard (Cox) HRs (adjusted only for age and sex due to the small number of events) . Because these estimates did not differ appreciably, we report only the estimates from Cox models. SPSS statistical software, version 15.0 (Norusis; SPSS Inc, Chicago, Illinois) and R software, version 2.8.1, were used for data analysis.
We identified 113 845 unique HIV-1-positive individuals from the COHERE and ART-CC cohorts, of whom 41 081 were treated with cART for <3 years, 50 495 did not have sustained viral suppression, and 368 had no available CD4 measurement at the start of the suppressed period. Of 21 901 individuals who achieved sustained viral suppression for ≥3 years, 16 193 had a CD4 count >200 cells/μL at start of the virally suppressed period and 158 had no available CD4 measurement at the end of the suppressed period, leaving 5550 individuals (20 291 person-years of observation; median follow-up time; 3.4 years [interquartile range, 1.6-5.3 years]) eligible for analyses (Figure 1). A histogram of the distribution of CD4 count at the start of the virally suppressed period is shown in Supplementary Data 2. The majority of these (4715 [85.0%]) achieved a CD4 count >200 cells/μL after 3 years of viral suppression. Half (2904 [52.3%]) were aged >40 years at the start of the suppressed period (Table 1).
Risk Factors for Failure to Achieve CD4 Count >200 Cells/μL
We found that risk of failure to achieve a CD4 count >200 cells/μL increased with increasing age and with decreasing CD4 count at start of the virally suppressed period (Table 2). Compared with MSM, men with heterosexual route of infection, injection drug users, and those with other or unknown transmission group had greater risk of incomplete CD4 recovery. Risk was also greater in those whose last viral load before start of cART was <100 000 copies/mL, those who initiated cART after 1998, and those who had 12 months or more from initiation of cART to start of the virally suppressed period. In models stratified on CD4 at start of the suppressed period, the impact of age and AIDS-defining illness on incomplete CD4 recovery appeared to be similar across CD4 strata (Supplementary Table 1). Patients infected via IDU had consistently higher risk of incomplete CD4 recovery than those infected via MSM transmission.
Time to Death From Any Cause
A total of 175 (3.2%) individuals died: 66 (7.9%) of those who did not attain a CD4 count >200 cells/μL and 109 (2.3%) of those who attained a CD4 count >200 cells/μL. Table 3 shows that individuals who did not attain a CD4 count >200 cells/μL after 3 years of sustained viral suppression had substantially increased mortality compared with those who achieved a CD4 count >200 cells/μL (adjusted HR, 2.60 [95% CI, 1.86-3.61]). The cumulative probability of survival stratified by CD4 count is presented in Figure 2. The estimated 5-year cumulative mortality (with 95% CIs) was 11.8% (8.9%-15.2%) in patients with a CD4 count <200 cells/μL at the end of the suppressed period, compared with 4.1% (3.1%-5.3%), 2.2% (1.4%-3.4%), and 2.2% (1.2%-3.7%) in patients with a CD4 count of 201-350, 351-500, and >500 cells/μL, respectively, at the end of the suppressed period. Compared with individuals with CD4 count >500 cells/μL at the end of the suppressed period, adjusted HRs in individuals with a CD4 count of 351-500, 201-350, and ≤200 CD4 cells/μL were 0.62 (95% CI, .32-1.19), 1.28 (95% CI, .74-2.23), and 2.62 (95% CI, 1.47-4.67), respectively.
Mortality Hazard Ratios Stratified by Risk Factors
Table 3 shows that the impact on mortality of not achieving a CD4 count >200 cells/μL was most pronounced in individuals whose CD4 count at the start of the suppressed period was 151-200 cells/μL (those in whom the increase in CD4 since start of ART was lowest). However, estimated HRs within CD4 strata were estimated imprecisely, and their CIs overlapped. The impact of incomplete CD4 recovery appeared similar across strata defined by age, mode of transmission, viral load before start of cART, year of cART initiation, time from initiation of cART to start of the virally suppressed period, and AIDS-defining disease (Table 3).
Of 2692 individuals who were eligible when a cutoff of 50 copies/mL for viral suppression was used, 2253 (83.7%) achieved a CD4 count >200 cells/μL. Estimated effects of risk factors for not achieving a CD4 count >200 cells/μL were broadly similar to those in the main analyses, although there was less evidence of elevated risk among heterosexual men (adjusted odds ratio, 1.17 [95% CI, .88-1.55]) and women (0.64 [95% CI, .45-.93]) (Supplementary Table 2). In this analysis, a total of 33 (7.5%) and 41 (1.8%) individuals with and without incomplete CD4 recovery, respectively, died during follow-up, and the adjusted HR for not attaining (compared with attaining) a CD4 count >200 cells/μL was 3.96 (95% CI, 2.36-6.66). Exclusion of 1625 (29.3%) individuals who did not report sexual route of transmission nor had HCV coinfection left 3925 individuals with 87 deaths. The adjusted HR for not attaining (compared with attaining) a CD4 count >200 cells/μL was 2.98 (95% CI, 1.85-4.79).
Time to Death From Specific Causes
Of 4135 individuals included in analyses of cause-specific mortality, 619 (15.0%) did not attain a CD4 count >200 cells/μL after the suppressed period and 121 (2.9%) died. Most deaths were from non-AIDS-defining causes, in both groups (Table 4). Mortality due to AIDS-defining, non-AIDS-defining, and unnatural causes of death was increased substantially in individuals who did not attain a CD4 count >200 cells/μL and was highest for hepatitis and non-AIDS-defining cancers (adjusted HRs, 6.76 [95% CI, 1.93-23.74] and 2.89 [95% CI, 1.44-5.28], respectively).
Based on data combined from 2 large international HIV cohort collaborations, we found that among HIV-positive individuals with 3 years of viral suppression on cART, those with incomplete CD4 recovery (CD4 count ≤200 cells/μL) had markedly higher mortality rates than those who achieved a CD4 count >200 cells/μL. These higher rates were observed consistently across strata defined by age, sex, route of transmission, and prior AIDS-defining illness. Rates of both AIDS- and non-AIDS-defining causes of death were elevated. We identified older age, transmission via male heterosexual sex or IDU, lower CD4 count at start of the suppressed period, and longer time from initiation of cART to start of the virally suppressed period as risk factors for incomplete CD4 cell recovery.
Strengths and Weaknesses
Because 15% of treated HIV-positive individuals have a CD4 count <200 cells/μL after long-term viral suppression, prognosis of such patients is a major concern. By combining data from 2 collaborations of HIV cohort studies, we assembled a sufficiently large data set to permit us to examine both risk factors and prognosis for all-cause and cause-specific mortality among patients with incomplete CD4 recovery. The contributing cohort studies represent a wide variety of countries and settings, and our results are therefore likely to be generalizable to treated HIV-positive individuals in Western Europe and North America. Not all contributing cohorts link their data with national death registries, which may lead to an underestimation of mortality rates. However, estimates of relative mortality comparing different groups should not be biased, providing that nonascertained deaths are missing at random . Serological tests for coinfection with HCV were not performed systematically in all cohorts; therefore, some coinfected individuals may have been misclassified. However, findings were similar in a sensitivity analysis excluding both injection drug users and HCV-coinfected individuals, and the impact of misclassification of HCV serostatus may have been limited because of its strong association with transmission via IDU, which was available from all cohorts. We did not have access to data on smoking, other comorbidity (eg, diagnosis of non-AIDS cancers) or non-cART medications (eg, cancer chemotherapy) and were therefore not able to adjust for such factors or assess whether they predict incomplete CD4 recovery.
Our definition of CD4 recovery (CD4 count >200 cells/μL after 3 years of viral suppression following cART initiation) differs from other definitions of recovery: for example, National Institutes of Health guidelines defined an adequate response as an increase in CD4 count in the range of 50-150 cells/μL per year . Our definition of CD4 recovery was based on 2 considerations. First, the increase in CD4 count over time since start of cART in virologically suppressed patients depends on baseline CD4 count . Second, the risk of mortality is strongly related to current CD4 count [16, 17].
Results in Context With Other Literature
Previous studies have identified age [8, 18, 19] and low baseline CD4 counts [4, 18, 20] as risk factors for incomplete CD4 cell recovery. However, 2 of these [16, 20] used designs that differed from ours in terms of inclusion criteria and length of the virally suppressed period. The effect of age observed in our study is consistent with its association with thymus size and activity and suggests that initiation of cART before immune incompetence occurs is especially important in older HIV-infected individuals [21-23].
Most studies that examined clinical endpoints according to achieved CD4 count in virally suppressed patients used short periods of viral suppression or did not have inclusion criteria relating to baseline CD4 count [9, 17, 24]. These studies mainly estimated short-term effects of changes in CD4 count, rather than elucidating the long-term impact of sustained low CD4 cell counts in patients who are virally suppressed long-term. Two previous studies examined implications of incomplete CD4 recovery among patients with sustained viral suppression, and our findings are in accordance with their estimates. Loutfy et al observed a 2.69-fold (95% CI, 1.26- to 5.78-fold) increase in mortality comparing 176 patients with a CD4 count <200 cells/μL with 1545 patients with a CD4 count >200 cells/μL after 2 years of viral suppression . That study had both a shorter duration of and less strict criteria for viral suppression as well as a much smaller and less generalizable sample. We previously reported the relative risk of death to be 3.4 (95% CI, 1.4-8.0) in a small study comparing 55 immunological nonresponders with 236 responders in the Danish HIV Cohort Study, using a design almost identical to that of the present study . Data on causes of death were not available in either of these 2 studies. Although we observed increased mortality in all subgroups of individuals with poor CD4 recovery, the relative risk of death was greatest among individuals with baseline CD4 count between 151 and 200 cells/μL. These individuals are characterized by the lowest increase (or even a decrease) in CD4 counts during the 3-year period of virological suppression, a phenomenon that may be related to non-HIV comorbidity such as cancer.
Consistent with our results, previous studies have found incomplete CD4 recovery to be associated with an increased risk of non-AIDS cancer [25-27]. It is still a matter of debate whether low CD4 counts lead to non-AIDS-defining cancers or whether common risk factors lead to low CD4 count and non-AIDS-defining cancers. Higher rates of non-AIDS mortality among individuals with incomplete CD4 recovery may be related to chronic immune activation in virally suppressed individuals [28, 29].
Implications and Conclusions
Our data underline the importance of early diagnosis of HIV and treatment with cART before patients have a low CD4 count. Although we have identified risk factors for poor CD4 recovery, no interventions to increase CD4 count in virally suppressed patients have been demonstrated to have beneficial effects on clinical endpoints and mortality. Previous studies have not consistently demonstrated differences between antiretroviral drug classes in effects on CD4 increases, and attempts to increase CD4 count with interleukin 2 were futile in terms of clinical benefit [30-32]. Virally suppressed patients with low CD4 counts should be monitored closely for diseases not conventionally considered to be HIV related, especially non-AIDS-defining cancers and liver diseases. Our study demonstrated an increased risk of non-AIDS causes of death in immunological nonresponders; further research is needed to elucidate the mechanisms that lead to persistently low CD4 counts despite viral suppression.