aForum for Collaborative HIV Research, UC Berkeley School of Public Health, Washington DC
bMid-Atlantic Permanente Research Institute, Kaiser Permanente, Rockville Maryland, USA.
Correspondence to Veronica Miller, University of California Berkeley Washington DC, USA. E-mail: firstname.lastname@example.org.
The question of whether HIV patients on antiretroviral therapy (ART) may expect to live a ‘normal’ lifespan is of major interest and has significant care strategy and financing policy implications. Independent of the public health benefit of immediate antiretroviral treatment, how early is early enough for achieving a normal lifespan? What are the long-term cost implications? If a normal lifespan is possible, what are the conditions under which this may happen?
Different approaches are available and have been used to investigate excess (or specific disease associated) mortality among HIV-infected populations. Some studies look for evidence that one treatment strategy (e.g. starting treatment earlier) improves mortality outcomes compared with a different strategy (e.g. starting treatment later) . Although this approach can tell us ‘whether’, and if so, ‘what’ treatment strategies reduce mortality, it does not tell us whether the reduced mortality approaches the ‘normal’ for non-HIV infected individuals. To best evaluate questions of extended life expectancy among HIV-infected patients, we need to compare HIV-infected populations with non-HIV infected populations, using methodologies developed for studying chronic diseases, such as actuarial projections. We argue that both approaches are necessary to address the question of lifespan. The first allows us to fine-tune treatment strategies to incrementally improve outcome. The second offers the ultimate reality check – have we done the best we can?
In a new study focusing on mortality rates in HIV-infected individuals compared with the general population, Rodger et al. focus on noninjection drug using individuals from the control arms of two randomized clinical trials. The SMART and ESPRIT were conducted from 2000 to 2006 in over 33 different countries, spanning Europe and North America. HIV infected population are compared with the general population using the Standardized Mortality Ratio (SMR). Patients included in the analysis (n = 3280) were treated with ART according to the historical and geographically appropriate standard-of-care. The authors found no evidence of increased mortality in individuals with an undetectable viral load and a most recent CD4 cell count more than 500 per μl. Those whose CD4 count fell in the range of 350–500 per μl did exhibit a higher SMR. The authors state (and we agree) that their data support the importance of early diagnosis and treatment to improve clinical outcome.
The main finding of the study, that the current CD4 count, if above 500 per μl, predicts mortality rates similar to the general population, implies that those who started with a lower count, but were able to respond with a CD4 cell increase to above 500 per μl, experience normal mortality rates. In this analysis, an effect of the CD4 nadir was not evident, but mortality (vs. CVD or other events influenced by inflammatory status) may be too crude a tool to discern such an impact.
This study confirms earlier findings – life expectancy for HIV-infected individuals approaches that of general populations under ‘ideal’ conditions [3,4]. What distinguishes this study is the fact that the HIV patient cohort was derived from two randomized clinical studies, with stringent criteria for death ascertainment, which may not always be true for clinic-based cohort studies or registries in many countries. This should lead to more confidence in the findings of the study. The question is: are we confident enough? Cohort and registry-based studies will always carry a risk of unmeasured confounding. Randomized clinical trials do not carry this risk.
Whereas randomized studies may contribute to questions such as the long-term impact of different treatment strategies, the question of whether a treatment strategy leads to normal lifespan can realistically only be addressed through large databases, registries, and long-term observational cohort studies. These studies need to be sufficiently long and with large enough n to ensure that any excess mortality, should it exist, has the opportunity of being expressed and is detectable over and above the general population rate all cause mortality. It is possible that comorbidities such as cardiovascular disease, kidney and liver disease, cancers other than AIDS-defining malignancies, may have greater impact as the HIV population ages, and are not adequately captured in present studies and cohorts. Smoking effects are usually not seen till later in life; thus, the study likely under-represents the true mortality rate of the cohort. As Rodger et al. note, the SMART and ESPRIT cohorts are relatively young, thus likely not capturing the true mortality rate. Encouragingly, health habits are changing; negative health habits such as smoking, which studies have shown have a greater prevalence in communities most impacted by HIV, are decreasing. Thus, temporal trends also will play a role in life expectancy determinations going forward. And while mortality is the ultimate ‘outcome metric’, focusing on that alone does not tell the whole story. If other events can occur with higher CD4 counts, including liver and cardiovascular events, lifespan may be an incomplete outcome goal for determining when to initiate therapy. These issues will not be resolved by randomized clinical trials, which by nature are not likely to be long enough to catch the late-in-life mortality events, which may or may not be disease associated.
The population being studied will also affect the outcome. For example, some studies focus on MSM who start early treatment , but general population rates for non-HIV infected MSM may not be available. Others, such as those by Rodgers et al.[2,4] exclude IDUs. The impact of these inclusions and/or exclusions on the ultimate mortality reduction effect may not be sufficiently large to really matter. While earlier studies reported increased morbidity and mortality in IDUs and racial/ethnic minorities (attributable to health disparities), more recent studies demonstrate equivalent morbidity and mortality rates across MSM, IDUs and heterosexual populations, including a majority of patients living below the national poverty rate, as long as the patients are retained in care .
If we add the findings of this study to our arsenal of studies supporting early initiation of antiretroviral treatment, where does that leave us? Clinical guidelines have called the shot: early treatment is recommended by DHHS, WHO, and IAS-USA HIV treatment guidelines.
To the question ‘Getting to normal: are we there yet?’, the answer appears to be getting closer and closer to ‘yes’.
Conflicts of interest
There are no conflicts of interest.
1. Kitahata MM, Gange SJ, Abraham AG, Merriman B, Saag MS, Justice AC, et al. Effect of early versus deferred antiretroviral therapy for HIV on survival
. N Engl J Med
2. Rodger A, Lodwick R, Schechter M, Deeks S, Amin J, Gilson R, et al. Mortality in well controlled HIV in the continuous antiretroviral therapy arms of the SMART and ESPRIT trials compared with the general population
3. van Sighem AI, Gras LA, Reiss P, Brinkman K, de Wolf F. Life expectancy of recently diagnosed asymptomatic HIV-infected patients approaches that of uninfected individuals
4. Lewden C, Bouteloup V, De Wit S, Sabin C, Mocroft A, Wasmuth JC, et al. All-cause mortality in treated HIV-infected adults with CD4 >/= 500/mm3 compared with the general population: evidence from a large European observational cohort collaboration
. Int J Epidemiol
5. Moore RD, Keruly JC, Bartlett JG. Improvement in the health of HIV-infected persons in care: reducing disparities
. Clin Infect Dis