Undetectable = Untransmittable and Your Health: The Personal Benefits of Early and Continuous Therapy for HIV Infection

(See the Major Article by Borges et al on pages 254-63.)

In 2016, the Prevention Action Campaign launched the “U = U” (Undetectable = Untransmittable) campaign to publicize the preventative benefits of human immunodeficiency virus (HIV) antiretroviral therapy (ART) [1]. Based on accruing evidence from randomized clinical trials [2] and observational cohorts [3], the campaign intends to widely disseminate the message that an undetectable viral load means the virus is untransmittable to sexual partners and offspring. Indeed, the preventative benefits of effective HIV therapy have been responsible for averting vertical transmission to an estimated 1.6 million children, and to countless others through reductions in sexual transmission.

Yet, it remains to be seen if the U = U campaign will be sufficient to promote early and uninterrupted HIV treatment to an extent requisite to curb the epidemic. Effectiveness data in support of treatment as prevention at the population level have not been completely promising. For example, a large population-based, test-and-treat trial in rural South Africa failed to decrease HIV incidence, largely because of long hiatuses between HIV diagnosis and initiation of ART [4]. Similarly, copious data from various global regions have demonstrated persistently low CD4 counts at the time of treatment initiation, despite free access to testing and treatment [5, 6].

Although interrupting transmission is a critical public health goal of HIV treatment, notably absent from the U = U messaging is information about whether early and consistent ART has an effect on personal health. Whereas protection of transmission to partners and neonates is a priority for many, it might not be seen as a panacea to important subpopulations, such as preadolescents, those who are not sexually active, and those in monogamous relationships with a known HIV-positive partner. Moreover, for clinicians treating patients living with HIV, the individual health benefits of ART are arguably paramount for promoting initiation of and adherence to treatment for those in their care. Consequently, an important priority for the scientific and the public health community is to reinforce and communicate the health benefits of early and continuous ART among people with HIV.

To that end, in this issue of the Journal of Infectious Diseases, Borges et al [7] present results of a pooled analysis of 2 of the largest studies comparing ART treatment strategies: early versus delayed therapy in the Strategic Timing of Antiretroviral Treatment (START) trial [8] and continuous versus interrupted therapy in the Strategies for Management of Antiretroviral Therapy (SMART) trial [9]. Notwithstanding the heterogeneity in the study populations observed, the combination of 2 large studies (total N = 10156) offers a compelling reinforcement of the merits of early and continuous HIV treatment. The authors demonstrate protective effects across a range of outcomes, including all-cause mortality, serious non-AIDS events, cardiovascular disease events, and cancer incidence (combined as both AIDS and non-AIDS cancers). Moreover, the benefits of early and continuous ART were cross-cutting: a reduced hazard of a composite outcome of death, AIDS, or serous non-AIDS events was seen across subgroups of age, both sexes, a range of CD4 counts at enrollment, and among residents of high- and low-income countries.

Although prior studies have identified benefits of early and persistent ART, these data offer an important opportunity to elaborate the magnitude and breadth of the individual health benefits of early and lifelong ART. For example, although reductions in all-cause mortality have been suggested by earlier trials that randomized individuals at relatively low CD4 count thresholds to immediate versus delayed ART, or 2 versus 3 active drugs [9–11], a clear demonstration of a significant mortality benefit from early ART has been absent from recent studies with higher CD4 count thresholds [8, 12, 13] (Figure 1). Consequently, an important strength of the current analysis is its use of pooled data across a broad swath of CD4 count thresholds, to strengthen both generalizability and power not available in prior clinical trials. Because it is highly unlikely that future studies will randomize individuals with HIV to a delayed treatment arm, such pooled analyses likely remain the only extant scientific opportunity to assess the treatment benefits of ART for outcomes with smaller effect sizes.

Figure 1.

Relative hazard of mortality with combination antiretroviral therapy in randomized clinical trials [8–13]. Abbreviations: ACTG, AIDS Clinical Trials Group; ART, antiretroviral therapy; HPTN, HIV Prevention Trials Network; SMART, Strategies for Management of Antiretroviral Therapy; START, Strategic Timing of Antiretroviral Treatment.

Yet, even a pooled analysis of over 10000 individuals enrolled in randomized controlled trials tests the limits of this approach. Because many clinical events are sufficiently rare and/or can escape detection in clinical trials, this analysis is likely under-powered to demonstrate differences in some important outcomes between subgroups. For example, the increased hazard of cardiovascular events was 1.77 in the SMART study, but only 1.18 in the START study, in which nadir CD4 counts were relatively high even in the delayed therapy group (>400 cells/μL). Although this difference—a 50% higher protective benefit comparing the SMART versus START study populations—was not statistically significant (P = .37), there is a strong rationale from both observational cohorts and mechanistic studies to suggest that protective effects of ART against cardiovascular disease might be most pronounced for those with more advanced disease at ART initiation [14–21]. Because interactive effects require approximately 4 times as much power to detect as individual treatment effects [22], the interaction terms presented in this article should not be interpreted to mean that subgroup effects were similar. Alternatively, as the authors discuss, differences in cardioprotective effects might have been due to unique participant risk profiles, or challenges in detecting cardiovascular disease events, particularly in resource-limited settings, where an important minority of START study subjects were based.

Nonetheless, the investigators did find a statistically significant difference in protective effects of ART against cancer, with a greater benefit in the START versus the SMART therapy groups (hazard ratio 3.10 for continuous versus intermittent ART, compared with 1.37 for immediate versus deferred ART, P value for interaction term .046). Notably, there were nearly twice as many total cancers reported in the START study overall (53 vs 28), and a major contributor to the benefit in START was through prevention of Kaposi sarcoma and non-Hodgkin lymphoma, which were reported in 21 and 4 participants in the delayed versus immediate treatment groups, respectively. This result was particularly meaningful in light of recent data suggesting persistently high risk of both AIDS and non-AIDS cancers among individuals on long-term ART [23]. Future pooled analyses, which include the TEMPRANO study, might better elucidate the benefit of early ART in subgroups, particularly for rarer outcomes, such as all-cause mortality and non-AIDS events.

A somewhat unexpected finding was that the beneficial effect of ART did not appear to differ greatly between those with higher versus lower levels of inflammatory markers prior to study initiation. On face value, this appears to contradict prior data from the SMART study, and others, that demonstrate how pretreatment levels of inflammation impact cardiovascular disease events and mortality [24–28]. However, more recently, other studies have suggested that time updated levels of disease control and inflammatory markers after ART initiation, as opposed to pretreatment levels, are more predictive of preclinical atherosclerosis and cardiovascular disease events [29–31], and that, particularly for non-AIDS events, inflammatory markers might not rise until just prior to events [32]. Moreover, SMART and START reported relatively nonspecific inflammatory and coagulation markers, as opposed to markers of monocyte activation or endothelial dysfunction, which have been more closely linked to non-AIDS events in persons with HIV. Importantly, this analysis appears to support the concept that early and continuous ART, irrespective of pretreatment biomarkers, has substantial health benefits.

To optimize clarity for patients who may request information about the full range of risks and benefits of earlier treatment, future work should attempt to disentangle toxicity data from other clinical outcomes. Borges et al [7] touch briefly on safety of early and continuous ART by describing rates of composite outcomes including grade 4 events, deaths, and hospitalizations separately in the 2 studies, due to differences in how toxicity data were collected. Similarly, although incidence data and hazard ratios offer important relative risk estimation, the absolute risk reductions, which were largely absent from this analysis, will enable clinicians and patients to better interpret the individual benefits of early and continuous therapy.

In summary, through increased power provided by a pooled analysis, Borges et al [7] contextualize the benefits of early and continuous ART for people with HIV infection. They strongly corroborate prior work demonstrating that immediate and continuous ART is not only crucial to protect the health of partners and offspring, but that it will also prolong life and health for those who are taking it, even if they have early and asymptomatic disease. Importantly, these health benefits increasingly appear to extend beyond AIDS and opportunistic infections, to include decreased risk of hospitalizations and noncommunicable diseases, which are particularly relevant to the aging population with HIV. Whereas many clinicians who treat HIV infection were likely aware of these effects before this study was published [33], there are accruing data that such protective individual benefits of ART are not always known among people with HIV [34, 35]. Reaffirming these data in the public sphere will be a critical step towards improving ART uptake, and improving health both for people with HIV and their partners and children.

Notes

Financial support. This work was supported by the National Institutes of Health (grant numbers K23 MH099916, P30 AG024409, R24 AG 044325, P30 AI060354 to M. J. S).

Potential conflicts of interest. Authors report no potential conflicts of interest. Both authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.