Survival benefits of antiretroviral therapy in Brazil: a model‐based analysis

Paula M. Luz, Michael P. Girouard, Beatriz Grinsztejn, Kenneth A. Freedberg, Valdiléa G. Veloso, Elena Losina, Claúdio J. Struchiner, Rachel L. MacLean, Robert A. Parker, A. David Paltiel, Rochelle P. WalenskyJournal of the International AIDS Society 2016 · DOI

TL;DR

Brazil's universal, free antiretroviral therapy (ART) program, initiated in 1996, has yielded substantial survival benefits for HIV-positive individuals. This model-based analysis quantifies these benefits, showing that while a significant portion has been realized, the majority of life-years saved are yet to come, driven by improvements in ART efficacy and earlier treatment initiation.

What Problem Does This Paper Address?

The paper addresses the critical need to quantify the population-level impact of Brazil's pioneering universal antiretroviral therapy (ART) program on survival. While the program's existence is well-documented, a comprehensive, model-based analysis of its cumulative and projected survival benefits over time, considering the evolution of treatment regimens and patient care, was lacking. This is crucial for demonstrating the program's value, securing continued public support, and providing evidence that could influence similar initiatives in other low- and middle-income countries facing similar HIV epidemics.

Methods

The researchers employed a microsimulation model called CEPAC-International, a well-established tool for modeling HIV disease progression and treatment outcomes. They adapted this model using data specific to Brazil to estimate life expectancy increases for HIV-positive patients initiating ART between 1997 and 2014.

The core methodology involved:

  • Era-Based Simulation: The study period was divided into six "eras" (1997-1999, 2000-2003, etc.) to reflect advancements in ART regimen efficacy, availability, and changes in patient characteristics (like CD4 count at ART initiation) over time.
  • Counterfactual Modeling: For each era, simulated cohorts of patients were modeled under two scenarios: one receiving ART according to Brazilian guidelines and another receiving only opportunistic infection (OI) prophylaxis (the standard of care without ART).
  • Life Expectancy Calculation: Life expectancy was calculated for each scenario, censored at December 2014 for 2014-censored benefits and projected to lifetime benefits.
  • Survival Benefit Quantification: The per-person survival benefit for each era was the difference in life expectancy between the ART and no-ART scenarios. This was then multiplied by the estimated number of patients initiating ART in that era to derive total survival benefits.
  • Sensitivity Analyses: To assess the robustness of their findings, the researchers conducted one-way sensitivity analyses by varying key model parameters (e.g., CD4 count at ART initiation, adherence rates, ART failure rates) within plausible ranges. They also performed best-case and worst-case scenario analyses.

Datasets

The study utilized data from several sources:

  • Brazilian Ministry of Health: Provided annual data on the number of patients receiving ART through the national program from 1999 to 2012, which was used to estimate patient numbers for each era.
  • HIV Clinical Cohort at the Instituto Nacional de Infectologia Evandro Chagas (INI): This cohort, comprising over 5500 HIV-positive patients since 1986, provided client-level data for model parameterization, including cohort characteristics and natural history parameters.
  • São Paulo Pharmacy Withdrawal Study: This study provided data on ART pharmacy withdrawals, used as a proxy for medication adherence.
  • United Nations Department of Economic and Social Affairs: Provided life tables for Brazil to estimate non-AIDS mortality.
  • Published Literature: Data from clinical trials and other studies were used to inform parameters related to ART efficacy, virologic suppression, and failure rates.

Key Insights and Ideas

  • Dramatic Survival Gains: The most striking insight is the immense survival benefit conferred by Brazil's ART program, translating into millions of "life-years saved" for the population.
  • Future Benefits: A significant portion of these life-years saved are yet to be realized, highlighting the long-term impact of sustained ART access.
  • Evolution of Treatment Matters: The study clearly demonstrates how improvements in ART regimens over time (reflected in the different eras) directly correlate with increasing survival benefits.
  • Early Treatment is Key: Higher CD4 counts at the time of ART initiation, a consequence of earlier diagnosis and treatment, significantly contribute to greater survival gains.
  • Model Robustness: The sensitivity analyses indicate that the core findings regarding survival benefits are relatively stable across a range of parameter variations, increasing confidence in the results.

Key Results

  • Total Projected Lifetime Survival Benefit: Approximately 9.3 million life-years are projected to be saved by the Brazilian national ART program for patients initiating treatment between 1997 and 2014.
  • Realized Benefits: As of December 2014, about 1.5 million life-years (16% of the total projected benefit) had been realized.
  • Increasing Per-Person Benefits: Per-person life expectancy increases with ART were substantial and grew across the eras. For instance, in Era 1 (1997-1999), the per-person benefit was 8.3 years, while in Era 6 (2014), it reached 19.9 years.
  • Significant Improvement in 20-Year Survival: The proportion of patients alive at 20 years post-treatment initiation increased dramatically from 0.2% without ART to 20.8% with ART in Era 1, and from 3.4% without ART to 60.6% with ART in Era 6.
  • Most Influential Parameters: Mean CD4 count at ART initiation was the most influential parameter in sensitivity analyses, followed by adherence rates and ART failure rates.

Limitations and Future Work

The authors acknowledge several limitations:

  • Single-Site Data: Input parameters for cohort characteristics and adherence were largely derived from a single reference center (INI) and a study in São Paulo, which may not perfectly represent the entire country.
  • Exclusion of Other Interventions: The analysis focused solely on the survival benefits of ART and did not explicitly model the impact of HIV prevention (e.g., mother-to-child transmission), treatment of children, or the reduction of horizontal transmission through suppressed viral load.
  • No Economic Returns: The study did not quantify the economic returns of ART treatment.
  • No Long-Term Co-morbidities: The current model version does not include long-term co-morbidities related to ART exposure, which could potentially reduce projected survival benefits.
  • Regional Heterogeneity: The analysis did not incorporate regional demographic variations in mortality or other key parameters, meaning it may not fully capture the heterogeneity across Brazil.
  • ART Initiation Criteria: While the model reflects the evolution of ART guidelines, it did not explicitly model changes in Brazil's ART initiation criteria, assuming patients initiated ART later than recommended in general.

Future work could involve incorporating the impact of other interventions, exploring economic returns, and developing models that account for long-term ART-related co-morbidities and regional differences. The authors also suggest that improved HIV testing, earlier ART initiation, better linkage to care, and interventions to enhance ART adherence could lead to even greater future survival benefits.

Summary updated

Abstract

Provision of ART through a national programme has led to dramatic survival benefits in Brazil, the majority of which are still to be realized. Improvements in initial and subsequent ART regimens and higher CD4 counts at ART initiation have contributed to these increasing benefits.
BibTeX
@article{luz2016survival,
  author = {Paula M. Luz and Michael P. Girouard and Beatriz Grinsztejn and Kenneth A. Freedberg and Valdiléa G. Veloso and Elena Losina and Claúdio J. Struchiner and Rachel L. MacLean and Robert A. Parker and A. David Paltiel and Rochelle P. Walensky},
  title = {{Survival benefits of antiretroviral therapy in Brazil: a model‐based analysis}},
  journal = {Journal of the International AIDS Society},
  year = {2016},
  doi = {10.7448/ias.19.1.20623},
}
Figure