Cumulative Human Immunodeficiency Viremia, Antiretroviral Therapy, and Incident Myocardial Infarction

Gaps in the usage and reporting of multiple imputation for incomplete data: findings from a scoping review of observational studies addressing causal questions

BACKGROUND

Missing data are common in observational studies and often occur in several of the variables required when estimating a causal effect, i.e. the exposure, outcome and/or variables used to control for confounding. Analyses involving multiple incomplete variables are not as straightforward as analyses with a single incomplete variable. For example, in the context of multivariable missingness, the standard missing data assumptions ("missing completely at random", "missing at random" [MAR], "missing not at random") are difficult to interpret and assess. It is not clear how the complexities that arise due to multivariable missingness are being addressed in practice. The aim of this study was to review how missing data are managed and reported in observational studies that use multiple imputation (MI) for causal effect estimation, with a particular focus on missing data summaries, missing data assumptions, primary and sensitivity analyses, and MI implementation.

METHODS

We searched five top general epidemiology journals for observational studies that aimed to answer a causal research question and used MI, published between January 2019 and December 2021. Article screening and data extraction were performed systematically.

RESULTS

Of the 130 studies included in this review, 108 (83%) derived an analysis sample by excluding individuals with missing data in specific variables (e.g., outcome) and 114 (88%) had multivariable missingness within the analysis sample. Forty-four (34%) studies provided a statement about missing data assumptions, 35 of which stated the MAR assumption, but only 11/44 (25%) studies provided a justification for these assumptions. The number of imputations, MI method and MI software were generally well-reported (71%, 75% and 88% of studies, respectively), while aspects of the imputation model specification were not clear for more than half of the studies. A secondary analysis that used a different approach to handle the missing data was conducted in 69/130 (53%) studies. Of these 69 studies, 68 (99%) lacked a clear justification for the secondary analysis.

CONCLUSION

Effort is needed to clarify the rationale for and improve the reporting of MI for estimation of causal effects from observational data. We encourage greater transparency in making and reporting analytical decisions related to missing data.

Systematic review on cumulative HIV viraemia among people living with HIV receiving antiretroviral treatment and its association with mortality and morbidity

BACKGROUND

We performed a systematic review to generate evidence on the association between cumulative human immunodeficiency virus (HIV) viraemia and health outcomes.

METHODS

Quantitative studies reporting on HIV cumulative viraemia (CV) and its association with health outcomes among people living with HIV (PLHIV) on antiretroviral treatment (ART) were included. We searched MEDLINE via PubMed, Embase, Scopus and Web of Science and conference abstracts from 1 January 2008 to 1 August 2022.

RESULTS

The systematic review included 26 studies. The association between CV and mortality depended on the study population, methods used to calculate CV and its level. Higher CV was not consistently associated with greater risk of acquire immunodeficiency syndrome-defining clinical conditions. However, four studies present a strong relationship between CV and cardiovascular disease. The risk was not confirmed in relation of increased hazards of stroke. Studies that assessed the effect of CV on the risk of cancer reported a positive association between CV and malignancy, although the effect may differ for different types of cancer.

CONCLUSIONS

CV is associated with adverse health outcomes in PLHIV on ART, especially at higher levels. However, its role in clinical and programmatic monitoring and management of PLHIV on ART is yet to be established.

Cumulative Human Immunodeficiency Virus (HIV)-1 Viremia Is Associated With Increased Risk of Multimorbidity Among US Women With HIV, 1997-2019

Background

To evaluate the effect of cumulative human immunodeficiency virus (HIV)-1 viremia on aging-related multimorbidity among women with HIV (WWH), we analyzed data collected prospectively among women who achieved viral suppression after antiretroviral therapy (ART) initiation (1997-2019).

Methods

We included WWH with ≥2 plasma HIV-1 viral loads (VL) <200 copies/mL within a 2-year period (baseline) following self-reported ART use. Primary outcome was multimorbidity (≥2 nonacquired immune deficiency syndrome comorbidities [NACM] of 5 total assessed). The trapezoidal rule calculated viremia copy-years (VCY) as area-under-the-VL-curve. Cox proportional hazard models estimated the association of time-updated cumulative VCY with incident multimorbidity and with incidence of each NACM, adjusting for important covariates (eg, age, CD4 count, etc).

Results

Eight hundred six WWH contributed 6368 women-years, with median 12 (Q1-Q3, 7-23) VL per participant. At baseline, median age was 39 years, 56% were Black, and median CD4 was 534 cells/mm3. Median time-updated cumulative VCY was 5.4 (Q1-Q3, 4.7-6.9) log10 copy-years/mL. Of 211 (26%) WWH who developed multimorbidity, 162 (77%) had incident hypertension, 133 (63%) had dyslipidemia, 60 (28%) had diabetes, 52 (25%) had cardiovascular disease, and 32 (15%) had kidney disease. Compared with WWH who had time-updated cumulative VCY <5 log10, the adjusted hazard ratio of multimorbidity was 1.99 (95% confidence interval [CI], 1.29-3.08) and 3.78 (95% CI, 2.17-6.58) for those with VCY 5-6.9 and ≥7 log10 copy-years/mL, respectively (P < .0001). Higher time-updated cumulative VCY increased the risk of each NACM.

Conclusions

Among ART-treated WWH, greater cumulative viremia increased the risk of multimorbidity and of developing each NACM, and hence this may be a prognostically useful biomarker for NACM risk assessment in this population.

Associations between HIV viremia during antiretroviral therapy and cardiovascular disease

OBJECTIVE

To investigate the association between HIV viremia exposure during antiretroviral therapy (ART) and cardiovascular disease (CVD) risk.

DESIGN

Nationwide observational cohort.

METHODS

Participants (age >15 years) from the Swedish nationwide InfCareHIV register initiating ART 1996-2017 were categorized in a time-updated manner into four viremia categories, starting from 12 months after ART initiation: suppression (<50 copies/ml), low-level viremia (50-199 copies/ml and 200-999 copies/ml, respectively), and high-level viremia (≥1000 copies/ml). In addition, cumulative viremia was estimated as the area under the log viral load (VL) curve. Proportional subhazard models adjusted for sex, age, pre-ART CD4 and VL, injection drug use, and country of birth were used to analyze the association between viremia exposure and CVD risk (ischemic heart disease, stroke, and heart failure; data obtained by linkage to national registers), accounting for the competing risk of non-CVD death.

RESULTS

In all, 337 cases of CVD were observed during 44 937 person-years of follow-up ( n  = 6562). Higher viremia exposure was associated with CVD, both when parameterized as cumulative viremia (adjusted subhazard ratio [aSHR] per 1 log 10  copy × year/ml, 1.03; 95% confidence interval [CI], 1.01-1.05) and as viremia category (aSHR for high-level viremia versus suppression, 1.45; 95% CI, 1.03-2.05). We observed no association between CVD and low-level viremia compared with those with suppression.

CONCLUSIONS

Higher exposure to HIV viremia was linked to CVD in ART recipients, whereas no increased risk was detected for people with low-level viremia compared with viral suppression. Causal inference is limited by the observational nature of this study.

Brief Report: Insomnia and Risk of Myocardial Infarction Among People With HIV

BACKGROUND

Insomnia is common among people with HIV (PWH) and may be associated with increased risk of myocardial infarction (MI). This study examines the association between insomnia and MI by MI type among PWH.

SETTING

Longitudinal cohort study of PWH at 5 Centers for AIDS Research Network of Integrated Clinical Systems sites.

METHODS

Clinical data and patient-reported measures and outcomes from PWH in care between 2005 and 2018 were used in this study. Insomnia, measured at baseline, was defined as having difficulty falling or staying asleep with bothersome symptoms. The Centers for AIDS Research Network of Integrated Clinical Systems centrally adjudicates MIs using expert reviewers, with distinction between type 1 MI (T1MI) and type 2 MI (T2MI). Associations between insomnia and first incident MI by MI type were measured using separate Cox proportional hazard models adjusted for age, sex, race/ethnicity, traditional cardiovascular disease risk factors (hypertension, dyslipidemia, poor kidney function, diabetes, and smoking), HIV markers (antiretroviral therapy, viral suppression, and CD4 cell count), and stimulant use (cocaine/crack and methamphetamine).

RESULTS

Among 12,448 PWH, 48% reported insomnia. Over a median of 4.4 years of follow-up, 158 T1MIs and 109 T2MIs were identified; approximately half of T2MIs were attributed to sepsis or stimulant use. After adjustment for potential confounders, we found no association between insomnia and T1MI (hazard ratio = 1.05, 95% confidence interval: 0.76 to 1.45) and a 65% increased risk of T2MI among PWH reporting insomnia compared with PWH without insomnia (hazard ratio = 1.65, 95% confidence interval: 1.11 to 2.45).

CONCLUSIONS

PWH reporting insomnia are at an increased risk of T2MI, but not T1MI, compared with PWH without insomnia, highlighting the importance of distinguishing MI types among PWH.

HIV, Aging, and Comorbidities Research in Clinical Cohorts: 3 Lessons Learned Using Examples From the CNICS Cohort

BACKGROUND

Owing to ongoing improvements in antiretroviral therapy, people with HIV (PWH) are achieving near-normal lifespans with many surviving into middle and old age. Despite this success, PWH have a higher than expected risk of developing non-AIDS comorbidities, multimorbidity, and functional decline at ages younger than those without HIV.

METHODS

As part of the Inter-CFAR (Center for AIDS Research) Symposium, HIV and Aging in the era of Antiretroviral Therapy and COVID-19, we presented a research update from HIV clinical cohorts and specifically described 3 lessons learned from the Centers for AIDS Research Network of Integrated Clinical Systems cohort that are important for HIV and aging research moving forward.

RESULTS

Adjudicated outcomes are particularly beneficial for less common comorbidities such as myocardial infarction. Multiple ascertainment approaches increase sensitivity over using diagnoses alone (89% vs. 44%). Adjudication eliminates false-positive events and allows myocardial infarction types to be identified. Comorbidity research has often relied on composite outcomes, such as all cardiovascular diseases, often to increase power. Mechanistic differences across outcomes demonstrate the importance of moving away from many composite outcomes. Timely data are needed to ensure findings are relevant to improve care or outcomes for the population of PWH who are currently aging.

CONCLUSIONS

A better understanding of the causes, mechanisms, prevention and treatment of functional decline, comorbidities, and multimorbidities is a crucial research focus as PWH are aging. Clinical cohorts with timely, comprehensive harmonized clinical data and carefully adjudicated outcomes are ideally positioned to improve understanding of these questions.

Determinants of loss to care and risk of clinical progression in PLWH who are re-engaged in care after a temporary loss

The risk of developing AIDS is elevated not only among those with a late HIV diagnosis but also among those lost to care (LTC). The aims were to address the risk of becoming LTC and of clinical progression in LTC patients who re-enter care. Patients were defined as LTC if they had no visit for ≥ 18 months. Of these, persons with subsequent visits were defined as re-engaged in care (RIC). Factors associated with becoming LTC and RIC were investigated. The risk of disease progression was estimated by comparing RIC with patients continuously followed. Over 11,285 individuals included, 3962 became LTC, and of these, 1062 were RIC. Older age, presentation with AIDS and with higher HIV-RNA were associated with a reduced risk of LTC. In contrast, lower education level, irregular job, being an immigrant and injecting-drug user were associated with an increased LTC probability. Moreover, RIC with HIV-RNA > 200 copies/mL at the re-entry had a higher risk of clinical progression, while those with HIV-RNA ≤ 200 copies/mL had a higher risk of only non-AIDS progression. Patients re-entering care after being LTC appeared to be at higher risk of clinical progression than those continuously in care. Active strategies for re-engagement in care should be promoted.

HIV Viremia and Risk of Stroke Among People Living with HIV Who Are Using Antiretroviral Therapy

BACKGROUND

Rates of stroke are higher in people living with HIV compared with age-matched uninfected individuals. Causes of elevated stroke risk, including the role of viremia, are poorly defined.

METHODS

Between 1 January 2006 and 31 December 2014, we identified incident strokes among people living with HIV on antiretroviral therapy at five sites across the United States. We considered three parameterizations of viral load (VL) including (1) baseline (most recent VL before study entry), (2) time-updated, and (3) cumulative VL (copy-days/mL of virus). We used Cox proportional hazards models to estimate hazard ratios (HRs) for stroke risk comparing the 75th percentile ("high VL") to the 25th percentile ("low VL") of baseline and time-updated VL. We used marginal structural Cox models, with most models adjusted for traditional stroke risk factors, to estimate HRs for stroke associated with cumulative VL.

RESULTS

Among 15,974 people living with HIV, 139 experienced a stroke (113 ischemic; 18 hemorrhagic; eight were unknown type) over a median follow-up of 4.2 years. Median baseline VL was 38 copies/mL (interquartile interval: 24, 3,420). High baseline VL was associated with increased risk of both ischemic (HR: 1.3; 95% CI = 0.96-1.7) and hemorrhagic stroke (HR: 3.1; 95% CI = 1.6-5.9). In time-updated models, high VL was also associated with an increased risk of any stroke (HR: 1.8; 95% CI = 1.4-2.3). We observed no association between cumulative VL and stroke risk.

CONCLUSIONS

Our findings are consistent with the hypothesis that elevated HIV VL may increase stroke risk, regardless of previous VL levels.

Effects of HIV Infection on Arterial Endothelial Function: Results From a Large Pooled Cohort Analysis

OBJECTIVE

To determine the effects of HIV serostatus and disease severity on endothelial function in a large pooled cohort study of people living with HIV infection and HIV- controls. Approach and Results: We used participant-level data from 9 studies: 7 included people living with HIV (2 treatment-naïve) and 4 had HIV- controls. Brachial artery flow-mediated dilation (FMD) was measured using a standardized ultrasound imaging protocol with central reading. After data harmonization, multiple linear regression was used to examine the effects of HIV- serostatus, HIV disease severity measures, and cardiovascular disease risk factors on FMD. Of 2533 participants, 986 were people living with HIV (mean 44.4 [SD 11.8] years old) and 1547 were HIV- controls (42.9 [12.2] years old). The strongest and most consistent associates of FMD were brachial artery diameter, age, sex, and body mass index. The effect of HIV+ serostatus on FMD was strongly influenced by kidney function. In the highest tertile of creatinine (1.0 mg/dL), the effect of HIV+ serostatus was strong (β=-1.59% [95% CI, -2.58% to -0.60%], P=0.002), even after covariate adjustment (β=-1.36% [95% CI, -2.46% to -0.47%], P=0.003). In the lowest tertile (0.8 mg/dL), the effect of HIV+ serostatus was strong (β=-1.90% [95% CI, -2.58% to -1.21%], P<0.001), but disappeared after covariate adjustment. HIV RNA viremia, CD4+ T-cell count, and use of antiretroviral therapy were not meaningfully associated with FMD.

CONCLUSIONS

The significant effect of HIV+ serostatus on FMD suggests that people living with HIV are at increased cardiovascular disease risk, especially if they have kidney disease.

Comparison of cumulative viraemia following treatment initiation with different antiretroviral regimens: a real-life study in Brazil

Introduction

The relative efficacy of different antiretroviral (ART) regimens has been extensively evaluated in the context of clinical trials, using HIV viral load (VL) measurements at pre‐specified timepoints after ART onset. However, data from real‐life studies using combined longitudinal measurements of cumulative viraemia are scarce. This study aimed to address the independent effect of different ART regimens on HIV cumulative viraemia over the first 12 months after treatment initiation, using programmatic data from the Ministry of Health of Brazil.

Methods

Retrospective cohort study analysing cumulative viraemia under the most frequently used ART regimens in Brazil (tenofovir, lamivudine and dolutegravir (regimen 1); tenofovir, lamivudine and efavirenz (regimen 2); tenofovir, lamivudine and ritonavir‐boosted atazanavir (regimen 3)).

Results and Discussion

We included 112,243 patients >12 years old who received their first ART prescription between January 2014 and August 2017. Univariate analysis indicated that cumulative viraemia was significantly lower in patients receiving regimen 1 as compared with those receiving regimens 2 or 3 (p<0.0001 for both pairwise comparisons). In a multivariable analysis adjusted for age, sex, baseline T CD4+ counts and baseline HIV VL, ART regimen persisted with statistically significant effect on 12‐month cumulative viraemia. The model predicted a 45‐unit increase in log₁₀ copy‐days/mL cumulative viraemia for regimen 2 as compared with regimen 1, and a 70‐unit increase in log₁₀ copy‐days/mL cumulative viraemia for regimen 3 as compared with regimen 1 (95%CI 41 to 49 and 61 to 79 respectively; p<0.001 for both comparisons). In models restricted to youths (13 to 24 years old) and female patients, ART regimen had similar effects. ART regimen with dolutegravir in association with a tenofovir‐lamivudine backbone was superior to regimens containing efavirenz or boosted atazanavir in reducing HIV VL, as shown by cumulative viraemia over the first 12 months after treatment initiation. The superiority persisted even after adjusting the analysis for potential confounders.

Conclusions

Our findings could bring direct benefits to patients as suggested by lower viral replication during treatment, lower risk of HIV transmission, and a potential reduction in resistance mutations in the initial 12 months under ART.

HIV infection and coronary heart disease: mechanisms and management

Antiretroviral therapy has largely transformed HIV infection into a chronic disease condition. As such, physicians and other providers caring for individuals living with HIV infection need to be aware of the potential cardiovascular complications of HIV infection and the nuances of how HIV infection increases the risk of cardiovascular diseases, including acute myocardial infarction, stroke, peripheral artery disease, heart failure and sudden cardiac death, as well as how to select available therapies to reduce this risk. In this Review, we discuss the epidemiology and clinical features of cardiovascular disease, with a focus on coronary heart disease, in the setting of HIV infection, which includes a substantially increased risk of myocardial infarction even when the HIV infection is well controlled. We also discuss the mechanisms underlying HIV-associated atherosclerotic cardiovascular disease, such as the high rates of traditional cardiovascular risk factors in patients with HIV infection and HIV-related factors, including the use of antiretroviral therapy and chronic inflammation in the setting of effectively treated HIV infection. Finally, we highlight available therapeutic strategies, as well as approaches under investigation, to reduce the risk of cardiovascular disease and lower inflammation in patients with HIV infection.

Long-term evolution of CD4+ cell count in patients under combined antiretroviral therapy

OBJECTIVE

Combined antiretroviral treatment (cART) results in profound immunologic improvement, but it is unclear whether CD4 cell counts return to levels similar to those of HIV-negative individuals. We explore long-term CD4 cell count evolution post-cART and its association with baseline levels, virologic suppression, pre-cART cumulative viremia and other factors.

DESIGN

Data were derived from the AMACS. Included individuals were adults who started cART, at least 2003, while previously ART-naive.

METHODS

Changes in CD4 cell counts were modeled through piecewise linear mixed models.

RESULTS

A total of 3405 individuals were included. The majority was male (86.0%), homosexual (58.8%) with median (IQR) age at cART initiation 36 (31-44) years and a median (IQR) follow-up of 3.9 (2.0-6.9) years. Most persons (57%) starting cART with less than 200 cells/μl did not reach 600 cells/μl after 7 years of treatment. Those starting cART with 200-349 CD4 cells/μl could reach 600 cells/μl within less than 2 years of fully suppressive treatment. Probability of CD4 normalization (i.e. >800 cells/μl) after 7 years of suppressive treatment was 24 and 46% for those starting treatment with less than 200 or 200-349 CD4 cells/μl, respectively. Lower pre-cART cumulative viremia was associated with faster CD4 recovery. CD4 cell count increases after 4 years were either insignificant or very slow, irrespectively of baseline levels.

CONCLUSION

cART initiation before CD4 cell count drops below 350 cells/μl is crucial for achieving normal CD4 levels. These findings underline the importance of timely diagnosis and cART initiation as the risk of both AIDS and non-AIDS-related morbidity/mortality remains increased in patients with incomplete CD4 recovery.