The effect of multivitamins and polyvalent cations on virologic suppression with integrase strand transfer inhibitors

Abstract

Integrase strand transfer inhibitors (INSTI) are the preferred drug anchor as part of antiretroviral therapy (ART) in people living with HIV/AIDS (PLWHA) [1]. Since 2014, clinicians have been increasingly advised of drug--drug interactions with INSTI and polyvalent cations (PVC), such as calcium, iron and zinc [1–4]. The mechanism of the drug--drug interaction is chelation, demonstrated by in-vitro studies [5,6], and recommendations for appropriate separation and/or administration vary by the individual INSTI and PVC involved [1–4]. PVC are also included in many multivitamins (MVI) and metal-based antacid preparations, where both raltegravir and dolutegravir concentrations were decreased by concomitant administration of antacids but not omeprazole [7,8]. However, the clinical significance of these interactions with INSTI is unclear with only a singular case report of virologic failure because of concomitant administration [9]. The objective of our study was to evaluate the impact of PVC on virologic suppression amongst PLWHA on INSTI-based ART. A retrospective electronic medical record (EMR) cohort of adult patients from an urban HIV clinic in Wilmington, Delaware was reviewed from July 2012 to September 2017 to identify all patients prescribed an ART regimen containing an INSTI (including dolutegravir, elvitegravir or raltegravir) for at least 6 months. The duration of 6 months was defined from the first resulted HIV quantitative viral load after initiating an INSTI-containing regimen. Medication lists were queried during the total time on an INSTI for concomitant administration of MVI and/or PVC. Individual patient EMRs were then evaluated for episodes of virologic failure, defined as an HIV viral load greater than 200 copies/ml, following initiation of the INSTI regimen. Follow-up time was calculated for all patients until virologic failure, study conclusion, or treatment discontinuation. Patients with treatment nonadherence and/or treatment interruptions, as documented by pharmacy refill or clinical history, were excluded from analysis. Analysis utilized descriptive statistics and Poisson regression with log of follow-up time specified as an offset to calculate the rate ratio of virologic failure given documented PVC. Models controlled for age in years, sex and race. Characteristics of the cohort appear in Table 1. Four hundred and ninety-three patients were prescribed an INSTI between July 2012 and September 2017; 360/493 (73%) were receiving an INSTI for at least 6 months during the study period. Among those prescribed an INSTI, elvitegravir, dolutegravir and raltegravir were received by 223 patients (62%), 115 patients (32%) and 22 patients (6%), respectively. Median follow-up time was 529 days (interquartile range: 382 days).Table 1: Cohort characteristics.One hundred and fifty-two patients (42%) on an INSTI were documented as receiving a PVC. Ninety-nine patients (65%) reported an MVI, with 18 of those receiving at least one additional source of PVCs. Seventy-three patients (48%) were in receipt of a PVC alone, with the majority of patients receiving calcium or iron supplementation. Forty-six patients (13%) experienced virologic failure during the study period. Virologic failure rates were 15%, 11% and 18% for patients receiving dolutegravir, elvitegravir and raltegravir, respectively. Patients receiving any PVC had on average 2.3 times the risk of virologic failure [95% confidence interval (CI) 1.2--4.4] compared with patients not reporting a PVC, controlling for age, race and sex. This association persisted irregardless of the specific PVC or MVI. Among the covariates included in the adjusted model, only male sex was independently associated with virologic failure (rate ratio: 4.1; 95% CI 1.8--10.8). We observed that concomitant receipt of PVC and INSTI-based ART was associated with more than double the rate of virologic failure in our HIV clinic. Although recommendations for appropriate separation were evolving over our study period, our data reinforce the importance of ongoing medication reconciliation and patient education. Patients may not equate vitamins, metal-based antacids and other supplements as ‘medications’ and may neglect to mention administration of these substances unless explicitly queried as part of a thorough medication history. Ongoing education about drug--drug interactions is vital for all patients, regardless of whether they have been taking their existing ART regimen for years or are either initiating or changing to an INSTI-based regimen. The analyzed data was limited to information retrospectively collected and documented in the EMR. This analysis did neither capture the separation (or lack thereof) of INSTIs from PVCs, nor did it assess whether education regarding drug--drug interactions was delivered to patients. Although a larger, prospective trial with the ability to capture exact timing of medications is warranted, our study demonstrates that these drug--drug interactions extend beyond theoretical concerns and have a clinically significant impact on virologic suppression. Acknowledgements Conflicts of interest There are no conflicts of interest.