Assistant Director, Epidemiology Regeneron Pharmaceuticals Tarrytown, United States
Background: Antiretroviral therapy (ART) has improved the morbidity and mortality of patients living with Human Immunodeficiency Virus (HIV). Older ARTs such as stavudine and zidovudine, however, are known to cause adverse events such as lipodystrophy (LD), with a reported adverse event rate of up to 85%. With the advent of newer HIV therapies, use of older ARTs has decreased dramatically and consequently, the prevalence of HIV-LD has purportedly decreased. However, there are no contemporaneous data on the use of older ARTs nor prevalence estimates of HIV-LD in the US.
Objectives: To estimate trends in the annual prevalence of HIV-LD and the proportion of patients with HIV who were dispensed stavudine or zidovudine in the US from 2006 to 2018.
Methods: The annual prevalence per 100,000 persons of HIV-LD and the proportion of HIV patients with a dispensation of stavudine or zidovudine were estimated using the following databases (years): IQVIA® Pharmetrics Plus (2006-2018), IBM® MarketScan Commercial/Medicare (2013-2018), and IBM® Multi-state MarketScan Medicaid (2013-2018). Patients continuously enrolled in their healthcare plan in a given calendar year were included and consisted of the prevalence denominator. To identify prevalent HIV-LD cases, we required ≥1 HIV/AIDS diagnosis (ICD-9-CM: 042.x or ICD-10-CM: B20) and ≥ 1 LD diagnosis (ICD-9-CM: 272.6 or ICD-10-CM: E88.1) in the same year. To assess use of older ART among HIV patients, we identified patients with ≥ 1 dispensation for stavudine or zidovudine who also had ≥1 HIV/AIDS diagnosis in a given calendar year.
Results: The prevalence of HIV-LD in Pharmetrics Plus was relatively stable between 2006-2018, ranging from 2.4-1.9 per 100,000. A similar trend was observed in the MarketScan Commercial/Medicare database where the prevalence ranged from 3.1-2.5 per 100,000 during 2013-2018. In contrast, the prevalence of HIV-LD steadily declined in the MarketScan Medicaid database from 1.9 to 1.0 per 100,000 individuals during 2013-2018. The use of medications known to cause HIV-LD decreased substantially in all databases. Using Pharmetrics Plus, we observed a decrease of 94% from 25.9% to 1.6%. Similar decreases were also seen in MarketScan Medicaid and Commercial/Medicare where dispensations for these medications fell by 74.3% and 76.5% from 10.9% to 2.8% and 6.8% to 1.6%, respectively.
Conclusions: The prevalence of HIV-LD has remained low over the last decade with < 4 cases per 100,000. Use of older ART has decreased substantially during the years studied, with fewer than 5% of HIV patients in the US using stavudine and zidovudine in 2018. However, the impact on HIV-LD prevalence was minimal, except for Medicaid patients, where a 50% reduction of HIV-LD was recognized.
