Users of TAF in PrEP more likely to develop high blood pressure or need statins

A study of people who started PrEP in a southern California health system has found that people who took PrEP containing tenofovir alafenamide (TAF) were more likely to be diagnosed with high blood pressure or start treatment with a statin to control cholesterol than those taking PrEP containing tenofovir disoproxil (TDF) during a two-year follow-up.

The findings, published this week in the journal JAMA Open Network, come from a retrospective study, and need to be confirmed in a larger prospective study.

Use of TAF in PrEP: international differences

The combination of tenofovir alafenamide and emtricitabine (Descovy) is available for PrEP use in a handful of countries as an alternative to tenofovir disoproxil/emtricitabine (TDF/FTC), now widely available as a generic product.

Gilead Sciences announced in October 2021 that it would not seek a license for use of Descovy as PrEP in the European Union. However, TAF has been approved for PrEP use in the four nations of the United Kingdom by the Medicines and Health Regulatory Agency. NHS England said in June 2023 that it would pay for the use of TAF as ‘second-line’ PrEP for people with kidney impairment, osteoporosis or at risk of bone fracture, and for under-18s. Use of Descovy for PrEP remains rare in the United Kingdom. A recent survey of a large Scottish PrEP clinic found that only 0.5% of PrEP users would qualify for Descovy because of bone or kidney problems.

But in the United States, Descovy is widely prescribed as PrEP. A survey of PrEP prescriptions in 2020 found that over 75,000 people were taking Descovy as PrEP during the second quarter of 2020 and that older people were more likely to have switched from TDF/FTC to Descovy.

Potential metabolic effects of TAF when used in PrEP

TAF use in HIV treatment has been associated with weight gain and increases in total and LDL cholesterol.

The DISCOVER trial compared the use of Descovy and Truvada (TDF/FTC) as PrEP and led to the licensing of TAF for PrEP use in the United States in 2019 for cisgender men and transgender women. It did not report on changes in blood pressure or the incidence of new hypertension, but participants who took Descovy in the study gained a greater amount of weight than participants who took TDF/FTC. They also experienced significantly greater increases in triglycerides, total cholesterol, LDL and HDL cholesterol than recipients of TDF/FTC, although changes in the total cholesterol: HDL ratio were not statistically significant, and all changes were modest in extent.

The only previous retrospective study of the metabolic effects of Descovy PrEP carried out since its approval looked at 4978 Descovy PrEP users in the United States, matched by demographics, comorbidities and medications with users of TDF/FTC. The study found that Descovy users were more likely to start statin treatment, but TDF users were more likely to have blood pressure above 140mmm/Hg, the threshold for high blood pressure. However, there was no difference between the two groups in the rate of formal diagnosis of hypertension. 

The study

To investigate further the metabolic outcomes of the two PrEP pills, researchers at Kaiser Permanente Southern California used electronic health records to carry out a retrospective study to identify new cases of high blood pressure within two years of starting PrEP in people who began PrEP between October 2019 and May 2022. They also looked for cases of initiation of statins within two years of starting PrEP, as a proxy for potentially harmful increases in cholesterol levels.

They identified two cohorts for the purposes of the study: one of people without hypertension at the time of their first PrEP prescription and one of people who had not begun statin treatment at the time of their first PrEP prescription. Hypertension was defined as two or more blood pressure readings above 140mm/Hg (systolic) or 90mm/Hg (diastolic) within a two-year period.

The two cohorts were each 97% male, 34% White non-Hispanic, 6% Black non-Hispanic and 41% Hispanic. The mean body mass index was 26.7 in the hypertension cohort and 27.4 in the statin initiation cohort.

Each participant taking TAF was matched with four people taking TDF/FTC by propensity score. The propensity score was calculated using a model that adjusted for demographic factors, insurance type, medical centre, year, body mass index, lipid levels, dyslipidaemia, diabetes and cardiovascular risk score. Then researchers calculated the difference in risk between the two groups and the time to either diagnoses of hypertension or initiation of statins.

In the hypertension cohort prior to matching (5523 people, 6% taking TAF), those taking TAF were older (36 vs. 33 years), more likely to have diabetes at baseline (4% vs 1%) and more likely to be non-Hispanic. They had higher cardiovascular risk scores and had a shorter follow-up time (276 vs 321 days).

The matched hypertension analysis included 371 people taking TAF and 1484 people taking TDF. During the two-year follow-up period, 2.2% taking TAF and 1.3% taking TDF developed hypertension. The absolute difference in the proportion who developed hypertension was 0.8% but it was higher in people who started PrEP aged 40 or over (2.4%). People taking TAF had a 64% higher risk of being diagnosed with hypertension (odds ratio 1.64, 95% confidence interval 1.05-2.56) and the risk was somewhat higher in the over-40s (OR 2.0). But there was no significant difference in the time it took for hypertension to be diagnosed between people taking TAF or TDF. An additional analysis which used a wider definition of hypertension (130 mm/Hg systolic or 80mm/Hg diastolic) also found a higher rate of hypertension in TAF users (10.8% vs 5.5%).

In the statin cohort prior to matching (6149 people, 6% taking TAF), those taking TAF were more likely to be non-Hispanic White, were older (36 vs 33 years), had higher cardiovascular risk scores and shorter follow-up (290 vs 324 days). The matched cohort consisted of 1855 people, 20% taking TAF. TAF users more often initiated statins during the two-year follow-up period (1.6% vs 1%) and had more than twice the odds of starting statin treatment (OR 2.33, 95% CI 1.41-3.85). Again, the absolute difference in the risk of starting statins was higher for the over-40s (2.4% risk difference between TAF and TDF users vs 0.85% for the entire cohort).

The study authors acknowledge that because this is a retrospective observational study, the results may be influenced by people at higher risk of hypertension being more likely to take TAF. But they say that the matching process should guard against this confounding factor.

The study is also limited by missing data relating to key variables. Over 60% of participants had missing cardiovascular risk scores, approximately 45% had missing total cholesterol data and around 20% lacked weight measurements. The researchers used a statistical technique called multiple imputation to correct for the missing data.

The researchers say that more studies with larger sample sizes, longer follow-up  and more diverse populations are needed to investigate the cardiometabolic impact of TAF in people taking it as PrEP.

The US Centers for Disease Control and Prevention recommend that people taking Descovy as PrEP should have their triglyceride and cholesterol levels checked each year, but make no recommendations on blood pressure monitoring.