Proportion of serious drug interactions dropping on Liverpool antiretroviral website
20th International Workshop on Clinical Pharmacology of HIV, Hepatitis & Other Antiviral Drugs. May 14-16, 2019. Noordwijk, the Netherlands


Mark Mascolini

Over the past 20 years, the proportion of potential serious drug-drug interactions involving first-line antiretroviral regimens dropped substantially, according to analysis of the esteemed University of Liverpool drug-interaction site [1]. The Liverpool team speculated that dwindling antiretroviral regimen pill burden over the years helped ease the interaction potential.


Two decades have passed since stalwart University of Liverpool researchers devised a simple online tool to help clinicians and others anticipate potentially harmful interactions between antiretrovirals and other drugs often prescribed for people with HIV infection [2].

The 1999 Liverpool site listed 7 antiretrovirals, compared with 36 antiretrovirals and/or combinations in 2019. The 142 comedications featured in 1999 jumped 5-fold to 728 comedications today. The 994 drug-drug interactions highlighted in 1999 are dwarfed by the 26,208 interactions today. At its inception the Liverpool site linked about 30% of interactions to further online information, whereas all of today's interactions have links to more data. The Liverpool team reported that today the site draws about 23,500 users monthly from about 220 countries.


This retrospective analysis compared current rates of antiretroviral interactions that require intervention with rates 20 years ago. The analysis focused on current and historic first-line regimens, not individual antiretrovirals, recommended by US, UK, or European guidelines. The Liverpool site ranks all potential antiretroviral-comedication interactions as red (contraindication/do not coadminister), amber (interaction requiring dose modification/monitoring),yellow (weak potential interaction), or green (no clinically significant interaction). Then the researchers split interactions into those requiring intervention (red and amber) and those requiring no intervention (yellow and green). The primary goal of the analysis was to compare current and historic rates of interventional interactions with antiretroviral regimens, that is, the percent of red or amber interactions among all interactions (red, amber, yellow, or green).


The Liverpool team counted 28 first-line regimens on the initial (historic) website and 16 first-line regimens today. Among the historic regimens assessed against the current comedication list, the proportion of interactions requiring intervention ranged from 37% to 57%. Among current regimens, the proportion of interventional interactions ranged from 8% to 54%. Looked at another way, the proportion of interventional interactions with the least-interacting antiretroviral regimens plunged from 37% with historic regimens to 8% today.

Among historic regimens recommended when the Liverpool site began, those involving nonnucleosides had a somewhat lower interventional interaction rate than those involving a protease inhibitor, 44% versus 52%. Among today's regimens, the interventional interaction rate is much greater with cobicistat- or ritonavir-boosted regimens than with unboosted regimens, 45% versus 12%. Compared with all other antiretroviral regimens, those including tenofovir alafenamide (TAF) had a slightly lower interventional interaction potential than those involving tenofovir disoproxil fumarate (TDF), 28% versus 34%.


Finally, the Liverpool investigators considered potential interactions with three clusters of comedications, those prescribed for mental health comorbidities (49 comedications), cardiovascular comorbidities (71 comedications), and metabolic comorbidities (36 comedications). Among current antiretroviral regimens, the interventional interaction potential proved much higher with boosted regimens than with unboosted regimens: 47% to 61% versus 0 to 4% for mental health comedications, 37% to 55% versus 1% to 10% for cardiovascular comedications, and 28% to 53% versus 0 to 14% for metabolic comedications.


The Liverpool team stressed the critical importance of "an awareness of the potential for drug-drug interactions with current first-line regimens, particularly when treating certain comorbidity clusters."



1. Gibbons S, Marzolini C, Moss C, McAllister K, Chiong J, Back D, Khoo S. Changes in drug interaction profiles for first-line HIV therapy over the 20 years of the Liverpool Drug Interaction website. 20th International Workshop on Clinical Pharmacology of HIV, Hepatitis & Other Antiviral Drugs. May 14-16, 2019. Noordwijk, the Netherlands. Abstract 5.

2. University of Liverpool. HIV Drug Interactions.