By 2024, there were more than 40 million people living with HIV, of whom over 31 million had access to treatment. Antiretroviral (ARV) treatments have significantly increased the life expectancy of people living with HIV (PWH), transforming HIV into a chronic condition requiring lifelong management (Marcus et al., 2020; Gueler et al., 2017; Samji et al., 2013). ARVs target different stages of the HIV lifecycle. NRTIs (nucleoside or nucleotide reverse-transcriptase inhibitors) and NNRTIs (Non-nucleoside reverse-transcriptase inhibitors) block reverse transcriptase’s enzymatic function and prevent completion of synthesis of the double-stranded viral DNA. INSTIs (Integrase Strand Transfer Inhibitors) prevent viral DNA integration into the host genome. PIs (protease inhibitors) inhibit viral protease, blocking virus maturation, and are often combined with boosters (ritonavir or cobicistat) that inhibit CYP3A4 to increase drug levels.

Modelling studies project a significant increase in the median age of PWH receiving ARV therapy, accompanied by a growing burden of comorbidities (Smit et al., 2015; Smit et al., 2017).

People with HIV (PWH) are at greater risk of developing comorbidities and experiencing polypharmacy (defined as taking five or more non-antiretroviral (ARV) medications concurrently) earlier than the general population, which leads to increased costs for managing comorbidities (Guaraldi et al., 2011; Yang et al., 2019; Gimeno-Gracia et al., 2016; Li et al., 2025; Hopwood-Raja et al., 2025). For example, a Spanish study revealed that polypharmacy is more prevalent among older HIV-positive individuals than among similarly aged members of the general population. This report found that HIV-positive individuals were prescribed more central nervous system (CNS) drugs and anti-infectives (Gimeno-Gracia et al., 2016). In a Canadian cohort of people aged 65 and older living with HIV, more than half of the participants had polypharmacy (Hopwood-Raja et al., 2025). According to 2023 epidemiological data, more than half (51%) of people living with HIV in Belgium were aged 50 years or over. The average duration since diagnosis for PWH in follow-up care increased from 7 years in 2006 to 14 years in 2024 (Deblonde et al., 2024). This ageing population faces heightened vulnerability to chronic diseases, which leads to increased medication use and higher rates of polypharmacy. This subsequently increases in the risk of drug‒drug interactions (DDIs) (Back and Marzolini, 2020; Mazzitelli et al., 2024). In elderly HIV-positive patients, the duration of HIV infection is a stronger predictor of multimorbidity and polypharmacy than age alone (Guaraldi et al., 2018).

DDIs occur through mechanisms such as the inhibition or induction of liver enzymes (e.g., CYP3A4), changes in drug absorption in the digestive tract (e.g., chelation or pH modifications), and altered renal drug excretion. They are particularly expected when the drugs share metabolic pathways (e.g., CYP450 enzymes).

DDIs in individuals receiving HIV treatment can alter medication levels, causing toxicity, reduced efficacy, or resistance to ARV medications (Back and Marzolini, 2020; Nhean et al., 2021).

Given the many options for HIV therapy, selecting a regimen for an individual should be guided by factors such as virologic efficacy, toxicity, pill burden, dosing frequency, and the potential for drug‒drug interactions. For example, PWH treated with ritonavir or cobicistat boosted protease or integrase inhibitors experienced clinically relevant interactions more frequently than those on other regimens (de Oliveira Costa et al., 2023).

Currently, regimens including an integrase strand transfer inhibitor (INSTI) are the most commonly prescribed. Unboosted INSTIs are often favoured for various reasons, including a lower risk of DDIs (Back and Marzolini, 2020; Deutschmann et al., 2021; Lepik et al., 2022; Peng et al., 2024). In addition, some dual ARV therapies achieve therapeutic efficacy comparable to that of triple therapy and are recommended as switch or initial treatments. Switching standard triple therapy to dual therapy for individuals receiving antiretroviral treatment could also help reduce the risk of DDIs, as dolutegravir and lamivudine were among the regimens with the lowest rates of relevant potential DDIs in a recent report (de Oliveira Costa et al., 2023).

Although PWH are ageing, the proportion of individuals with potential DDIs could thus decrease over time with a switch to newer therapies that could reduce the number of DDIs. In addition, an alert system was implemented following previous alarming results showing an important number of DDIs in our centre in 2012 and 2016 (El Moussaoui et al., 2020).

The present study aimed to analyse the evolution of the prevalence and types of contraindicated drug-drug interactions (DDIs), also known as ‘red-flag’ DDIs, among PWH between 2017 and 2022 and to compare our findings with those of a previous study conducted at the University Hospital of Liège (2012–2016). ‘Red-flag interactions’ refer to severe, potentially life-threatening drug interactions that are absolutely contraindicated or require immediate intervention to avoid serious adverse effects.

We conducted a retrospective longitudinal cohort study of individuals aged 18 years and over living with HIV and attended the University Hospital of Liège, Belgium, as outpatients, between 2017 and 2022. Participants who did not attend a medical consultation every year were excluded from the study, to enable annual data collection. All of the other participants were included in the study.

Demographic data included age, sex, ethnicity, and country of origin. Clinical data included weight, height, BMI, the date of the first positive HIV test result, the date of the first infectious disease consultation, the mode of transmission, and information on alcohol consumption status, smoking status, drug use, and concomitant medications. Biological data included the nadir CD4 count, the presence of other conditions (including hepatitis B or C), and HIV type (1 or 2). Data on non-ARV medications were collected at each visit to an infectious disease specialist and categorized according to the Belgian Center for Pharmacotherapeutic Information (CBIP) standards. The CBIP is an official drug evaluation organization that provides a standardized therapeutic classification system.

We used the Liverpool Interaction Database to identify contraindicated (red flag) drug interactions (Liverpool HIV Interactions, 2023). The charts used in this study were available on the Liverpool Interaction Checker and were last revised on 31 May 2023. This study focused exclusively on contraindicated drug-drug interactions (DDIs) between antiretroviral (ARV) and non-ARV medications.

To enable comparison with the years 2012 and 2016 (El Moussaoui et al., 2020), participants were selected who had been followed up in both 2012 and 2016, as well as between 2017 and 2022. Red-flag interactions for these participants were analysed using the same charts, which were available on the Liverpool Interaction Checker and revised on 31 May 2023.

This longitudinal retrospective study assessed the evolution and type of red-flag DDIs in a cohort of people living with HIV in Liège, Belgium, from 2017 to 2022. We also compared our findings with the results of previously published studies that analysed DDIs in 2012 and 2016 (El Moussaoui et al., 2020). Our results revealed a significant increase in the number of non-ARV comedications used, which has been identified as a risk factor for DDIs, as reported by other groups (Ok et al., 2020; Ruellan et al., 2021; Altunal et al., 2023). This is not surprising, given that the average age of PWH is increasing, both in our cohort and globally. The proportion of participants with polypharmacy increased significantly in our study, rising from 20.7% to 28.9% between 2017 and 2022. Polypharmacy is undoubtedly a growing issue that needs to be addressed (De Bellis et al., 2025), as polypharmacy has been linked to non-adherence, adverse drug events, falls, opioid overdoses and increased complexity of medical regimens, thereby contributing to inappropriate prescribing (Justice et al., 2021).

Despite the substantial increase in comedications used, we observed a continuous reduction in the total number of DDIs since 2017, which contrasts with the findings of our previous reports from 2012 to 2016 (El Moussaoui et al., 2020). A decline in red-flag and/or orange-flag DDIs has also been documented in other studies (Deutschmann et al., 2021; Lepik et al., 2022). These reductions were mostly attributed to a shift from PI-based regimens to INSTI-based regimens. The decrease in red-flag DDIs in our study was related to multiple factors.

Firstly, there has been a shift from higher DDI-risk ART, mostly boosted regimens, to those with lower risk, particularly regarding the widespread usage of unboosted INSTIs, both in triple and dual therapy. The large reduction in the use of boosted regimens is a key factor associated with the reduction in DDIs, as a booster is associated with a 30-fold greater risk of red-flag DDIs. We thus confirmed other results showing a strong negative impact of a booster on the risk of DDIs (de Oliveira Costa et al., 2023; Altunal et al., 2023; Jakeman et al., 2017; Peng et al., 2024). In particular, boosted elvitegravir was implicated in 37% of DDIs in 2017 and 26% of DDIs in 2022 in our study, a score largely driven by its red-flag interaction with topical steroids and lercanidipine. Bictegravir- and dolutegravir-based regimens are now preferred because of their higher resistance barrier and better profile considering DDIs.

Second, the simplification of antiretroviral therapy regimens may contribute to the observed trend, with a growing usage of dual therapy and a shift away from more complex treatment combinations (Ruellan et al., 2021). We confirmed that the combination of NRTI/INSTI (3TC/DTG), which has been widely used in our centre and globally in recent years, is safe in terms of the risk of DDIs. Nevertheless, NNRTI/INSTI (RPV/DTG) dual therapy was associated with a higher risk of DDI, particularly when associated with proton pump inhibitors (Back and Marzolini, 2020; Hodge et al., 1999; Livio and Marzolini, 2019).

Finally, following our previous reports from 2012 to 2016, we implemented an alert system to warn clinicians that concurrent prescription may be inappropriate. This alert system is based on the University of Liverpool HIV drug interaction database. We regularly run this program that allows us to identify contraindicated interactions between ARVs and other medications listed in the electronic medical records. Clinicians are then notified in the event of contraindicated interactions. This system has probably contributed to changing clinicians’ habits and reducing the number of contraindicated interactions. However, this effect is difficult to isolate from other factors involved in reducing these interactions.

One of the identified risk factors for DDIs in our study was the duration of HIV infection rather than age itself. This may reflect an earlier onset of multimorbidity and subsequent polypharmacy in PWH, a phenomenon well known as premature ageing (Guaraldi et al., 2011; Gimeno-Gracia et al., 2016; Guaraldi et al., 2018).

In terms of comedications, topical corticosteroids emerged as the most frequently implicated non-ARV drugs in DDIs, accounting for almost half of the red-flag DDIs in 2022. This finding is consistent with previous studies, which have identified corticosteroids as one of the non-ARV drugs most frequently implicated in DDIs (de Oliveira Costa et al., 2023; López-Centeno et al., 2020). These medications mostly interact with boosted ARV medications, which inhibit CYP3A4, thereby increasing the risk of Cushing’s syndrome and adrenal insufficiency. Although the systemic absorption of topical corticosteroids is usually low, it can sometimes be significant, particularly when high-potency formulations are used or applied to large areas of skin. In such cases, the risk of interaction with pharmacokinetic boosters may become relevant. This risk could be mitigated by the prescription of alternative corticosteroids, such as beclomethasone (Liverpool HIV Interactions, 2023; EACS, 2023; World Health Orga nization, 2021). If coadministered, dose reduction of the glucocorticoid should be considered with close monitoring of local and systemic effects (Liverpool HIV Interactions, 2023).

The proportion of DDIs involving proton pump inhibitors (PPIs) and RPV remained high and fairly stable over time, accounting for almost a quarter of the red-flag DDIs in 2022. PPIs are frequently well-known causes of DDIs, as shown in other studies (de Oliveira Costa et al., 2023; Deutschmann et al., 2021; Steulet et al., 2024). Coadministration may significantly decrease the plasma concentration of rilpivirine. Indeed, PPIs reduce gastric acid secretion and thereby increase gastric pH, which can substantially impair the absorption of certain antiretroviral drugs that require an acidic environment for optimal bioavailability—such as rilpivirine and atazanavir. In the case of rilpivirine, coadministration with PPIs can lead to significantly reduced plasma concentrations, potentially compromising virologic suppression and increasing the risk of treatment failure. Alternatives include switching to intramuscular RPV or to another ARV unaffected by gastric pH (Liverpool HIV Interactions, 2023; EACS, 2023).

Cardiovascular and anti-haemostatic drugs each accounted for 17% of the DDIs in our study, a finding frequently reported in the literature (de Oliveira Costa et al., 2023; Lepik et al., 2022; López-Centeno et al., 2020). The molecules lercanidipine, clopidogrel and rivaroxaban were among the most frequently involved. PIs and boosters significantly increase the risk of adverse reactions by increasing the plasma concentrations of some of these drugs, including some statins, potentially leading to serious outcomes such as rhabdomyolysis. They also increased the risk of bleeding when taken alongside anticoagulants or clopidogrel. The clinical effect of DDIs between antiplatelets and antiretroviral therapy (ART) on bleeding, thrombosis, and other major adverse cardiovascular events (MACE) remains unclear (Matsikas et al., 2025). Regarding antihypertensive drugs, boosters further amplified plasma lercanidipine levels. Alternative antihypertensives should be considered to prevent these potentially severe interactions (Back and Marzolini, 2020; Nhean et al., 2021; Liverpool HIV Interactions, 2023; EACS, 2023).

PWH are particularly vulnerable to mental health conditions, which often necessitate the use of neuroleptics. According to the literature, neuroleptics are commonly prescribed to PWH (Deutschmann et al., 2021). We indeed identified specific DDIs involving haloperidol and quetiapine. In particular, quetiapine is known to be involved in several DDIs and, together with other atypical antipsychotics, is listed by the EACS guidelines as one of the top medications to be avoided and not prescribed to elderly individuals with HIV in certain conditions (Liverpool HIV Interactions, 2023; EACS, 2023).

Of course, the issue of drug-drug interactions (DDIs) is not limited to people living with HIV (PWH). Antiviral drugs are also used in other contexts, such as for treating hepatitis B and C, and for treating patients with SARS-CoV-2 (Hodge et al., 1999; Livio and Marzolini, 2019). In the context of SARS-CoV-2 infection, the most frequent DDIs involve nirmatrelvir/ritonavir (Hodge et al., 1999), which is consistent with our results. Therefore, SARS-CoV-2 infection in PWH is associated with an increased risk of DDIs, necessitating vigilance and a personalised therapeutic approach (Conti et al., 2024; Hendrick et al., 2024).

Our findings should be interpreted in light of several limitations. Patients often fail to disclose all over-the-counter medications or dietary supplements used, and short-term treatments such as antibiotics are difficult to track accurately, potentially introducing bias into the study. Additionally, we did not analyse interactions between ARV medications themselves.

In conclusion, we showed a significant and continuous increase in the number of non-ARV comedications used in our cohort. The risk of DDIs remains closely linked to the number of non-ARV comedications used, although the number of red-flag DDIs decreased with time, a reduction mostly triggered by the switch to an unboosted ARV regimen. Topical steroids, when associated with a booster, and PPIs, when associated with RPV, are the most frequent comedications associated with DDIs. Our findings emphasize the importance of ongoing vigilance in identifying and managing DDIs in PWH. Active monitoring, along with the implementation of alert systems, can help healthcare providers avoid potentially harmful interactions. Favoring ARVs with a low risk of interaction could help reduce the risk of interaction. In addition, it is crucial to raise awareness among healthcare providers, including not only infectious disease specialists, but also general practitioners. Patients themselves need to be made aware of the risk of drug interactions, and how to ensure that they do not occur when a new drug is prescribed or taken without any prescription. Distribution points also have a role to play in preventing drug interactions, especially where prescription-free medication access exists.