Introduction

Antiretroviral drugs (ARVs) have significantly decreased AIDS-related mortality and morbidity [1, 2]. In high-income countries, the life expectancy of people living with HIV/AIDS now approaches that of the HIV-negative population [3, 4]. While current treatment guidelines recommend daily ARV use, many patients experience treatment fatigue after years of adherence. Moreover, less than two thirds of patients maintain the ≥90% adherence associated with optimal viral suppression [5]. Similarly, pre-exposure prophylaxis (PrEP) has demonstrated high efficacy in HIV prevention trials and is recommended by the World Health Organization for populations with HIV incidence of ≥3% [6]. However, adherence to PrEP ranged from 28 to 98% in clinical trials [7]. Adherence can be an even greater challenge when there is co-existing mental illness, active substance abuse, unstable housing or limited social support, and when the regimen is complex such as multiple pills and/or doses per day [8,9,10,11,12,13].

Long-acting or extended release ARVs (LA-ARVs) promise new options for HIV treatment and prevention, and enable novel strategies to address poor adherence or treatment fatigue. A consensus definition including optimal dosing frequency of LA-ARVs has not been determined; however, these agents have an extended half-life due to engineered drug release kinetics or direct properties of the drug or its bioactive metabolite(s). In this paper, we review the leading LA-ARVs in development and summarize clinical studies of their use for HIV treatment and prevention. We conclude with a discussion of the possible challenges to the use of these compounds in routine care.

Long-Acting ARV Formulations

Nanoformulation

Long-acting drugs have been available for more than 30 years as antipsychotics, contraceptives, and biologics, but available ARV formulations were either not soluble in suspensions or required high drug volumes to achieve efficacy. Advances in nanotechnology have enabled some ARVs to be milled into nanometer-sized particles and crystals that are stable and highly bioavailable at low drug concentrations in suspension [14]. However, only a select few of the current ARVs have compatible aqueous solubility, antiviral potency, and systemic clearance kinetics for nanoformulation, and have advanced to clinical studies.

Long-Acting Cabotegravir

Cabotegravir (CAB) is an integrase strand transfer inhibitor (INSTI), structurally similar to dolutegravir. As a crystalline free form, it demonstrates low aqueous solubility and a half-life of 40 h after oral administration [15]. As an injectable, depot-controlled nanosuspension, drug particles slowly dissolute into interstitial fluid surrounding the drug depot [16]. Consequently, the injectable depot formulation of long-acting CAB (CAB LA) demonstrates an approximately 25-fold higher elimination half-life (25–54 days) compared to the oral formulation [17]. Measurable plasma levels occur within hours of injection, and therapeutic levels within days. In one of the initial studies, plasma levels exceeded the protein adjusted IC90 (PA-IC90) for approximately 16 weeks following a single 800-mg intramuscular (IM) injection in healthy subjects and measurable levels persisted for up to 52 weeks in some individuals [17]. Drug absorption of the depot formulation varies by sex and body mass index (BMI). Slower rates of absorption have been observed with increasing BMI and female sex (−35%) [18]. Conversely, faster rates of absorption have been observed in very low BMI [18]. Overall exposures to CAB, however, appear similar in females and males, suggesting that extent of absorption are comparable despite different absorption rates. CAB’s metabolism occurs mainly via biliary excretion, with the majority excreted as unchanged drug; enterohepatic circulation of reabsorbed drug may contribute to its prolonged pharmacologic profile [19]. It has not been shown to inhibit or induce major cytochromes or other hepatic, intestinal, or renal drug transporters, making it an ideal candidate for co-administration with other ARVs [15].

Long-Acting Rilpivirine

Rilpivirine (RPV) is a non-nucleoside reverse transcriptase inhibitor (NNRTI) that demonstrates potent activity against both wild-type HIV-1 and some mutants resistant to first-generation NNRTIs [20]. RPV is much better tolerated than earlier NNRTIs, with the most common adverse effects (AEs) being nausea, headache, disrupted sleep, somnolence and rash [21]. While daily oral formulations have been available for several years, a long-acting injectable rilpivirine formulation (RVP LA) has recently been developed as a crystalline nanosuspension [22]. The drug is detectable in serum within hours of IM dosing, and peak serum concentrations are reached within 6–8 days [23•]. Its plasma elimination half-life is 44–61 days and sub-therapeutic concentrations have been detected in plasma and female genital fluids over 18 months after a single IM injection [24, 25]. Peak plasma concentrations of RPV were approximately 30% lower among females and decreased by approximately 2.3% for every kilogram per square meter increase in BMI; however, only gender was associated with overall RPV exposure, and no effect of gender or BMI was noted on plasma concentrations after 28 days post-dose [23•]. Females with BMI >25kg/m2 appear to be at risk of low peak concentrations after injection.

Drug Loaded Inserts or Implants

Dapivirine Vaginal Ring

Vaginal rings were first developed in the 1970s for delivery of hormone replacement therapy and are most commonly used for contraception [26]. Initial attempts to use a vaginal ring with an HIV microbicide failed as nonoxynl-9 was found to increase the risk of HIV transmission with repeated application. Now, flexible silicone rings with ARVs stabilized in matrices or reservoirs that allow slow, local drug diffusion have been developed. The most advanced of these incorporates dapivirine, an NNRTI. Dapivirine vaginal ring is a silicone elastomer matrix developed primarily for PrEP. There is no pill formulation. In healthy female volunteers, 25-mg dapivirine vaginal ring resulted in maximum plasma drug within a median of 7 days [27]. Dapivirine concentrations in the vaginal fluid were at least 2800-fold higher than the IC99, as estimated from a cervical tissue explant model, near the ring, cervix, and intriotus within 1.5 h of ring insertion. This level was sustained or higher until the ring was removed after 28 days. After ring removal, the plasma drug elimination half-life was 67 h. The ring was well tolerated; the most common AEs were metrorrhagia, vulvovaginal discomfort, and headache. There was no evidence of significant alterations to the vaginal fluid pH or flora.

Tenofovir Alafenamide Implant

While oral and vaginal formulations of tenofovir have shown promise for PrEP, nanoformulation is not an option for tenofovir alafenamide (TAF) due in part to its high aqueous solubility. The drug is however amenable to sustained release formulation as a subdermal implant that employs the geometry of widely used contraceptive implants. In beagle dogs, a prototype subdermal TAF implant produced intracellular levels of the active moiety (tenofovir diphosphate) that were 11–32 higher than protective levels (24–48 fmol/106 cells) estimated from the iPrEX study [28]. The geometry of the device allows easy insertion and replacement on the inside of the upper arm. It is estimated that the implant would need to contain 51 mg of TAF for a 1-year dose. Further studies of this prototype and other implants are ongoing.

Microparticles

Drug formulation as microparticles may be an alternative approach for some drugs as has been demonstrated with nevirapine, an NNRTI. A subcutaneously administered LA formulation of nevirapine consisting of microparticles (>50 μm) coated with biocompatible polymers showed promising results in a rat model [29]. Human simulation showed that an optimally formulated and dosed injection may produce plasma levels for 90 days that exceed 0.2 g/ml, which is five times the plasma levels thought to be needed for infant prophylaxis.

Targeted Mutagenesis

Broadly neutralizing monoclonal antibodies (bNAbs) have generated interest for HIV prevention and treatment. One of the first of these compounds is VRC01 that targets the CD4 binding site of HIV-1 and requires intravenous administration every few weeks [30]. The potential use of VRC01 for treatment/prevention was increased by development of VRC01 LS, which has an extended half-life. VRC01 LS incorporates site-directed mutagenesis where methionine in the C-terminus of the heavy chain was changed to leucine (L) and asparagine to serine (S). These mutations markedly increase the bNAb’s binding affinity for the neonatal Fc receptor, consequently reducing systemic clearance [31]. The pharmacokinetic and antiviral properties of VRC01LS are being investigated in VRC 606 and VRC 607, respectively (NCT02599896, NCT02840474). Other bNAbs will likely undergo similar engineering to acquire long-acting properties.

Other Compounds

Other compounds with long-acting properties are shown in Table 1. Long-acting formulations of ritonavir boosted atazanavir and raltegravir are being investigated with the aim to achieve high drug concentrations in lymphoid tissues and the mucosal sites of HIV transmission. A novel nucleoside reverse transcriptase translocation inhibitor and a small molecule fusion inhibitor have demonstrated promising pre-clinical and early clinical data. Additionally, a monoclonal antibody that blocks CCR5 binding is moving into phase II/III trials for heavily treatment experienced HIV-infected patients and for HIV prevention.

Table 1 Other selected long-acting antiretrovirals
Full size table

Long-Acting ARV Strategies for HIV Treatment

The efficacy of CAB plus RPV two-drug regimen for maintenance of viral suppression was first demonstrated in LATTE 1, a phase II, dose-ranging trial that compared oral CAB plus RPV maintenance (after a three-drug induction period) to an efavirenz-containing three-drug regimen [38••]. LATTE 1 provided proof of principle for LATTE 2, an ongoing, phase IIb, multicenter, open-label trial comparing IM CAB LA and RPV LA administered every 4 weeks (Q4W) or 8 weeks (Q8W) to an oral three-drug regimen of CAB plus abacavir/lamivudine ABC/3TC after a 20-week induction period of oral CAB plus ABC/3TC [39]. At 48 weeks, all arms demonstrated comparable antiviral activity (virologic suppression to <50 copies/ml in intent-to-treat maintenance exposed analysis was 91, 92, and 89% for Q4W, Q8W, and daily oral dosing, respectively) [40••]. Protocol defined virologic non-response occurred in <1, 7, and 2% participants receiving Q4W, Q8W, and daily oral dosing, respectively. The most common AEs in the LA maintenance arms were injection site reactions (ISR), comprising pain (67%), nodules (7%), and swelling (6%). The majority were of mild (82%) or moderate (17%) severity and 90% resolved within 7 days. Reports of ISR declined from 86% on day 1 to 29% at week 48. Only 2/230 participants withdrew from the study because of an ISR and about 90% reported they would be very satisfied to continue the injections. While all study arms are being followed through week 96, the every 4-week CAB LA plus RPV LA dosing has been selected for phase III evaluation in ATLAS (NCT02951052) and FLAIR (NCT02938520).

Long-Acting ARV Strategies for HIV Prevention

Cabotegravir LA

CAB concentrations at HIV transmission sites achieved with CAB LA are low due in part to its high protein binding; median cervical, and vaginal tissue to plasma ratio ranged from 16 to 28% while median rectal tissue to plasma ratios were ≤8% [41]. However, despite these low genital tissue concentrations, CAB LA prevented simian/human immunodeficiency virus infection (SHIV) via the vaginal, rectal, and intravenous routes in macaques [42•, 43, 44•]. ÉCLAIR is an ongoing phase IIa randomized, multicenter, double-blind study of CAB LA in 127 adult men at low risk for HIV infection [45]. Participants were randomized 5:1 to oral CAB (n = 106) 30 mg daily for 4 weeks then CAB LA 800-mg IM every 12 weeks for three doses versus saline placebo (n = 21). The primary objective was to evaluate safety and tolerability of CAB LA through week 41. Participants are to be followed through week 81. The median age of participants at entry was 31 and 30 years in CAB and placebo arms, respectively. Half of participants were white and greater than three-quarters reported sexual contact with a man as their risk factor for HIV acquisition. Two participants acquired HIV by week 41, one in the placebo arm and the other in the CAB LA arm 24 weeks after the last injection. There were no serious AEs attributable to CAB LA; however, 93% of CAB LA recipients reported an ISR compared to 57% in placebo arm. The most common ISR was pain, reported in 92% of CAB LA injections and 27% of placebo injections, none of which was grade 4. Pain in the CAB LA arm lasted about 5 days, and 4/94 (4%) dropped out of the CAB LA arm due to injection intolerance. The target drug concentration was ≥4 × PA-IC90 as this concentration demonstrated 100% protection against SHIV infection via rectal challenge in a primate model [44•]. Thirty-one percent, 37%, and 30% of participants achieved this target drug concentration after the first, second, and third injection, respectively. Going forward, 600-mg injections of CAB every 8 weeks will be studied in HPTN 083 (NCT02720094).

Rilpivirine LA

In the first study to evaluate RVP concentrations in human genital compartment following RPV LA injection, single 300, 600, and 1200 mg IM doses of RPV LA were administered to 60 women and the 600-mg dose to 6 men [23]. Drug concentrations achieved in the cervicovaginal fluid and vaginal tissue approximated plasma, while rectal tissue concentration was similar or higher than plasma levels, but rectal fluid concentrations ranged from 9 to 33% of the plasma concentration. Three women who received the 300-mg dose had no detectable RPV in the cervicovaginal fluid and one female on the same dose acquired HIV at day 41 of the study. HIV drug resistance testing demonstrated that the patient was infected with wild-type virus then acquired the K101E mutation, leading to 4-fold resistance to RPV and 4–8-fold cross-resistance to other NNRTIs [46••]. RPV LA was well tolerated and there were no serious AEs. The most common AEs were self-limited injection site pain and/or transient non-tender nodules.

In the MWRI-01 study, 36 participants (68% female) were enrolled into two cohorts of either a single 600 or 1200 mg IM dose of RPV LA, and followed for 4 months [47••]. At 28 days post-dose, the 1200-mg dose resulted in a higher geometric mean plasma concentration, consistent with the earlier study. The drug concentrations achieved in rectal tissues were about twice the level in cervical and vaginal tissues. Viral inhibition studies in tissue explants conducted as a surrogate of efficacy at the sites of HIV transmission demonstrated suppression of viral replication in rectal tissue but not in vaginal or cervical tissue. The pharmacokinetic and pharmacodynamic characteristics of multiple injections of RPV LA are being evaluated (NCT01656018). Studies to date portend an uncertain future for RPV LA as a PrEP agent. Nevertheless, a phase IIB, multisite, double-blind, placebo-controlled trial of RPV LA IM every 8 weeks for 40 weeks has completed enrollment and is ongoing (NCT02165202).

Dapivirine

Two large phase III randomized, double-blind, placebo-controlled trials of the dapivirine vaginal ring have been conducted. The Microbicide Trials Network-020—A Study to Prevent Infection with a Ring for Extended Use (MTN-020-ASPIRE) enrolled 2629 women and randomized (1:1) them to monthly 25-mg dapivirine vaginal ring or placebo ring for a median follow-up of 1.6 years [48••]. The median age of participants was 26 years, 41% were married, and 99.5% reported having a primary sexual partner. The dapivirine arm had 27% protection from HIV acquisition. Efficacy of protection differed significantly according to age: 61% for women >25 years versus 10% in younger women. The intervention’s efficacy was closely correlated to treatment adherence. Women older than 21 years of age, among whom efficacy of protection was 56%, demonstrated greater than 70% adherence based on plasma drug concentrations and residual amounts of dapivirine in the used vaginal rings. In post hoc analysis, adherence was categorized into four quartiles using a time varying model that considered the amount of time a woman had the vaginal ring, among other variables. This concluded that protection could have been as high as 92% in the most adherent women [49]. The intervention and placebo arms had similar rates of NNRTI mutations: 12 versus 10%, respectively. The drug was well tolerated and 14% of participants in either arm experienced any serious adverse event.

The second trial of dapivirine, RING study/International Partnership for Microbicides–027 (IPM-027), was also a phase III randomized, double-blind, placebo-controlled trial in African women [50]. One thousand nine hundred fifty-nine healthy, non-pregnant women between the ages of 18 and 45 years were enrolled for a follow-up of 24 months. Participants were randomized (2:1) to monthly 25-mg dapivirine vaginal ring or a placebo ring. The median age of participants was 25 years and 8% were married. Dapivirine ring resulted in a 30.7% reduction in the risk of HIV infection, 37.5% in women older than 21 years. Similar to ASPIRE, the drug was well tolerated and there was no significant difference in the incidence of adverse events between arms. Ongoing or planned studies will further evaluate the effectiveness of dapivirine ring, the impact of adherence, and whether there are biological explanations for lower protection rates in younger women (NCT02858037, NCT02702895, NCT02028338).

Acceptability of Long-Acting Agents: Patients’ Perspectives

Treatment

The attitudes toward LA-ARV for HIV treatment appear to be favorable among HIV-infected patients, while some individuals clearly prefer oral therapy. A survey of 400 HIV-infected individuals revealed that 72% were willing to try the injectable method. Willingness positively correlated with dosing interval, as 61% indicated willingness to try once weekly dosing and 84% indicated willingness to try once monthly dosing [51]. Interestingly, respondents who had also reported missing ARV doses in the 4 days prior to the survey were significantly more willing to try injectable ARV no matter the dosing interval (weekly, every 2 weeks, monthly), suggesting ARV non-adherent individuals may be more motivated to use LA agents. Further, active substance users were significantly more willing to try injectable ARV than non-users. As active substance use is highly correlated with poor ARV adherence, injectable ARV may be a plausible alternative for this group as well [52].

Prevention

Before the results of the first clinical long-acting pre-exposure prophylaxis (LA-PrEP) trials became public, Meyers et al. conducted a study in New York City [53]. One hundred ninety-seven HIV-negative, young men who have sex with men completed a computer-based survey after receiving information about both LA-PrEP and daily oral PrEP. The mean age was 21 years and 80% were insured. Approximately 40% reported low socioeconomic status, but greater than 75% had attended some post-secondary school. The participants were at relatively high risk for HIV infection as 71% reported inconsistent condom use and 54% reported multiple sexual partners in the preceding 3 months. Approximately 81% of participants reported definite or probable willingness to use LA-PrEP. Participants with higher behavioral risk factors, either more sexual partners or history of sexually transmitted infection, were more willing to use LA-PrEP. Also, participants with a high school or some post-secondary school education were more willing to use LA-PrEP than those who had completed college. In the ÉCLAIR study, 75% of participants reported satisfaction with receiving an injection compared to taking a pill; 79% of participants reported a willingness to continue LA-PrEP as compared to 13% who reported that they would not [54].

Women reported a high acceptance of the dapivirine vaginal ring in a 12-week phase I/II double-blind study of dapivirine ring in sub-Saharan Africa (N = 266), average age 25 years. Over 95% of participants reported that the ring was comfortable and that they were willing to continue it, though approximately 15 and 20% were concerned that the ring may fall out or become lost in their body, respectively [55]. Women also expressed concerns about their sexual partner feeling the ring during intercourse, but less than a quarter reported this occurring and none reported that this had a significant negative impact on acceptability. These issues are being further evaluated. Evidence of suboptimal adherence, including reports that some women deliberately removed their ring prior to intercourse in the phase II studies, underscores the importance of rigorous behavioral studies and user education.

Challenges to Long-Acting Agent Use

While the novel formulations and prolonged half-lives of LA-ARVs confer distinct advantages over traditional agents, this property introduces unique management challenges. For example, once administered, current injectable LA agents cannot be recalled or dialyzed, thus potentially exposing patients to intractable AEs. For this reason, the pragmatic approach in trials to date involves an oral lead-in in order to uncover significant systemic AEs and avoid the LA formulation in affected individuals. Extended exposure to sub-therapeutic plasma concentrations heightens the risk of viral resistance to the LA-ARV and possibly other agents in the same class if the recipient becomes viremic from non-adherence or treatment failure, or in the case of PrEP use, from new HIV infection. There is also a risk of prolonged fetal exposure if pregnancy occurs prior to systemic clearance of the LA-ARV.

Cost is another important consideration that may limit the uptake of LA-ARV. In a mathematical model, LA-ARV was found to be cost-effective for treatment primarily when targeted at poorly adherent individuals with multiple prior treatment failures [56]. Its use as second-line therapy or first-line therapy in ARV-naïve patients only became cost-effective when the annual drug cost was set below 31,000 and $24,000/year, respectively. In the context of prevention, Walensky et al. estimated the number of new HIV infections that needed to be prevented by LA-PrEP in order to offset the additional costs associated with use of LA-PrEP over daily oral PrEP or no PrEP [57]. Applying the Cost-Effectiveness of Preventing AIDS Complications International Model to South Africa, they assumed a 75% effectiveness of LA-PrEP, 62% effectiveness of oral PrEP, and an annual HIV incidence of 5% for South African women between the ages of 16 to 26 years. The estimated monthly costs of drug as well as laboratory and clinic visits were 18.60 and $12.30 for LA-PrEP and oral PrEP, respectively. In the simulations, LA-PrEP was cost saving in comparison to no PrEP and the incremental cost-effectiveness ratio compared to oral PrEP was $150/life-year saved. Cost savings of LA-PrEP over oral PrEP were sensitive to LA-PrEP effectiveness, HIV incidence, and LA-PrEP pricing. It was estimated that $1.6 billion dollars would be needed to provide LA-PrEP to 50% of South Africa’s high-risk women, a prohibitive cost for the country and most countries globally.

Implementation of LA-ARV may expend additional infrastructure. Most HIV patients on successful ART are currently seen for routine follow-up every 4 to 12 months; hence, a requirement for more frequent visits to a healthcare facility to receive injectable LA-ARV may create additional strain on the healthcare system and be unattractive to some patients. Surgical procedures, even if minor, will be needed to insert implantable agents, while agents that require cold chain storage—such as RPV LA—will be particularly difficult to scale up in resource-limited settings. Overall, scale up of LA therapy will require collaborative efforts of healthcare providers, pharmaceutical industry, and global agencies. Lessons learnt from the modest uptake of daily oral PrEP indicate that innovative strategies will be needed to maximize uptake of LA agents for PrEP. Patient and provider factors that stymie the widespread use of daily oral PrEP, and which may also be barriers to LA-PrEP uptake will need to be addressed. Strategies that may be worth evaluating include co-location of LA-PrEP with sexual health clinics, substance abuse treatment centers, and mobile needle exchange units.

Conclusion

LA-ARVs have the potential to further decrease the morbidity and mortality of HIV infection. The pipeline of LA-ARV compounds is expanding. Evidence of long-term safety, a better understanding of the correlates of prevention efficacy, and creative solutions to anticipated implementation challenges are necessary to realize the potential of these promising agents.