Strategic use of salvage long-acting antiretrovirals in the setting of resistance

Abstract

Purpose

Long-acting cabotegravir/rilpivirine (LA-CAB/RPV) was approved for use in virally suppressed patients with human immunodeficiency virus (HIV) in January 2021. While this was a paradigm shift for many patients living with HIV, as LA-CAB/RPV was the first injectable complete regimen for the treatment of HIV, several patient populations, including those lacking virologic suppression, have not been able to easily access this advance in science and care.

Summary

In this article, we provide an update on 2 patients from our previous report and describe one further patient who experienced treatment failure following initiation of LA-CAB/RPV. Additionally, we review reports published to date of the clinical outcomes of patients with viremia who have accessed LA-CAB/RPV in the setting of baseline resistance-associated mutations (RAMs) to either component and any resulting RAMs at virologic failure. On the basis of this evidence, we recommend that hybrid or all-injectable regimens be considered for patients who have struggled with adherence to oral antiretroviral therapy or have partial or full resistance to one component of LA-CAB/RPV.

Conclusion

The case series reported here adds to literature supporting the notion that LA-CAB/RPV can be successfully used in patients who are viremic.

Key Points.

• The case series reported here adds to literature supporting the notion that long-acting cabotegravir/rilpivirine (LA-CAB/RPV) can be successfully used in patients who are viremic.

• Hybrid or all-injectable regimens should be considered for patients who have struggled with adherence to oral antiretroviral therapy (ART) or have partial or full resistance to one component of LA-CAB/RPV.

• Patients with a large viral reservoir may experience an extended time to viral suppression even while receiving injectable ART, at the same time seeing improvements in their immune system and averting HIV-related sequelae.

We previously described our center’s experience with youth with human immunodeficiency virus (HIV) and adherence struggles who were viremic at baseline and were initiated on long-acting antiretroviral agents as a salvage strategy.¹ In this previously published case series, 2 of the 3 patients (67%) were able to achieve viral suppression after initiation of long-acting cabotegravir/rilpivirine (LA-CAB/RPV), in one case with the addition of other antiretrovirals. These patients rapidly achieved viral suppression, in one case for the first time in decades. We now report on the longer-term clinical outcomes of our center’s cohort of patients who received salvage, off-label combination long-acting antiretroviral therapy (ART) while they were HIV viremic, the majority of whom had non-nucleoside reverse transcriptase inhibitor (NNRTI) resistance-associated mutations (RAMs) at baseline, which has been identified as one of the baseline risk factors for virologic failure on LA-CAB/RPV across phase 3 trials.² We also report on an additional patient who experienced virologic failure and was subsequently found to have had baseline NNRTI resistance. Further, we examine the growing body of literature reporting on patients who were started on LA-CAB/RPV while viremic and achieved viral suppression.^3-6 There is a need to gather evidence on this growing topic to support a potential shift in recommendations and effect a needed change in payor coverage criteria. Currently, US clinical guidelines recommend use of LA-CAB/RPV based on the Food and Drug Administration (FDA)–approved indication, with LA-CAB/RPV used only as “switch therapy” in patients who have already achieved virologic suppression.^7,8 This approach continues to overlook patients who may benefit the most from these medications due to (1) past or ongoing difficulties with adherence; (2) lack of identification as a “long-acting ART candidate” by their providers; and (3) lack of randomized controlled trial data because viral suppression has been a prerequisite for participation in all the formative trials. Patients with perinatally acquired HIV or those who have lived with HIV since the beginning of the epidemic have endured the learning curves of scientific advances, including serial monotherapy resulting in multidrug-resistant (MDR) HIV. These patients should rightly be considered high priority for access to long-acting injectables as a tool for survival.⁹ With this paper, we aimed to provide an update on patients at our institution who accessed the long-acting injectable while viremic and report on a new patient who experienced failure, as well as review the growing body of literature on the outcomes of patients initiating LA-CAB/RPV while viremic.

Case series

Case 1 (virologic suppression with RAMs)

At 30 years old, Ms. X had last achieved viral suppression 14 years before publication of our previous report.¹ She was started on hybrid injectable therapies (CAB/RPV + ibalizumab [IBA]) in the winter of 2021 after an extended acute immunodeficiency syndrome (AIDS)–related hospital stay in the intensive care unit. Her medical history, including RAMs and barriers to care that led to her hospitalizations, was previously summarized.¹ Thirteen months into her hybrid injectable therapy, her IBA home infusions were discontinued, with disruptions to vascular access named as the leading cause among several reasons for discontinuation (none of which were related to efficacy). For a 2-month period, the patient received only oral once-daily darunavir/cobicistat (DRV/c) and once-monthly LA-CAB/RPV. With concerns still looming over her slowly decreasing viral load and her ability to consistently take oral ART in conjunction with LA-CAB/RPV, the decision was made to change DRV/c to DRV/c/tenofovir alafenamide fumarate/emtricitabine (DRV/c/TAF/FTC) for additional coverage while attempting to gain access to twice-yearly lenacapavir (LEN) subcutaneous injections. This circumvented the need for indwelling vascular access and eliminated the need for frequent home visits from the nursing team. A new appeal letter to continue her LA-CAB/RPV was required by the payor, as annual reauthorization of the medications was denied citing her viral load at the time of reauthorization of 367 copies/mL, which was above their requirement of fewer than 50 copies/mL. With implementation of injectable antiretrovirals in the patient’s regimen, her number of hospitalizations decreased from 8 AIDS-related admissions in the previous year to 2 non–AIDS-related admissions in the subsequent year. LEN was ultimately initiated 2 months after discontinuation of IBA, at month 15 in the LA-CAB/RPV course. Since initiation of LEN, the patient has not had any further hospitalizations. Viral suppression to fewer than 20 copies/mL was finally attained 22 months after starting a continuous regimen of injectable antiretrovirals.

Case 2 (virologic suppression with RAMs)

Ms. Y is a 40-year-old who previously struggled with oral adherence since being diagnosed with HIV at 5 years of age. She was started on LA-CAB/RPV in the summer of 2021 and later requested to be transitioned to every 2-month dosing. Upon transitioning to this dosing schedule, it was discovered that the A98G mutation that was present at baseline had been recategorized as conveying low-level resistance to RPV (2-fold change).¹⁰ This mutation was originally found in the “novel mutations detected” section of the genotype report, which usually contains RAMs that are not typically analyzed for their effect on antiretroviral susceptibility. The risks of switching to every 2-month dosing were explained to the patient. Given the risks when discontinuing LA-CAB/RPV, which included reversion to nonadherence on an oral regimen and the possibility of acquiring further resistance to long-acting ART, the patient chose to continue with once-monthly dosing. She has maintained continuous viral suppression since starting LA-CAB/RPV.¹ Despite low-level resistance to RPV being detected after initiation, we wanted to highlight this patient’s continued viral suppression and the potential prevention of virologic failure by maintaining once-monthly dosing and therefore higher antiretroviral trough concentrations.¹¹

Case 3 (virologic failure with a RAM)

Mr. D is a 38-year-old with HIV that was diagnosed 6 years prior and had a baseline CD4⁺ T cell count of 243 (11%) and HIV RNA load of 371,829 copies/mL. He was initiated on efavirenz/tenofovir disoproxil fumarate/FTC (EFV/TDF/FTC) while in the inpatient setting. Baseline resistance testing revealed resistance to all NNRTIs (except RPV, as it was not at that time listed as an NNRTI on the laboratory report), and his regimen was changed to elvitegravir/cobicistat/TDF/FTC (EVG/c/TDF/FTC) 1 month later when the genotype results were obtained. At the time of the switch, his HIV RNA viral load had decreased to 99 copies/mL. EVG/c/TAF/FTC was fully suppressive, as was bictegravir/TAF/FTC (BIC/TAF/FTC), which was begun almost 3 years before initiating CAB and RPV oral lead-in in the fall of 2021. Mr. D was already virally suppressed upon starting once-monthly LA-CAB/RPV in late 2021 and maintained viral suppression for 5 months, at which point he was switched from once-monthly to every 2-month LA-CAB/RPV based on his preference for less frequent injection visits. Approximately 10 months after switching from a once-monthly to an every 2-month schedule, an HIV viral load of 2,990 copies/mL caused concern; the test was repeated with a result of 212 copies/mL. In again reviewing his initial 2015 genotype results, it was recognized that at baseline, before initiation of any ART, he had an Y188HL major mutation; however, despite RPV being FDA approved in 2011, it was not listed with the other medications in the NNRTI class on the patient’s 2015 genotype report. It is now known that this RAM reduces RPV susceptibility by 5-fold. The patient’s body mass index while on LA-CAB/RPV was stable at 21 kg/m². He was started on oral DRV/c/TAF/FTC. No updated viral load or additional resistance data were available at the analysis cutoff.

Literature review

Several reports have been published of the outcomes with LA-CAB/RPV in patients who were initiated on this therapy while viremic. A group from the University of California San Francisco (UCSF) reported initiating long-acting injectable therapy in a cohort of patients, of whom 57 of 133 (43%) were viremic; 95% of these patients achieved viral suppression on long-acting injectable therapy.³ Two of the patients who experienced virologic failure had baseline minor mutations, and one had a drug interaction that likely precipitated failure (Table 1). Ten of 12 (83%) patients from the University of Mississippi’s 2 patient cohorts had baseline RAMs to either NNRTIs or integrase strand transfer inhibitors (INSTIs).⁴ The first cohort of 6 patients were started on a once-monthly schedule and then transitioned to every 2-month dosing, while the second cohort of 6 patients was initiated on every 2-month dosing (one patient had an INSTI RAM, E157Q, and a second had NNRTI RAMs K103N and P225H). At the time of publication, all 12 patients had achieved viral suppression within 3 months. A group from Community Health in North Carolina (CAN) reported on one patient who achieved viral suppression despite INSTI- and NNRTI-associated mutations.⁵ Moreover, a description of compassionate use of LA-CAB/RPV was published by D’Amico et al,⁶ in which 80% of patients were viremic at the time of initiation. Patients with primary mutations were excluded, with a few exceptions. Of this cohort, 73% of patients with perinatally acquired HIV were found to achieve viral suppression on LA-CAB/RPV, demonstrating efficacy in patients who have been exposed to numerous regimens.

Table 1.

Summary of Reports of Patients Initiated on LA-CAB/RPV With Baseline HIV RNA Level of >50 Copies/mL

	UCSF3	Johns Hopkins Hospital1	University of Mississippi4	CAN5	ViiV Healthcare compassionate use program6
Number reported on injectable ART	133	122	12	1	35
Patients viremic at baseline	57 (43%)	3 (2.5%)	12 (100%)	1 (100%)	28 (80%)
Patients with virologic failure at data cutoff	3/57a	1/3	NA	NA	7/35b
Mean baseline HIV RNA level	16,218 copies/mL	6,620 copies/mL (range, 4,310-8,930 copies/mL)	152,657 copies/mL (range, 2,410-566,000 copies/mL)	341,000 copies/mL	60,300 copies/mL (range, 86 to >10,000,000 copies/mL)
Median time to viral suppression	33 days (range, 28-56 days)	12.5 months (range, 3-22 months)	3 months	3 months	5 months (range, 1-31 months)
Median time to viral failure	33 weeks	10 months	NA	NA	NR
Patients with known RAMs at baseline	2	2	10	1	5
RAMs detected at baseline	Pt. 1c: RT: V179I; Pt. 2d: IN: T97A	Pt. 1: RT: K101E, Y181YC; Pt. 2: RT: A98G; Pt. 3: RT: Y188HL	RT: K103N (2), V106I (1), P225H (1); IN: D232N (1), N155H (1), E157Q (3)	RT: K103N; IN: E157Q, S230N	RT: M184V (2), K219E (1), K238KR (1), E138G (2), V118IV (1), V179I (1), R211K (1), K103N (2), T215Y (1); IN: E138D (1)
RAMs detected at failure	Pt. 1c: RT: L100I, Y181I; Pt. 2d: RT: E138K; IN: R263K	Pt. 1: NA; Pt. 2: NA; Pt. 3: RT: Y188HL	NA	NA	RT: E138E/K (1), E138G (2), M230L (1), Y181I (1), K101E (2),Y181Y/C (2), K219N (1); IN: G118G/R (1), Q148R (2), N155H (1), M50M/I, E138D (1), E157E/Q (1), E138E/K (1), G163E (1)
Other factors contributing to failure	Drug interaction (lingering induction effects of rifabutin usage 2 weeks before LAI) with one patient	Pt. 1 likely has large viral reservoir due to frequent viral loads in millions off ART Pt. 3 had baseline major NNRTI RAM not reported on assay that was a risk factor when transitioned to every 2-month dosing	NA	NA	1 of the 3 patients with BMI >30 kg/m2 did not achieve viral suppression; the other 2 patients experienced a viral blip of 55 copies/mL (1) or viral rebound (1)
Outcome	Viral suppression achieved in 95% of patients with viremia by data cutoff 19% of patients achieved viral suppression for the first time Of the 3 patients with virologic failure, one resuppressed on oral boosted PI-based ART, the second refused oral ART and has remained on injectable ART, and the third has remained on LA-CAB/RPV with the addition of LEN	Viral suppression was achieved in 66% of patients from the initial report. The last patient reached viral suppression 22 months after initiation and after being switched from ibalizumab to LEN + LA-CAB/RPV Two patients with perinatally acquired HIV achieved or maintained viral suppression One patient experienced virologic failure almost 10 months after switching to every 2-month dosing; a baseline NNRTI RAM was discovered at failure	Viral suppression achieved in 100% of patients by data cutoff	Patient achieved virologic suppression	Viral suppression achieved in 57% of viremic patients by data cutoff 73% of patients with perinatally acquired HIV achieved or maintained virologic suppression Seven patients discontinued due to incomplete virologic response ◦ Two experienced viral rebound before failure (one patient was suppressed at initiation) ◦ Four did not achieve viral suppression after receipt of 2-11 injections ◦ All 6 patients with baseline viremia had new NNRTI RAMs at discontinuation (2 of these had E138G at baseline) ◦ New INSTI RAMs were observed in 3 patients at discontinuation (2 of these patients had RAMs conferring resistance to CAB)

Abbreviations: ART, antiretroviral therapy; BMI, body mass index; CAB, cabotegravir; CAN, Community Health in North Carolina; HIV, human immunodeficiency virus; IN, integrase; INSTI, integrase strand transfer inhibitor; LA-CAB/RPV, long-acting cabotegravir/rilpivirine; LAI, long-acting injectable; LEN, lenacapavir; NA, not applicable; NNRTI, non-nucleoside reverse transcriptase inhibitor; NR, not reported; PI, protease inhibitor; RAMs, resistance-associated mutations; RT, reverse transcriptase; UCSF, University of California San Francisco.

^aViral rebound: one patient with a higher BMI of 32 kg/m² and shorter needle use likely contributing to low drug levels experienced viral rebound after previous viral supp ression. No RAMs were observed in either the reverse transcriptase or integrase gene. This patient immediately achieved resuppression with oral ART.

^bViral rebound: the patient was suppressed at entry and experienced a viral blip of 55 copies/mL at injection 19. They were resuppressed on darunavir/cobicistat + tenofovir alafenamide fumarate/emtricitabine. At the time of failure, a new cabotegravir RAM, G118G/R, was detected. No NNRTI RAMs were detected.

^cLA-CAB/RPV discontinued. Patient declined oral ART.

^dThe decrease in HIV RNA at first maintenance injection was less than 2 log₁₀, but by the third injection viral load was <30 copies/mL. Patient declined oral ART at database closure.

In our center’s experience, patients demonstrated benefit from regimen simplification with long-acting injectable ART, despite an often extended time period to achieve viral suppression in the setting of a large viral reservoir, as in case 1.^1,3,6 Similarly, an individual from the UCSF cohort had LEN added to LA-CAB/RPV when their HIV RNA level was 182 copies/mL following 5 injections of LA-CAB/RPV. An updated viral load status for this patient has not yet been reported.³ The addition of LEN at a lower HIV RNA level than in the original studies is another example of novel utilization of market-available tools succeeding in helping patients to achieve viral suppression.^12,13

Considering broader findings from these studies, the immunological outcomes in each report are also largely positive, with mean increases in CD4⁺ T cell numbers to a range of around 150 to 200 cells in relatively short periods of time.^1,3-6 Most reported failures were in patients who accessed LA-CAB/RPV via the manufacturer’s compassionate use program.⁶ D’Amico et al⁶ reported that 6 of the 7 patients (86%) who had experienced incomplete virologic response at the time of the data cutoff were viremic at initiation; new NNRTI mutations were found in 6 patients, and 4 developed new INSTI mutations (Table 1). Despite ultimately experiencing virologic failure, each of the 6 patients saw a decrease in HIV RNA levels that might not have occurred had they not accessed LA-CAB/RPV. In some of these patients, it is unclear whether the recommended 2-inch needle was utilized for those with a body mass index of greater than 30 kg/m², given these patients’ early access to the medication before needle length specifications were worked out.^6,7,14 Overall, the majority of patients who experienced virologic failure in each report were placed on a boosted darunavir–based regimen (some regimens added boosted darunavir to LA-CAB/RPV or TAF/FTC).^1,3,6 Final clinical outcomes for some of these patients in terms of whether they were able to resuppress on oral ART after experiencing virologic failure on a long-acting injectable remain unknown.

Discussion

Commonly reported themes for patients with viremia accessing long-acting ART include psychological barriers (pill burden, stigma, cognitive delay, mental health), physical challenges (gastrointestinal issues, difficulty swallowing), housing instability, and substance use disorder.^1,3-6 In the majority of reports of use of LA-CAB/RPV in patients with HIV viremia, primary mutations were exclusionary for initiation of injectables. However, given the global prevalence of preexisting NNRTI resistance, particularly in regions with widespread rollout of EFV-based regimens, people with known baseline resistance to either NNRTIs or INSTIs should be considered for a hybrid regimen, rather than avoiding use of LA-CAB/RPV. This could include addition of an oral antiretroviral from a different class or a different injectable antiretroviral with an alternative mechanism of action. For patients for whom therapy has failed, it may be prudent to test for additional RAMs to CAB and RPV during the pharmacokinetic “tail phase” after the terminal injection, to guide future therapy decision-making, especially if the patient is unable to start oral ART after discontinuing LA-CAB/RPV.

Laboratory results of RAMs that have not yet been classified as clinically relevant to antiretroviral efficacy or RAMs that have been reclassified as in case 2 warrant particular attention and caution upon initiation of or transition to LA-CAB/RPV. There is often a lag time between detection of a mutation and awareness of its clinical relevance, leading to a discrepancy in the interpretation of antiretroviral susceptibility between resistance surveillance agencies and the agencies that create the laboratory reports that clinicians review, as in case 3. This lag time is likely due to differences in laboratory processing times and procedures but should further reinforce the caution needed when utilizing LA-CAB/RPV as a component of an antiretroviral regimen.^15-17 Cumulative resistance should be interpreted, utilizing manufacturer package inserts and available resistance databases.

Conclusion

On the basis of the available literature characterizing successful treatment of viremic patients (many of whom had baseline resistance) with LA-CAB/RPV, providers can consider initiating LA-CAB/RPV in this setting with appropriate tweaks and adjunctive ART. However, this should be an individualized approach, and addition of other long-acting antiretrovirals such as LEN or IBA should be considered in patients with MDR HIV, a history of pill aversion or intolerance, or cognitive impairment that limits their ability to achieve viral suppression with the use of oral ART alone. Attainment of viral suppression for patients in whom it has been elusive will require innovation—both in the creation of new tools and in optimal, individualized use of available tools such as long-acting antiretrovirals.

Data availability

The data underlying this article cannot be shared publicly due to the privacy of the individuals who participated in the study.

Disclosures

Dr. Weld was supported during the writing of this article by a National Institutes of Health K23 mentored career award under award number 1K23AI165290. Dr. Agwu has participated on the expert scientific advisory board (Gilead, Merck, and ViiV) within the past 36 months. Additionally, she has served as a site principal investigator for a multisite study (Merck and Gilead) and an investigator-initiated study (Gilead). Dr. Kilcrease has declared no potential conflicts of interest.