Safety and efficacy of deltoid intramuscular long-acting cabotegravir and rilpivirine,a case series

Abstract

Objective

The objective was to evaluate the safety, pharmacokinetics and efficacy of an alternative deltoid site for CAR-LA.

Design

This is a prospective cohort of people living with HIV (PLWHIV) switching for intramuscular long acting cabotegravir and rilpivirine (CAR-LA).

Methods

Clinical and biological data were collected at baseline, month 1 (M1), month 3 (M3), and every three months thereafter, when available.

Results

A total of 15 patients were enrolled between August 2022 and October 2024. Of these, 11 were transgender women, 1 cisgender woman, and 3 cisgender males. The median duration of plasma HIV-1-RNA (pVL) <50 copies/mL before switching to CAR-LA was 42 months. The main criterion for selecting PLWHIV was the presence of injected silicone in the gluteal region (12 out of 15 cases). Median duration of CAR-LA was 13 months. All patients maintained a pVL<50 copies/mL at M1 and M3. No virological failure was observed in the 13 patients who were followed up to M12. One PLWHIV discontinued deltoid injections due to low plasma concentrations at M2, while another discontinued due to non-adherence which led to cabotegravir resistance at M13. Three PLWHIV reported moderate pain at the injection site, with one discontinuing treatment at M12. All CAB and RPV plasma concentrations measured after deltoid CAR-LA were comparable with those obtained following gluteal CAR-LA phase 3 studies.

Conclusion

Deltoïd CAR-LA seems to be a reliable and well-tolerated alternative for PLWHIV with contraindications for gluteal injections.

Introduction

Treatment with intramuscular (IM) long acting cabotegravir (CAB) and rilpivirine (RPV) (CAR-LA) in people living with HIV (PLWHIV) optimizes adherence to antiretroviral therapy, particularly for individuals with compliance challenges or those with frequent travel requirements. To date, the recommended administration site is exclusively the gluteal region. However, some individuals are unable to benefit from this site due to contraindications, such as the presence of injected silicone from illicit cosmetic injection for gender-affirming procedures, gynoid adipose tissue distribution or treatment site fatigue. A study involving 15 healthy volunteers demonstrated that when CAR-LA was administered via the vastus lateralis (lateral thigh) muscles, CAB and RPV plasma concentrations (plC) were adequate. ¹ Additionally, case reports of two patients have shown that deltoid injections result in plC comparable to those achieved with gluteal injections, maintaining a suppressed HIV viral load. ² In this study, we report on 15 cases of PLWHIV who received CAR-LA at the deltoid site.

Methods

We conducted a single-center prospective cohort including virologically suppressed PLWHIV and treated with CAR-LA at Saint Antoine University Hospital (Paris, France). All patients who received IM CAR-LA at the deltoid site between August 2022 and October 2024 were included.

Clinical and biological data (including immunological, virological, and antiretroviral pharmacokinetics) were collected at baseline, month 1 (M1), month 3 (M3), and every three months thereafter, when available. Total duration of follow up have been recorded.

The recommended initial loading doses are a single CAB 600 mg/3 mL IM and a single RPV 900 mg/3 mL IM injection at the same visit, and one month later, followed by the same doses every two months, for maintenance treatment. All deltoid IM injections were administered by the same certified nurse. The same side were used for each injection–one side for RPV and another site for CAB.

Plasma HIV RNA determination (pVL)

The follow up pVL was determined at M0, M3, M7 and M12 using COBAS® 6800 system (Roche Molecular Systems, Branchburg, NJ, USA) or Alinity M. Virological failure (VF) was defined as one pVL >100 c/mL or at least two pVL >50 c/mL.

Genotypic resistance testing

Genotypic resistance testing on either viral RNA or DNA was performed according to the using procedures and primer sequences described at https://www.hivfrenchresistance.org/. Results of historical genotypic resistance testing were retrieved to determine viral subtype, and genotypic resistance testing was performed on samples collected at time of VF to detect potential drug resistance-associated mutations according to ANRS algorithm (version of April 2024).

Determination of plasma drug concentrations (plC)

plC were determined by validated UPLC-MS/MS with limits of quantification below 5 ng/mL for both CAB and RPV (UPLC-MS/MS Acquity, Waters Corporation Milford, MA, USA). ³

CAB and RPV plC results were interpreted according to the respective protein-adjusted concentrations required for 90 % inhibition of viral replication (PA-IC90) for CAB (166 ng/mL) and RPV (12 ng/mL). ⁴

Statistical analyses

The data were collected anonymously from the patients’ medical records and presented as medians and range or IQR for plasma concentrations.

Ethics

Signed informed consent for the use of biomedical data in medical research was obtained from all patients. This study was approved by the Sorbonne University ethics committee (CER-2025-CARLA-00479). In France, administration to a site other than the approved gluteal site constitutes off-label use. Each case was reviewed during a multidisciplinary meeting involving clinicians, virologists, pharmacologists, and adherence nurses, and the off-label deltoid route was proposed only when considered safe and appropriate. This option was offered exclusively to patients deemed capable of fully understanding the information provided. They were clearly informed that the deltoid administration differs from the approved gluteal site and therefore constitutes off-label use, and they provided explicit informed consent prior to administration.

Results

A total of 15 PLWHIV were enrolled in the study. Of these, 11 were transgender women, 1 cisgender woman, and 3 cisgender males (Table 1). Patients have been living with HIV for a median duration of 12 years (min 3; max 35 years) and have received a median of three lines (min 2; max 7) of oral antiretroviral (ARV) treatment prior to switching to CAR-LA.

Table 1.

Clinical and virological data of patients receiving deltoid cabotegravir and rilpivirine long acting intramuscular injection.

No	Gender	Age (years)	Race/ethnicity	Last antiretroviral treatment before switch	BMI D0 (kg/m2)	BMI M7 (kg/m2)	BMI M12 (kg/m2)	Plasma viral load D0 (copies/mL)	Plasma viral load M3 (copies/mL)	Plasma viral load M7 (copies/mL)	Plasma viral load M12 (copies/mL)	HIV-1 sub-type	Resistance mutations to strategy at D0
1	MtF	39	South America	BIC/FTC/TAF	31,4	35	36	<50	<50	<50	<50	BC	Non amplifiable in plasma
2	MtF	49	South America	BIC/FTC/TAF	32,6	34	33	<50	<50	<50	<50	F	None on RT, non amplifiable Integrase o, plasma
3	MtF	30	South America	DOR/3TC/TDF	28	28	NA	<50	<50	<50	<50	B	None
4	MtF	49	South America	DTG/3TC	25,9	25	24.8	<50	<50	<50	<50	F1	R263K/R* on integrase gene in genotype on PBMC
5	H	38	South America	BIC/FTC/TAF	21,1	21	21,4	<50	<50	<50	<50	B	None
6	H	37	South America	TDF/FTC/EFV	28,1	27	28,7	<50	STOP	STOP	STOP	B	None
7	MtF	51	South America	BIC/FTC/TAF	34,5	28	29	<50	<50	<50	<50	Undetermined non B	non amplifiable Integrase
8	MtF	39	Caribbean	DOR/3TC/TDF	39,2	36	36,2	<50	<50	<50	<50	CRF02AG	None
9	MtF	59	South America	TDF/FTC/RAL	28,1	ND	NA	<50	<50	<50	<50	B	None
10	MtF	31	South America	BIC/FTC/TAF	19,2	19	NA	<50	<50	<50	<50	CRF108_BC	None
11	F	58	Sub Saharan Africa	RAL/RPV	34,7	36	33.8	<50	<50	<50	<50	D	179D, 138K* on RT gene in genotype on PBMC
12	MtF	39	Asia	RPV/FTC/TAF	23,1	23.1	NA	<50	<50	<50	<50	CRF01-AE	None
13	MtF	32	South America	BIC/FTC/TAF	27,0	25.6	27.3	58	<50	<50	<50	C	None
14	MtF	38	South America	RPV/FTC/TAF	26.7	NA	NA	<50	<50	<50	NA	CRF01-AE	non amplifiable RT in plasma
15	H	38	South America	BIC/FTC/TAF	29.3	30.9	31.4	<50	<50	<50	<50	CRF31-BC	None

*Apobec mutations.

BIC/FTC/T: Bictegravir/Emtricitabine/Tenofovir Alafenamide.

DOR/3TC/TDF: Doravirine/Lamivudine/Tenofovir disoproxil fumarate.

DTG/3TC: Dolutegravir/Lamivudine.

RPV/FTC/TAF: Rilpivirine/Emtricitabine/Ténofovir alafenamide.

TDF/FTC/EFV: Tenofovir disoproxil fumarate/Emtricitabine/Efavirenz.

TDF+FTC/RAL: Tenofovir/disoproxil fumarate/Emtricitabine/Raltégravir.

3TC/ABC/RPV: Lamivudine/Abacavir/Rilpivirine.

NA: Data not available.

The median CD4 count and nadir were 612 cells/mm³ (range: 271-1173) and 404 cells/mm³ (range: 4-664), respectively.

The median duration of pVL <50 copies/mL prior to the initiation of CAR-LA was 42 months. One patient (No. 13) had a pVL of 58 copies/mL at CAR-LA initiation. Injectable therapy was started for this patient due to poor adherence to oral ARV therapy. At M1, pVL was undetectable. Median duration of deltoid CAR-LA was 13 months (min 2, max 20 months).

All PLWHIV demonstrated immunity against hepatitis B virus, with anti-HBs antibodies >10 IU/mL, either from prior infection or vaccination. The median BMI at CAR-LA initiation was 28 kg/m² (range: 19.2–39.2), which remained unchanged during follow-up.

Seven PLWHIV were beneficiaries of the French state medical aid (AME), and one faced an interruption in health insurance coverage during the study period.

Eleven PLWHIV were receiving hormone therapy, of which 8 were transgender women receiving estrogen therapy, and three were cisgender male sex workers on testosterone therapy.

The primary reason for selecting the deltoid injection site was the presence of silicone in the gluteal region (12 out of 15 cases). These 12 participants never received IM injections (CARLA or antibiotics). One patient, who was morbidly obese (BMI of 39.2 kg/m²), presented challenges with accessing the gluteal muscle despite the use of long needles. Another PLWHIV had nodules at the gluteal injection site attributed to previous antimalarial injections. Participant No.6 exhibited inadequate plC of CAB (<5 ng/mL) and RPV (29 ng/mL) after CAR-LA gluteal administration at M5. He had a BMI of 27 kg/m² and no drug interactions.

All PLWHIV maintained a pVL <50 copies/mL at M1.

At M3, 14 participants had an undetectable pVL. Participant No.6, had subtherapeutic CAR-LA levels following gluteal injections, which led to VF at M5 with an HIV pVL of 798 copies/mL. He was subsequently switched to deltoid injections and achieved an undetectable viral load two months later, although the plC remained subtherapeutic. This prompted the discontinuation of CAR-LA and the transition to an oral regimen (lamivudine/dolutegravir), resulting in sustained pVL suppression. The HIV genotype test revealed no resistance mutations.

At M7, 14 PLWHIV had a pVL<50 copies/ml. At M12, 13 PLWHIV had undetectable pVL, while one had not reached yet the 12 months on follow-up.

Overall, three PLWHIV discontinued treatment: patients No.6 (as previously described), No.1, and No.10. Patient 1 experienced VF at M13, with pVL=161 copies/mL, primarily due to non-adherence to treatment associated with social instability. This led to the discontinuation of CAR-LA and the emergence of multiple resistance mutations (74I, 97A, 155H), linked to CAB resistance. The reverse transcriptase gene could not be amplified. In response, the treatment regimen was modified to include emtricitabine, tenofovir disoproxil fumarate, darunavir, and ritonavir. Following this adjustment, the patient achieved virological suppression one month later.

Patient No. 10 stopped injections at M12 due to pain at the injection site and treatment fatigue and opted to resume oral treatment. He had an undetectable pVL at M12 and restarted treatment with bictegravir, emtricitabine and tenofovir alafenamide. He had a sustained viral suppression on oral therapy after the switch.

All patients except patient No. 6 had plC within the therapeutic range. CAB and RPV plC determined at follow up visits (subject to sample availability) are presented in Figure 1. Mean +/- SD of interval between last deltoid CAR-LA and sampling were 26 +/- 4 days (M1, post-loading dose at initiation, n=10), 54 +/- 9 days (M3 or M5, n=10) and 48 +/- 18 days (M7 or M13, n=8), respectively. Median (IQR 25-75 %) CAB and RPV Cpl at M1, M3 or M5 and M7 or M13 were 1,612 ng/mL (1,131 – 1,965) and 44 ng/mL (34 – 52), 1,598 ng/mL (1,365 –1,943) and 44 ng/mL (35 - 68) and 1,253 ng/mL (1,107 – 2,174) and 67 ng/mL (47 - 80), respectively. Among all determined plC, none was below the in vitro protein-adjusted concentrations required for 90% viral replication inhibition (PA-IC₉₀), (166 and 12 ng/mL for CAB and RPV, respectively) and only one had CAB plC below the corresponding 4 x PA-IC₉₀ (664 ng/mL for CAB) at M1.

Figure 1.

Plasma concentrations of RPV and CAB determined at follow up visit.

All reported adverse events were mild to moderate. Three PLWHIV reported myalgias at the injection site following deltoid injection, with one PLWHIV opting to discontinue the injections after 12 months of treatment. One PLWHIV experienced presyncope without loss of consciousness after an injection at 18 months of treatment, with no identifiable etiology.

Discussion

To our knowledge, this is the first real life case series including 15 PLWHIV treated with long-acting CAB and RPV administered at the deltoid muscle site. The follow-up results reported here demonstrate that injections at the deltoid site seem to be both effective and well-tolerated.

All but one CAB and RPV plC determined post deltoid CAR-LA were within the range of those obtained post gluteal CAR-LA injections. ⁵

Only one PLWHIV had VF due to poor adherence secondary to social instability. It is therefore important to note that social precarity and treatment adherence must be considered when prescribing long-acting treatments.

The injections were well tolerated with only one PLWHIV who discontinued the injections due to pain. This approach therefore provides an important alternative, especially for individuals with contraindications to intragluteal injections. In other contexts, such as psychiatric treatment, the deltoid route has also been proposed, showing good acceptance and accurate dosing.^6,7 Han et al. highlighted that previous studies have reported 2 mL as the maximum recommended volume for intramuscular (IM) injections in the deltoid region in adults, while gluteal IM injections have been documented as tolerable up to 5 mL. ¹ Nursing societies recommend maximum IM injection volumes of 2 mL for the deltoid, 4 mL for gluteal sites, and 5 mL for the vastus lateralis. ⁸ Although the volume of CAR-LA injections was 3 mL, only a few patients reported mild to moderate pain at the injection site, that led to treatment discontinuation in only one PLWHIV after 12 months.

Individuals with a gynoid adipose tissue distribution, particularly in the gluteal region, may experience VF, as reported by de Malliard et al., ⁹ due to low plC of CAB and RPV. In such cases, using ultrasound to measure subcutaneous fat thickness at the injection site could help prevent unintentional subcutaneous administration. ⁹ However, this method is costly, not widely accessible, and requires specialized training. Use of IM deltoid administration could be a good alternative for these patients. Our findings are supported by Cossu et al. who reported two cases in which CAB and RPV plC were comparable 24 weeks after switching from the gluteal to the deltoid IM injection sites. Optimal immune-virologic control was maintained, with no reported adverse effects related to CAR-LA injections. ² A study by Han et al. demonstrated that IM administration of CAR-LA in the vastus lateralis (lateral thigh) muscles had comparable efficacy and safety to gluteal injections. ¹

Among the three PLWHIV using steroids for bodybuilding, one exhibited a very low CAB and RPV plC, while the other two maintained therapeutic levels. Serum testosterone levels were not measured. Two potential explanations for the subtherapeutic drug level were identified. The first is drug interactions with over-the-counter anti-aromatase medications, commonly used to prevent gynecomastia, which induces CYP3A4 activity. No other drug has been found to be inducer of UGT1A1 or UGT1A9. The second is enhanced drug clearance due to increased muscle blood flow associated with anabolic androgen steroid use for athletic and body building purposes. ¹⁰

The administration of long-acting cabotegravir and rilpivirine in the deltoid region appears to be a reliable and well-tolerated alternative for patients with contraindications to gluteal or thigh injections. However, further studies with a larger patient population are needed to confirm these findings.