Treatment-related early discontinuations and adverse events among newly diagnosed people living with HIV initiating integrase inhibitors in a real-world setting

Abstract

Background

Cohort studies suggest higher discontinuation rates with integrase strand transfer inhibitors (INSTIs) than are seen in clinical trials. We assessed discontinuations and adverse events (AEs) that were considered related to the initial INSTI in the first year following initiation among treatment-naïve people living with HIV (PLWH).

Methods

Newly diagnosed PLWH initiating raltegravir, elvitegravir/cobicistat, dolutegravir or bictegravir in combination with emtricitabine/tenofovir alafenamide or emtricitabine/tenofovir disoproxil fumarate between 10/2007 and 1/2020 at the Orlando Immunology Center were included. Unadjusted incidence rates (IRs) and incidence rate ratios (IRRs) were calculated for treatment-related discontinuations and AEs associated with the initial INSTI in the first year following initiation.

Results

Of 331 enrolled, 26 (8%) initiated raltegravir, 151 (46%) initiated elvitegravir/cobicistat, 74 (22%) initiated dolutegravir and 80 (24%) initiated bictegravir. Within the first year, treatment-related discontinuations occurred in 3 on elvitegravir/cobicistat (IR 0.02 per person-years (PPY)) and 5 on dolutegravir (IR 0.08 PPY); no treatment-related discontinuations occurred among those initiating raltegravir or bictegravir. Eleven treatment-related AEs occurred in 7 on raltegravir (IR 0.46 PPY), 100 treatment-related AEs occurred in 63 on elvitegravir/cobicistat (IR 0.72 PPY), 66 treatment-related AEs occurred in 37 on dolutegravir (IR 0.97 PPY) and 65 treatment-related AEs occurred in 34 on bictegravir (IR 0.88 PPY). Unadjusted IRRs did not reveal any significant difference between INSTIs in terms of early treatment-related discontinuations or AEs.

Conclusions

In our cohort, treatment-related AEs occurred in 43% initiating INSTIs but were responsible for early discontinuation in only 2% with no treatment-related discontinuations observed among those initiating RAL or BIC.

Keywords

INSTI discontinuations adverse events real-world safety tolerability

Background

Clinical trials of integrase strand transfer inhibitor (INSTI)-based regimens in treatment-naïve people living with HIV (PLWH) have revealed that discontinuations due to adverse events (AEs) occurred in <5%, and discontinuations due to neuropsychiatric AEs (NPAEs) occurred in <2%.^1,2 Real-world data from cohorts of treatment-naïve and -experienced PLWH report INSTI discontinuation rates due to AEs of 3.3–9.9% in the first year following initiation. NPAEs, which have been more frequently observed in DTG users have been associated with discontinuation rates of 0.7–5.6%.^3–6

There are few descriptions of the real-world safety and tolerability of BIC/emtricitabine (F)/tenofovir alafenamide (TAF) among newly diagnosed PLWH. Data from the OPERA cohort, evaluating treatment-naïve PLWH with advanced HIV initiating first-line 3-drug regimens, demonstrated significantly fewer discontinuations of BIC/F/TAF compared to other regimens. However, reasons for discontinuation were not available in approximately 50% of cases and only 10% experienced treatment-related AEs prior to discontinuation, suggesting that safety was not the primary driver of differences in discontinuation.⁷ In a Spanish cohort following 1584 PLWH initiating BIC/F/TAF (89% treatment-experienced), there were 61 treatment discontinuations by month 12, with toxicity being the primary reason for discontinuation (two-thirds due to gastrointestinal (GI) and NPAEs).⁸ Data from a German cohort of 943 PLWH initiating BIC/F/TAF (93.4% treatment-experienced) also demonstrated a higher treatment-related discontinuation rate of BIC/F/TAF (5.3%) compared to that seen in clinical trials and observed discontinuations due to NPAEs in 3.3%.⁹

Given that BIC/F/TAF is widely used for treatment initiation, it is important to understand its real-world tolerability compared to other INSTIs in this setting. We assessed ‘early’ treatment-related discontinuations and AEs defined as those occurring within the first year following initiation of RAL, EVG/c, DTG and BIC in combination with emtricitabine/tenofovir disoproxil fumarate (F/TDF) or F/TAF among newly diagnosed PLWH.

Methods

This retrospective cohort study evaluated early treatment-related discontinuations and AEs among treatment-naïve PLWH initiating RAL, EVG/c, DTG or BIC in combination with F/TDF or F/TAF between October 2007 and January 2020 at the Orlando Immunology Center (OIC). The nucleoside reverse transcriptase inhibitor (NRTI) backbone was limited to F/TDF or F/TAF to allow for a cleaner comparison of the INSTIs given the diversity of toxicities related to different NRTI backbones. PLWH were excluded if they initiated an INSTI in combination with F/TDF or F/TAF and an additional antiretroviral.

Demographics, lab values and clinical parameters were extracted from the charts of all eligible PLWH through Week 48 of INSTI treatment or until the time of INSTI discontinuation, if this occurred prior to Week 48 (discontinuation was defined as a change or stop in the initial INSTI). The primary endpoint was incidence of early treatment-related discontinuations associated with the initial INSTI. We also reported incidence of treatment-related AEs associated with the initial INSTI. All AEs were graded using the Division of AIDS (DAIDS) Table for Grading the Severity of Adult and Pediatric Adverse Events.¹⁰

Descriptive statistics were calculated for baseline demographic and clinical characteristics, and treatment-related discontinuations and AEs throughout the study. For discontinuations and AEs, treatment-relatedness was based on whether a possible relationship between the discontinuation or AE and initial INSTI was documented (discontinuations and AEs related to the NRTIs were excluded). The following symptoms were classified as NPAEs: insomnia, sleep disturbance, dizziness, headaches, depression, poor concentration and anxiety. Unadjusted incidence rates (IRs) and incidence rate ratios (IRRs) with associated 95% confidence intervals (CIs) were calculated for early treatment-related discontinuations and AEs and were compared using chi-square testing. The Sterling Institutional Review Board (IRB) determined that the study met IRB exemption criteria based on the observational nature of the study, which utilized retrospective data collected as a part of routine clinical care (Sterling IRB ID 7410).

Results

Between October 2007 and January 2020, 487 treatment-naïve PLWH received a prescription for an INSTI in combination with F/TDF or F/TAF at the OIC. Ninety-eight were subsequently referred to a clinical trial prior to antiretroviral initiation and 59 were lost-to-follow-up after the initial visit, leaving 331 eligible for enrollment. The median age (range) was 32 years (16, 73), 11% were female and 42% were non-White (Table 1). Twenty-eight percent had baseline CD4⁺ count <200 cells/mm³ and 35% had baseline HIV-1 RNA ≥100,000 copies/mL. Twenty-six (8%) started RAL, 151 (46%) started EVG/c, 74 (22%) started DTG and 80 (24%) started BIC (Table 1). One-hundred and twenty (36%) started an INSTI in combination with F/TDF and 211 (64%) started in combination with F/TAF.

Table 1.

Baseline demographic and clinical characteristics.

Characteristic	N = 331
Median Age (range)	32 (16, 73)
Sex
Male, n (%)	294 (89)
Female, n (%)	37 (11)
Race
Caucasian, n (%)	192 (58)
Black, n (%)	96 (29)
Asian, n (%)	4 (1)
Other, n (%)	39 (12)
Ethnicity
Hispanic/Latino, n (%)	40 (12)
Not Hispanic/Latino, n (%)	291 (88)
Baseline CD4⁺ cell count, median (range), cells/mm³	367.5 (4, 1445)
≥200 cells/mm3, n (%)	238 (72)
<200 cells/mm3, n (%)	93 (28)
Baseline HIV-1 RNA, median (range), c/mL	56,000 (10, 21,800,000)
<100,000 c/mL, n (%)	217 (65)
≥100,000 c/mL, n (%)	114 (35)
HIV disease status
Asymptomatic, n (%)	199 (60)
Symptomatic, n (%)	39 (12)
AIDS, n (%)	93 (28)
Initial INSTI regimen
RAL-based regimens, n (%)	26 (8)
RAL + F/TDF, n (%)	26 (8)
RAL + F/TAF, n (%)	0
EVG/c-based regimens, n (%)	151 (45)
EVG/c/F/TDF, n (%)	73 (22)
EVG/c/F/TAF, n (%)	78 (23)
DTG-based regimens	74 (22)
DTG + F/TDF, n (%)	21 (6)
DTG + F/TAF, n (%)	53 (16)
BIC-based regimens, n (%)	80 (24)
BIC/F/TAF, n (%)	80 (24)

Abbreviations. C/mL, copies/mL; AIDS, acquired immunodeficiency syndrome; INSTI, integrase strand transfer inhibitor; RAL, raltegravir; F, emtricitabine; TDF, tenofovir disoproxil fumarate; TAF, tenofovir alafenamide; EVG/c, elvitegravir/cobicistat, DTG, dolutegravir, B, bictegrarvir.

Eight (2%) discontinued INSTIs early due to an AE, yielding IRs of 0.022 discontinuations per person-years (PPY) among 3 on EVG/c and 0.08 discontinuations PPY among 5 on DTG. No treatment-related early discontinuations occurred among those on RAL or BIC (Table 1). Among 3 on EVG/c, 1 on EVG/c/F/TDF discontinued due to ‘ongoing bone loss’ as an updated DEXA scan completed 9 months after EVG/c initiation demonstrated a change in diagnosis from ‘osteopenia’ to ‘osteoporosis’ compared to a DEXA scan 12 months prior to HIV diagnosis. Two on EVG/c/F/TAF discontinued due to GI upset (1) and weight gain (1). Among 5 on DTG, 3 on DTG + F/TDF discontinued due to gastrointestinal upset (2) and anorexia (1). Two on DTG + F/TAF discontinued due to weight gain (1) and insomnia (1). Unadjusted IRRs demonstrated no significant difference in the incidence of early treatment-related discontinuations among the different INSTIs (Table 2).

Table 2.

Incidence of treatment-related early discontinuations and adverse events in the first year following INSTI initiation.

Incidence of early discontinuation due to adverse events
INSTI	Total N	Number of early DCs	Time at risk (p-y)	Incidence	Unadjusted IRR (95% CI)	P-value
RAL	26	0	139.1	0	0 (0; NA)	0.47
EVG/c	151	3	24.0	0.02	Ref	Ref
DTG	74	5	65.9	0.08	3.52 (0.84, 14.72)	0.07
BIC	80	0	73.8	0	0 (0; NA)	0.21
Incidence of treatment-related adverse events
INSTI	Total N	Number of AEs	Time at risk (p-y)	Incidence	Unadjusted IRR (95% CI)	P-value
RAL	26	11	24	0.46	0.64 (0.34, 1.19)	0.16
EVG/c	151	100	139.4	0.72	Ref	Ref
DTG	74	66	68.3	0.97	1.35 (0.99, 1.84)	0.06
BIC	80	65	73.8	0.88	1.23 (0.9, 1.68)	0.2

Note. Significant p-values (<0.05) have been bolded for ease of interpretation.

Abbreviations. INSTI, integrase strand transfer inhibitor; RAL, raltegravir; EVG/c, elvitegravir/cobicistat; DTG, dolutegravir; BIC, bictegravir, DCs, discontinuations; p-y, person-years; IRR, incidence rate ratio; AEs, adverse events; CI, confidence interval.

Within the first year, 7 on RAL experienced 11 treatment-related AEs (IR 0.46 PPY), 63 on EVG/c experienced 100 treatment-related AEs (IR 0.72 PPY), 37 on DTG experienced 66 treatment-related AEs (IR 0.97 PPY), and 34 on BIC patients 65 treatment-related AEs (IR of 0.88 PPY) (Table 2). Unadjusted IRRs demonstrated no significant difference in the incidence of early treatment-related AEs among the different INSTIs (Table 2).

Among RAL initiators, the most common treatment-related AEs were fatigue (3/7) and Gl upset (3/7). NPAEs occurred in 2/26. All treatment-related AEs were grade 1–2 except for 2 persons with Grade 3 transaminitis that was transient and resolved. Among EVG/c initiators, the most common treatment-related AEs were GI upset (35/63), and hypertriglyceridemia (25/63). NPAEs occurred in 13/151. All treatment-related AEs were Grade 1–2 except for 15 persons with Grade 3 laboratory abnormalities (8 with hypertriglyceridemia, 4 with hypercholesterolaemia, and 3 with transaminitis, the latter were transient and resolved). Among DTG initiators, the most common treatment-related AEs were insomnia (12/37) and GI upset (8/37). NPAEs occurred in 15/74. All treatment-related AEs were Grade 1–2 except for 1 person with Grade 4 transaminitis that was transient and resolved. Among BIC initiators, the most common treatment-related AEs were GI upset (9/34) and weight gain (9/34). NPAEs occurred in 4/80. All treatment-related AEs were Grade 1–2. There were no serious AEs documented during the study period.

Discussion

In our real-world cohort of treatment-naïve PLWH, discontinuations due to AEs occurred in ≤2%, however, other real-world analyses have observed a higher incidence of early treatment-related INSTI discontinuations (3–10%).^2–5,11 Though many of these included both treatment-naïve and -experienced PLWH, our cohort only included treatment-naïve PLWH. Prior data suggests that treatment-naïve PLWH may be more motivated to continue antiretrovirals despite experiencing tolerable AEs compared to PLWH switching therapy,¹² thus potentially explaining our results compared to other real-world studies.

Previous studies have demonstrated a higher incidence of early treatment-related discontinuations among those on EVG/c, typically due to gastrointestinal, general and renal-related AEs, with NPAE-related discontinuations occurring more frequently in those on DTG^5,11 In our study, we observed no significant difference in early treatment-related discontinuations among INSTIs, with zero early treatment-related discontinuations among BIC and RAL patients (Table 2). Though, DTG was associated with a non-significant increased incidence of early treatment-related AEs; only 5 on DTG discontinued early due to AEs. NPAEs were more common among those on DTG in our cohort; however, only 1 on DTG discontinued due to a NPAE.

Grade 3–4 AEs occurred infrequently and mostly consisted of fasting hypertriglyceridemia among those on EVG/c and transient transaminitis in 2 on RAL, 1 on EVG/c and 1 on DTG. Higher rates of hypertriglyceridemia have been previously described in association with EVG/c¹³; possible mechanisms include EVG-induced impairment of adipocyte function and increase in proinflammatory cytokines.^13,14 Notably, there were no Grade 3–4 AEs leading to INSTI discontinuation.

Study limitations include the retrospective nature of the analysis, the long duration of follow-up during which awareness of INSTI and other antiretroviral therapy (ART) side effects evolved, and the inability to control for different provider’s knowledge of the latter and how this potentially affected their determination of drug-relatedness. For example, ‘ongoing bone loss’ was documented for 1 patient as an AE related to EVG/c/F/TDF only 9 months after initiation, however, it is now well understood that ART-associated bone loss is more of a long-term consequence generally associated with at least 1–2 years of ART use.¹⁵ Other limitations were that data are from a single center limiting generalizability to other populations and that exclusion of RAL and DTG given in combination with other NRTIs may have resulted in selection bias. However, this is one of few real-world cohorts evaluating the safety and tolerability of BIC compared to other INSTIs, and our inclusion of only treatment-naïve PLWH initiating a TAF or TDF-based regimen allowed for a cleaner comparison of INSTIs compared to prior studies, which included PLWH on a variety of nucleoside backbones and often combined those treatment-naïve and -experienced.^3,5,11

In conclusion, we observed early treatment-related AEs in 43% initiating INSTIs, however these led to discontinuations in only 2%. Tolerability of BIC was favourable and similar to other INSTIs in this cohort.

Footnotes

Author note

Data were previously presented at IDWeek 2021, Abstract 889, September 29-October 3rd, 2021; Virtual. Dr. Charlotte-Paige Rolle has received research grants and honoraria from Gilead Sciences, ViiV Healthcare, and Janssen Infectious Diseases during the conduct of this study. Dr. Federico Hinestrosa has received honoraria from Gilead Sciences, Merck, and AbbVie during the conduct of this study. Dr. Edwin DeJesus has received honoraria from Gilead Sciences during the conduct of this study.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a research grant from Gilead Sciences, IN-US-985-5360 (PI: Rolle). The funder had no role in the study design; the collection, analysis and interpretation of data; the writing of the report; and in the decision to submit the article for publication.

ORCID iD

Charlotte-Paige Rolle

References

Kolakowska

Maresca

Collins

, et al. Update on adverse effects of HIV integrase inhibitors. Curr Treat Options Infect Dis 2019; 11(4): 372–387.

Yombi

. Dolutegravir neuropsychiatric adverse events: specific drug effect or class effect. AIDS Rev 2018; 20(1): 14–26.

Llibre

Montoliu

Miro

, et al. Discontinuation of dolutegravir, elvitegravir/cobicistat and raltegravir because of toxicity in a prospective cohort. HIV Med 2019; 20(3): 237–247.

Cuzin

Pugliese

Katlama

, et al. Integrase strand transfer inhibitors and neuropsychiatric adverse events in a large prospective cohort. J Antimicrob Chemother 2019; 74(3): 754–760.

Hoffmann

Welz

Sabranski

, et al. Higher rates of neuropsychiatric adverse events leading to dolutegravir discontinuation in women and older patients. HIV Med 2017; 18(1): 56–63.

Hoffmann

Llibre

. Neuropsychiatric adverse events with dolutegravir and other integrase strand transfer inhibitors. AIDS Rev 2019; 21(1): 4–10.

Mounzer

Brunet

Fusco

, et al. Advanced HIV infection in treatment-naive individuals: effectiveness and persistence of recommended 3-drug regimens. Open Forum Infect Dis 2022; 9(3): ofac018.

Ambrosioni

Rojas Lievano

Berrocal

, et al. Real-life experience with bictegravir/emtricitabine/tenofovir alafenamide in a large reference clinical centre. J Antimicrob Chemother, 2022. 77(4): p. 1133–1139.

Hoffmann

Schewe

Fenske

, et al. Short-term neuropsychiatric tolerability of bictegravir combined with emtricitabine/tenofovir alafenamide in clinical practice. Antivir Ther 2020; 25(2): 83–90.

10.

U.S. Department of Health and Human Services, NIOH . National Institute of Allergy and Infectious Diseases, Division of AIDS. Division of AIDS (DAIDS) Table for Grading the Severity of Adult and Pediatric Adverse Events, Corrected Version 2.1. Available from: https://rsc.niaid.nih.gov/sites/default/files/daidsgradingcorrectedv21.pdf

11.

Lepik

Yip

Ulloa

, et al. Adverse drug reactions to integrase strand transfer inhibitors. AIDS 2018; 32(7): 903–912.

12.

Palella

Chmiel

Moorman

, et al. Durability and predictors of success of highly active antiretroviral therapy for ambulatory HIV-infected patients. AIDS 2002; 16(12): 1617–1626.

13.

The RSG . Incidence of dyslipidemia in people with HIV who are treated with integrase inhibitors versus other antiretroviral agents. AIDS 2021; 35(6): 869–882.

14.

Moure

Domingo

Gallego-Escuredo

, et al. Impact of elvitegravir on human adipocytes: alterations in differentiation, gene expression and release of adipokines and cytokines. Antiviral Res, 2016. 132: p. 59–65.

15.

Moran

Weitzmann

Ofotokun

. Bone loss in HIV infection. Curr Treat Options Infect Dis 2017; 9(1): 52–67.