Differences in Antiretroviral Adherence Behaviors, Treatment Success, and Eligibility for Long-Acting Injectable Treatment between Patients Who Acquired HIV in Childhood vs. Those Who Acquired It in Adolescence/Early Adulthood

This study investigates the impact of the age at which HIV was acquired on adherence. There was no difference in adherence between patients who acquired HIV in childhood vs. those who acquired it in adolescence/early adulthood (83% vs. 90%; p = 0.24), but achievement of virological/immunological efficacy (78.8% vs. 93.5%, p = 0.02) was less likely in patients who had acquired HIV in childhood. On the basis of resistance, patients who acquired HIV in adolescence/early adulthood tended to be more eligible for cabotegravir/rilpivirine treatment (90.3% vs. 80.3%; p = 0.11).


Introduction
Despite simpler and more tolerable regimens for treating HIV infection, suboptimal adherence remains common, particularly among young patients [1][2][3][4][5][6][7]. Young adults may face multiple challenges, such as identity issues, unveiling, lack of a family model, and thinking in terms of short-term benefits [8,9]. This study aims to compare adherence depending on the timing of infection acquisition, i.e., infancy/childhood versus adolescence/early adulthood, to explore the causes of non-adherence and to compare the two groups in terms of the HIV viral load, resistance to antiretroviral therapies, and eligibility for injectable cabotegravir-rilpivirine.

Materials and Methods
This study included a retrospective chart review and a survey in order to identify factors of non-adherence from the point of view of the healthcare team, and that of the patient, respectively.

Retrospective Study
This study included all patients infected with HIV younger than 25 years of age, currently aged between 18 and 30 (as of 1 May 2021), and seen in three Canadian clinics in the Montreal area (Centre Hospitalier de l'Université de Montréal (CHUM), McGill University Health Center (MUHC), and Charles LeMoyne Hospital (CLMH)). Patients for whom no antiretroviral had been prescribed, or lacking minimal biological data in the medical file (HIV viral load and CD4 count) were excluded. The population was divided into two groups: patients with HIV diagnosed before the age of 10 (group 1), and patients diagnosed with HIV between the ages of 10 and 25 (group 2). Data were collected between 1 June 2021 and 27 August 2021, in Montreal. For the cross-sectional study, patients already receiving long-acting injectable antiretrovirals were excluded.
Data collected included adherence to antiretrovirals as systematically reported in the health care providers' notes, potential viral resistance determined by genotype testing, and the complexity of the antiretroviral regimen. Poor/insufficient adherence was defined as omitting to take 20% or more of monthly doses, based on the minimum threshold of 80% required for therapeutic adherence in the context of antiretrovirals [6]. We defined good immunological and virological efficacy when the viral load was below 200 copies per ml or undetectable, and when the CD4+ T-lymphocytes count was above 200 cells per mm 3 . Eligibility for cabotegravir/rilpivirine was determined on the basis of the absence of specific mutations, i.e., E138A/G/K/Q/R, K101P/E/Q, V179L, H221Y, F227C, M230I/L, and Y181C/I/V for rilpivirine and S147G, R263K, H51Y, and S153F/Y for cabotegravir, as defined by the Stanford University HIV Drug Resistance Database, and International Antiviral Society-USA [10,11]. In the absence of genotype, patients were assumed to be free of mutations of interest.

Survey
For the cross-sectional study, all patients included in the retrospective chart review were invited, by email or letter, to complete an electronic questionnaire in order to identify self-reported causes of non-adherence to antiretrovirals, except for those already on injectable treatment. With the author's permission, we used the Godin questionnaire, in English and French, which is validated for antiretroviral therapy (ART) adherence in the Canadian population [12]. It is a brief and simple self-reporting questionnaire evaluating the number of antiretroviral pills missed in the previous 7 days, social activities in the past week (leisure activities, going to bars, restaurants, etc.), and impact on ART adherence (Appendix A). In addition, we included a visual analog scale and some questions to explore the causes of non-adherence and the patient's interest in a long-acting injectable treatment. The RedCap ® software (survey function) was used to collect the data. An HTML link to the questionnaire and a QR-Code with the same function were included in the letter or email inviting participation in the study. To encourage participation, for those who used email for communication, the email was sent 3 times at 2-week intervals. For the others, a letter inviting participation was sent twice by postal mail. About half of the patients were also invited by phone to complete the questionnaire, as this was permitted by the ethics committee at only one of the three participating sites. Informed consent was signed electronically by answering the first question. Participants had the opportunity to enter a draw for a CAD 100 gift card. This study received multicentric approval by the CHUM institutional review and research ethics board.

Statistical Analysis
Groups were described using means, with a 95% confidence interval (95% CI), and numbers, with proportions, for continuous and categorical variables, respectively. The Mann-Whitney U test was used for a comparative analysis of continuous variables, while chi-square tests and Fisher's exact test were used for a comparative analysis of categorical variables. Analyses were performed using SPSS Statistics v.26 (IBM Corporation, Armonk, NY, USA).

Results
Out of 148 patients screened, 128 were included (CHUM = 29, MUHC = 92, and CLMH = 7). The reasons for exclusions were: diagnosis of HIV infection over the age of 25 (n = 14), inability to determine the date of diagnosis (n = 4), insufficient biological data in the medical file (n = 1), and absence of a prescribed antiretroviral (n = 1). In all, 66 patients were included in group 1 (HIV diagnosed before the age of 10) and 62 in group 2 (patients diagnosed between the ages of 10 and 25) ( Table 1).

Retrospective Study
HIV transmission was mainly perinatal in group 1 (95.5%) and sexual in group 2 (88.7%). No transmission through injection drug use was noted in the charts, but 10 patients had acquired HIV by another or unknown route.
Regarding the eligibility of patients for long-acting injectable cabotegravir/rilpivirine on the basis of resistance, there seemed to be more eligible patients in group 2, but this relation was not statistically significant (80.3% vs. 90.3%; p = 0.11). Nine patients (six in group 1 and three in group 2) had mutations that made them ineligible for treatment with cabotegravir/rilpivirine: E138A/K (n = 4), K101P/E/Q (n = 4), and Y181C (n = 1) ( Table 2).

Cross-Sectional Study
Four patients were excluded from the cross-sectional study because they were already receiving a long-acting injectable antiretroviral combination. In group 1, out of the 65 who received the questionnaire, 16 people responded (25%). In group 2, out of the 59 who received the questionnaire, 15 people responded (25%).
We did not find a statistically significant association between group and treatment adherence considering the mean number of missed pills over 2 days (0.25 (95% CI: 0-0.61) vs. 0.11 (0-0.33); p = 0.95), and the mean number of missed pills over 7 days (0.56 (95% CI: 0-1.18) vs. 0.20 (0-0.43); p = 0.57). There was, however, a significant difference between adherence noted in their files and self-reported adherence, since only 45% of the patients reported the same level of adherence as that assessed by a health professional (prescribing physician, clinical pharmacist, or clinical nurse).
Among patients with "optimal" chart adherence, 72% self-rated their adherence as such. Noticeably, all patients whose adherence was judged to be "poor or insufficient" considered their adherence to be "optimal" and reported no omissions over 7 days. Eleven patients considered their adherence to be optimal, although it was described as imperfect or insufficient by health professionals. Conversely, five patients considered that their treatment compliance was not optimal, while the healthcare professional considered it as optimal.
The patients included in group 1 generally reported more social activities, and among them, two said that these situations interfered with taking their antiretroviral treatment properly. Compared to three patients in group 2, five patients in group 1 reported that ART negatively impacted their quality of life. In both groups, the main self-reported cause of non-adherence was simple forgetfulness (59%). Patients also mentioned interference with social life (12%), lack of time to take the medication (10%), and the fear of a possible disclosure of their HIV-positive status (10%) as being etiologies of non-adherence. The semi-structured questionnaire identified additional causes of non-adherence as forgetting renewal at the pharmacy (n = 1), going outside the home without bringing along the pills (4%), and "Taking a daily treatment for a pathology imposed since birth" (mentioned by one patient). Another patient reported the size of the pills as an obstacle to adherence.
Regarding interest in long-acting ART, patients in group 2 were more likely (73.3%) to be interested in the injectable cabotegravir/rilpivirine combination than patients in group 1 (56.3%), but this difference was not significant (p = 0.32) ( Table 3).

Discussion
Our study has found an association between adherence and immune-virological efficacy. The relation between adherence and virological efficacy is well established in the literature [4,6,13]. Our data do not support the assumption that adherence is different depending on the age of seroconversion. Similarly, Vanthournout et al. [4] did not find a significant association between ART adherence, history of treatment failure, the clinical stage of the pathology, age at diagnosis disclosure, and the duration of taking ART. However, fewer patients with perinatally acquired HIV achieved immunological and virological controls, and their virus showed more resistance mutations, as expected considering the difference between the presumed mean duration of infection in the two groups. Consequently, ART regimens were significantly more complex in this group.
Clinicians see the new injectable therapies as an attractive option for improving adherence, particularly among the younger population, who often do not take other medications. Indeed, 65% of those surveyed expressed interest. However, compared with those who had acquired HIV later, there was a trend toward lower eligibility among those with perinatally acquired HIV, as their virus had more accumulated mutations. This is not surprising, as they had both longer exposure to antiretrovirals and exposure to older agents.
Adolescents and young adults living with HIV are exposed to periods of great change, particularly through puberty. Thus, it is necessary to consider the barriers to adherence related to developmental factors: physical, cognitive, social, emotional changes, combined with an emerging recognition of their sexual identity [8]. These barriers to adherence include in particular motivational barriers related to the acceptance of the disease, and factors related to the impact of potential stigmatization [8]. Social barriers to compliance are favored by chaotic and unstructured lifestyles (disorganization, drug addiction, etc.), material constraints, or even a lack of support from the patient's social environment (family and friends) [9]. In our study, more than half the respondents reported a negative impact of their antiretrovirals on their quality of life.
This study is not without limitations, especially with regard to the survey component. The 25% response rate to the questionnaire may have affected representativeness. Our study may have a response bias. For example, people who are more involved in their healthcare or with a less busy schedule may be more likely to respond. Potential participants were contacted multiple times, and, in our opinion, more solicitation would not have been acceptable. We do not believe that a written questionnaire would have improved the response rate, as our study population was under 30 years of age. Moreover, it would have posed greater confidentiality issues. In addition, in both groups, on average, higher ART adherence was noted when it was self-assessed using the online questionnaire, compared to adherence according to the notes in the medical file, probably showing a social desirability bias. Finally, due to the small sample, our study might have been underpowered to detect small differences between groups.
In conclusion, we did not find a difference in antiretroviral adherence between patients who acquired HIV in childhood vs. those who acquired it in adolescence/early adulthood. Regardless of the transmission mode, young people living with HIV need appropriate support and methods to reduce various adherence barriers, and to maximize their chances of obtaining a long-term undetectable viral load. Even though many young patients may be interested in long-acting injectable antivirals, it is important to look at previous genotypes to ensure eligibility.  • Go to a restaurant?

Wake-up/Breakfast/Morning Lunch/Afternoon Supper/Evening/Bedtime
• Go to a bar?
• Go to a party?
• Sleep away from home?
• Visit friend(s) or family member(s)?
• Receive a visit from friend(s) or family member(s)?
• Attend a meeting

4.
During the last 7 days, did one of the situations listed in Question 3 prevent you from taking all your antiretroviral pills?

5.
During the last 7 days, how many times in total, did you miss taking one or more of your antiretroviral pills? (If you haven't missed any, write down the number « 0 ») Long-acting antiretroviral treatments are now available: the medication should be monthly or every two months by a nurse, an injection in each buttock.

Number of antiretroviral pills that you missed…
11. If this treatment was possible for you, would you like to switch from a treatment taken by mouth to an injected treatment because it does not need to be taken every day?
Yes No 12. If not, for which reason(s)? 13. Do you think this treatment would improve your treatment adherence? (Treatment adherence = when you take the medicine correctly, as prescribed by your doctor) (many answers possible)

Yes
No, my compliance is already optimal No, because of the constraint of a monthly or every two months appointment No, because I'm afraid of the pain caused by the injection Other, please specify Reference question 1 to 5: Godin, G., Gagné, C., Naccache, H. Validation of a self-reported questionnaire assessing adherence to antiretroviral medication. AIDS Patient Care 2003, 17, 325-332.