Completion of isoniazid preventive therapy for latent tuberculosis infection among children and adolescents compared to adults living with HIV in Kinshasa, Democratic Republic of the Congo

Little is known about isoniazid preventive therapy (IPT) completion rates among children or adolescents compared to adults living with HIV in Kinshasa, Democratic Republic of the Congo (DRC).


INTRODUCTION
People living with HIV (PLHIV) are at high risk for tuberculosis (TB), and TB remains a major cause of death in PLHIV; thus, TB prevention in PLHIV is a priority [1,2].Early initiation of antiretroviral therapy (ART) for asymptomatic PLHIV has a potent TB preventive effect, with even more benefits for those with advanced immunodeficiency [3].The Democratic Republic of the Congo (DRC) has endorsed the World Health Organisation recommendations that all PLHIV initiate ART regardless of clinical stage or CD4 cell count, and this policy has been shown to provide a TB preventive benefit at the population level in high HIV prevalence settings [3].Furthermore, it has been shown that among PLHIV, 6-9 months of isoniazid preventive therapy (IPT) combined with ART significantly reduces the risk of TB and mortality, compared with ART alone [4][5][6][7][8].However, optimal IPT adherence (>80% of prescribed course doses) and completion are necessary to prevent the development of active TB from latently infected individuals [3].
There is an abundance of published data to show that IPT is effective in preventing active TB among PLHIV [9][10][11].However, its wide implementation for both children and adults in high TB-burden African countries has been limited by several structural factors, including poor health infrastructure and inefficient public health programming [10,12].Moreover, treatment completion among PLHIV who start IPT is poor.A meta-analysis of 70 cohorts including more than 94,000 PLHIV from high-and low-income countries found that only 33%-41% of PLHIV started and completed tuberculosis prevention therapy, most of which were IPT [13].
Per the 2022 Global TB Report, DRC is among the top 30 countries with the highest TB, TB/HIV, and multidrugresistant TB burdens globally [1].Gaps in IPT uptake and completion contribute to the high TB incidence and overall TB burden, including among PLHIV.A retrospective study in Kinshasa, DRC reports similar high IPT completion rates among adults and children on ART (88.2% vs. 86.2%,respectively, p > 0.05) [14].However, the study was conducted in two relatively large health facilities, limiting the generalizability of the findings.It is important to assess programmatic performance with respect to IPT among PLHIV from a more representative population, especially in countries with high burden of both TB and HIV, to strengthen evidence-informed improvement plans.
Furthermore, these assessments should be done separately for children and adults, given the limited available data from DRC as mentioned above.Also, reported provider-, caregiver-, or health system-related challenges for latent TB treatment in children include health providers' perceived inability to rule out active disease, fear of drug resistance, non-implementation of existing guidelines, poor completion with prolonged IPT with reliance on caregiver, and fear of side effects [15,16].In low-and middle-income countries (LMICs), the limited availability of paediatric formulations of TB prevention drugs exacerbates the challenges faced by prevention programs [15,17].
We performed a multi-year retrospective assessment of IPT completion among children and adolescents compared to adults living with HIV in Kinshasa, DRC.Given the challenges pertaining to TB prevention for paediatric patients outlined above, we hypothesized that IPT completion will be lower in children and adolescents compared to adults.Our results will help advance our understanding of age-related disparities in IPT completion and guide researchers and policymakers for future implementation science research in DRC.

Study design, population, and setting
We conducted a retrospective cohort analysis of data routinely collected from adults, adolescents, and children living with HIV enrolled in an FHI 360-implemented HIV care and treatment program in six health zones of Kinshasa, DRC.Our study used data from PLHIV who had initiated IPT between 2004 and 2020.To be included in the study, patients also had to meet the following criteria: (1) ARTnaïve or -experienced; (2) at least 6 months of follow-up data available.ART-experienced was defined as taking either a non-nucleoside reverse transcriptase inhibitor (NNRTI, e.g., efavirenz) or Integrase Strand Transfer Inhibitor (INSTI) (e.g., dolutegravir)-based regimen or other.Followup continued from initiation of ART until (a) loss to followup/transferred; or (b) death.
The primary outcome was the proportion of children and adolescents versus adults who completed IPT.Our database included a variable indicating whether patients had completed IPT according to the DRC's national TB treatment guidelines.Before 2020 (our study period), the guideline recommended a 6-month isoniazid only regimen for IPT.After 2020, the recommendation changed to 3-month daily isoniazid + rifampicin or 3-month once weekly isoniazid + rifapentine regimens [18].PLHIV with confirmed or suspected latent TB were eligible to receive IPT. Pyridoxine was routinely administered daily during treatment with INH during IPT, in line with the standard of care in the DRC.Per the national TB treatment guidelines, latent TB is confirmed when patients have a positive tuberculin skin test (TST) or Interferon Gamma Release Assay (IGRA) and no symptoms or signs of active TB.Latent TB is suspected when PLHIV meet the following conditions: (1) Patients aged <12 months living with HIV who were exposed to TB and do not have symptoms or signs of active TB; (2) Patients aged ≥12 months living with HIV who are unlikely to have active TB based on symptoms or signs and no history of TB exposure; (3) All children living with HIV who completed TB treatment; (4) Adolescents and adults living with HIV with a positive or unknown TST or IGRA, regardless of the level of immunosuppression [18].Of note, IGRA testing was not available in DRC before 2020.
Our study population was treated within a U.S. President's Emergency Plan for AIDS Relief (PEPFAR) program using the 95-95-95 framework and UNAIDS indicators.We classified the population age groups using the PEPFAR categories in an effort to standardise our indicators.Our primary analyses compared IPT non-completion in children (<11 years), adolescents (defined as ages 11 through 19 years, inclusive) and adults (age ≥20 years), based on age at ART initiation.Secondary outcome included all-cause mortality.Other covariates in the analysis included sex, WHO clinical stage at ART initiation, CD4 count, duration on ART, and health zone.

Data management
Data were retrieved by trained program staff directly from patient files at the health facilities using an electronic national register called Tier.Net.Data collection and entry in Tier.Net was monitored to ensure compliance with quality and integrity standards.Data were validated at the health facility and then sent to the health, provincial, and national levels.Access to Tier.Net and ability to enter and edit patient data were restricted to data managers and monitoring and evaluation staff with different permission levels to maintain data integrity.The secure server hosting the Tier.Net data, located at the Ministry of Health, includes cloud backup, password protection, and verification and validation tools.We checked the database for previous or repeated IPT and duplicate records, including duplicates from cohort aging, using the patients' ID numbers and found none.

IPT completion
IPT completion was checked monthly or bimonthly at clinic visits by a health care worker who inquired about missed doses and reviewed the number of INH tablets dispensed to the individual from the pharmacy refill records.Patients were counselled about the importance of taking INH daily and completing the full course of treatment.

Statistical analysis
Continuous variables were expressed as median with interquartile range since they were non-normal data.Categorical variables were expressed using frequencies and percentages.Fisher's exact test was used to assess the association between the categorical variables and age groups, and the Wilcoxon rank-sum test to assess the equality of the median of the continuous variables between children, adolescents, and adult patients.Log-binomial regression was used to assess significant association between demographic and clinical factors and IPT non-completion.Factors associated with IPT non-completion at p < 0.15 in unadjusted univariable log-binomial regression were included in a multivariable model to identify predictor variables associated with IPT non-completion [19].Adjusted risk ratios and their 95% CIs were used as a measure of association.Factors with p < 0.05 were considered significantly associated with IPT noncompletion.The Kaplan-Meier (KM) curves were used to assess differences between time to non-completion and time to death among the three age categories.All statistical analyses were performed using Stata (V.16, Stata Corp, College Station, Texas, USA) and R (V.4.1.0,R Core Team) with R Studio (V.1.4.1, R Studio Team) statistical software.1).Furthermore, children had a significantly shorter follow-up duration on IPT than adults (median 6.8 months [IQR: 6-9] vs. 9 [8.2-9.2],p < 0.001).The results of the KM analyses showed that the probability of death was not statistically different between completion and non-completion groups (p = 0.150) (Figure 1), but the probability of death among the three age categories was statistically significant, p < 0.001 (Figure 2).Children were less likely to die as compared to adolescents and adults.In a Cox regression multivariable analysis, after adjusting for gender, health zone, and calendar time; children remained at less risk of death (aHR: 0.42; 95% CI: 0.22-0.81,p < 0.001).

Factors associated with IPT non-completion in unadjusted and adjusted multivariable binomial regression
Results of log-binomial regression analysis are shown in Table 2; compared to adults, children were 1.29 (95% CI: 1.11-1.50,p = 0.001) times and adolescents 1.32 (95% CI: 1.20-1.45;p < 0.001) times at risk of IPT noncompletion in the unadjusted univariable model.Adolescents were 32% more likely not to complete IPT treatment than adults (RR: 1.32 (95% CI: 1.20-1.45);p < 0.001).The risk of IPT non-completion was also higher among males than females (RR: 1.14; 95% CI: 1.06-1.22).Also, patients with CD4 count less than 200 cells/μL were at higher risk of IPT non-completion than those with CD4 greater than 500 cells/μL (RR 1.20; 95% CI: 1.04-1.38,p = 0.011).Patients on an Efavirenz-based regimen had 26% lower risk of not completing IPT than those on a dolutegravir-based regimen (RR: 0.74; 95% CI: 0.67-0.82),p < 0.001).Similarly lower risks of IPT non-completion were seen for patients on a nevirapine-based regimen and other regimen versus a dolutegravir-based regimen.The risk of IPT non-completion was also significantly lower for patients residing in Limete, Matete, Mont-Ngafula, and Ndjili health zones compared to those living in Kinshasa health zone.
After adjustment for known and available confounding factors (e.g., sex, baseline CD4+ count, residence, ART regimen), there were no significant differences by age group and by gender.PLHIV from Lingwala had 1.26 (95% CI: 1.05-1.51,p = 0.012) times the risk of IPT non-completion than those in Kinshasa health zone, while those in Limete, Lingwala, Matete, and Ndjili had significantly lower risks of non-completion as compared to those in Kinshasa.Lastly, patients who were initiated on ART before 2010 (RR: 0.82; 95% CI: 0.73-0.93),p = 0.002) and between 2010 and 2015 (RR: 0.88; 95% CI: 0.79-0.99),p = 0.027) were less likely to not complete IPT as compared to those who initiated ART between 2015 and 2020 (Table 2).

DISCUSSION
We documented an overall sub-optimal IPT completion of 61.5% among PLHIV in this observational cohort from DRC.In unadjusted analyses, children and adolescents had significantly higher risks of not completing IPT compared to adults.After adjustment, IPT non-completion was independently associated only with health zone of residence and type of ART regimen.Also, our time-to-event analysis showed that survival was associated with IPT completion.However, this finding did not reach statistical significance (log-rank test p = 0.15), likely because we were underpowered for the secondary outcome of all-cause mortality (Type-2 error).
A previous study reported IPT completion rates of 86.6% in children and 88.2% in adults in Kinshasa, DRC [14], which is considerably higher than the levels we documented.The difference could be explained by the fact that the above study was limited to two health facilities, but our sample was larger and probably more diverse.However, the difference could also indicate a decline in IPT completion over the years.This aligns with our results indicating that the risks of IPT non-completion increased significantly after 2015.Further investigations are needed to understand the reasons for this decline.
There is limited published data on IPT completion in children/adolescents compared to adults [20], but some previous research in Sub-Saharan Africa has reported differences in IPT completion between children and adults [14,21].In this analysis, the risk of non-completion was higher among children and adolescents than adults in unadjusted analysis; this finding partially supports our main hypothesis.Age differences in IPT completion are driven by a host of factors impacting IPT access for younger compared to older PLHIV [15][16][17].
The higher risk of IPT non-completion among patients on dolutegravir versus other regimen merits further investigation.Significant drug interactions were documented among patients taking isoniazid + rifapentine and dolutegravir [22], a factor that could explain the high level of IPT non-completion among patients on both drugs in our cohort.However, our database does not specify the IPT regimen taken by patients in this analysis, limiting our ability to test this hypothesis.The differences in IPT completion by health zones highlighted in our findings could be indicative of regional differences in quality of care and other factors of the health system that impact IPT compliance.Our study has several important clinical and policy implications for the DRC HIV national program.First, the low completion rate of IPT is indicative of gaps in the follow-up of PLHIV under IPT in the DRC.There is an urgent need for further research to investigate these gaps, particularly the individual, structural, and social determinants of sub-optimal IPT completion in this population.Second, latent TB infection treatment regimens that are shorter and equally effective, such as a weekly dose of rifapentine and isoniazid for 3 months (3HP); a daily dose of rifampicin plus isoniazid for 3 months (3HR); a daily dose of rifapentine plus isoniazid for 1 month (1HP); or a daily dose of rifampicin for 4 months (4R) [1, 23,24], recently introduced by the DRC's National TB Program, should be scaled up to improve treatment compliance.This effort should follow clear guidelines to improve compliance, including those around the prevention and management of drug interactions and side effects resulting from the use of the TB preventive treatments.Third, in settings where the above newer and shorter regimens are unavailable, implementation of targeted interventions to improve IPT completion is needed.Finally, increased funding for children-and adolescent-focused programs, de-coupled from general population care and treatment programs, should be considered to help intensify the scaleup of evidence-based interventions to improve outcomes in this vulnerable population.
Although informative, our study has several limitations.Retrospective analysis of routine data collected under programmatic conditions is vulnerable to missing data.However, in our study missing data appears to have occurred randomly.Individuals with missing data for specific explanatory or outcome variables of interest did not differ from those without missing data in terms of demographic characteristics.Our dataset was not originally designed as a comprehensive research tool and therefore was not structured to capture some important information, such as the specific reasons for IPT non-completion or cause of death.IPT adherence and completion were assessed in the program clinics by self-report and review of refill information from pharmacy dispensing records.This may overestimate the degree of adherence and treatment completion and is a major limitation of our study.Additionally, several other predictors of IPT non-completion, such as stigma due to HIV co-infection, other opportunistic infections, and low educational level, were not included in our dataset [20,25].Our dataset did not include the date at which any of the individuals treated with IPT subsequently developed active TB.This limitation does not allow us to determine whether patients developed TB before or after completion/noncompletion.Furthermore, our study was underpowered for the secondary outcome of all-cause mortality.Finally, our results are based on data from individuals treated with IPT in several health zones in Kinshasa, DRC and may not be generalizable to other settings.Further studies are needed to get a fuller understanding of the factors underlying IPT non-completion and related age disparities.
In summary, the overall sub-optimal IPT completion rate in adults as well as children/adolescents in this setting is of great concern.We call for future prospective studies to determine specific barriers to IPT completion in this population as well as implementation science, including the development of cost-effective and more patient-friendly interventions (such as shorter regimens and increasing availability of paediatric formulations of drugs for TB preventive therapy) to improve IPT completion in LMICs.

FF
I G U R E 1 Kaplan-Meier of time-to-death stratified by IPT completion versus non-completion.I G U R E 2 Kaplan-Meier of time-to-death stratified by age group.

Table 1 .
Adolescents and children were less likely to be female compared to adults (431/804, 53.6% and 216/429, 50.3%, respectively, vs. 7523/10458, 71.9%).The Lingwala Health Zone had the highest proportion adolescent participants (51.5%) in the study, and the Ndjili Health Zone had the highest proportions of children and adults (28.2% and 26.7%, respectively).Compared to adults, children had a shorter follow-up duration on ART (34 months [IQR:17-72] vs. 44 months [IQR:19-87], p < 0.001).Furthermore, children were less likely and T A B L E 1 Demographics and clinical characteristics of the study population by age categories (children, adolescents and adults).
a Missing percentages were calculated out of the total column and not included in the tests done.b Other ART regimens are any other combination regimens that are not dolutegravir-based, efavirenz-based, or nevirapine-based.c p-value use the adults as comparision group.TROPICAL MEDICINE & INTERNATIONAL HEALTH adolescents more likely than adults to initiate ART with HIV stage IV (15% and 35.5%, respectively, vs. 25.9%,p < 0.001), and the CD4 cell count at ART initiation was considerably lower in adolescents (26 cells/μL [IQR:0.17-462])than in adults (278 cells/μL [IQR:150-449]), p < 0.001.Children and adolescents were also less likely than adults to be on dolutegravir-based ART (27.5% and 63.6%, respectively, vs. 65.3%, p < 0.001 for each individually vs. adults).Among those who were initiated on IPT, 5625 out of 11,691 PLHIV had IPT completion outcome results, and an overall 61.5% completion rate was documented.Compared to adults, children and adolescents were less likely to complete IPT [104/199 (52.3%) and 268/525 (51.0%), respectively, vs. 3085/4901 (62.9%)] (Table T A B L E 2 Factors associated with IPT non-completion in unadjusted and adjusted multivariable binomial regression.
a Adjusted for age, sex, health zone, WHO stage at ART initiation, CD4 count at ART initiation, ART regimen, and calendar time.TROPICAL MEDICINE & INTERNATIONAL HEALTH