Antiretroviral therapy outcomes among adolescents and young adults in a Tertiary hospital in Kenya

Background Limited data is available on the treatment outcomes of HIV infected adolescents and young adults (AYA) in sub-Saharan Africa. HIV-infected adolescents and young adults (AYA) are at high risk of developing antiretroviral treatment failure. Objective To determine the clinical, immunological and virologic outcomes of AYA at a tertiary hospital in Kenya. Methodology A longitudinal study was conducted among AYA age 10–24 years attending Kenyatta National Hospital comprehensive care center. Clinical data was abstracted from electronic medical records for study participants with at least 6 months of follow-up using a structured data abstraction sheet. Results A total of 250 AYA age 10 to 24 years were included. The median age was 16 years. The median CD4 cell count was 650.6 cells/mm3 (IQR 350.7–884.0). More than half (60.6%) of AYA had a CD4 cell count higher than 500 cells/mm3. Overall, 76.9% of AYA had achieved viral suppression (viral load <1000 copies/ml). There was a significant increase in virologic failure with higher age and late adolescents and young adults were more likely to have a viral load > 1000 copies/ml p<0.012. Conclusion The overall virologic suppression in this cohort of AYA was sub-optimal. Both immunological and virologic outcomes were worse among late adolescents (18–19 years) and young adults (20–24 years).


Introduction
Adolescents and young adults (AYA) represent an increasing proportion of people living with HIV in the world. Over 1.6 million adolescents age 10-19 and 5 million young people age 15-24 are living with HIV 1 . Out of these 1.5 million (89%) HIV infected adoles-cents live in sub-Saharan Africa. Adolescents have low levels of testing, linkage and retention to care and poor adherence to antiretroviral treatment 2,3 . Consequently clinical, immunological and virologic outcomes in adolescents have consistently shown poor results 4 . The number of AIDS-related deaths among adolescents tripled between the year 2000 and 2015 5 . Adolescents and young adults are at high risk of developing treatment failure and drug resistance, which has dire consequences in low resource settings such as Kenya, where there are limited choices of antiretroviral regimens. Across sub-Saharan Africa, many children who were infected at birth are now becoming adolescents. The number of adolescents and young adults on antiretroviral treatment has been increasing reflecting treatment success of HIV infected children 6 .
AYA have unique characteristics that predisposes them to treatment failure and drug resistance such as poor insight into the disease and treatment 7 . This makes close follow up and frequent assessment of treatment outcomes necessary. The existing literature on improving treatment outcomes is mainly focused on children and adults, and very limited in regards to AYA. Despite major advances in treatment and access to antiretroviral treatment, little is known about treatment outcomes in this age group. The study assessed the treatment outcomes of AYA actively on follow up at a tertiary hospital setting to determine their clinical, immunological and virologic outcomes.

Study Design
A longitudinal study was conducted as part of an interventional study to determine the effect of mobile phone calls on retention to care among adolescents and young adults on care and treatment at Kenyatta National Hospital Comprehensive Care Clinic between January 2017 and November 2017. A review of medical records was performed to determine the clinical, immunological and virologic outcomes in AYA 10 to 24 years on active follow.

Study Site
The study was carried out at Kenyatta National Hospital Comprehensive Care Centre which at the time of the study had 9,915 active HIV patients on care and treatment, among them 362 adolescents age 15-19 years and 693 children between 0-14 years. Patient's data is entered into an electronic medical records database (IQ care) during routine clinic visits. Data on clinical variables, outcomes, treatment regimens and laboratory records were abstracted from the electronic medical records data which is routinely updated at each clinic visit. AYA are typically seen in the clinic monthly or every three months for clinical monitoring, with a bi-annual viral loads and CD4 cell count monitoring.

Study Outcomes
The main study outcomes were improvement in WHO staging, immunological status and virologic failure. Immunological failure was defined as CD4 cell count < 500 cells/mm 3 and virologic failure as a viral load of > 1000 copies/ml.

IRB Approval
The study was approved by Kenyatta National Hospital/University of Nairobi Ethical Review Committee (ERC) and University of Washington Institutional Review Board.

Data Management and Statistical Analysis
Data were collected and entered into a password-protected MS-Access database. The database was set within required limits and restrictions to reduce errors. Exploratory data analysis was carried out to identify extreme values and inconsistencies. On completion of the data entry, the data was exported into SPSS Version 21.0 for analysis. Continuous values were summarized as means, medians, standard deviation and interquartile ranges. Discrete values were summarized using frequencies and percentages. Categorical variables were compared using chi-square test and Fishers exact test. Continuous variables were analyzed using Student t-test and Mann Whitney test by comparing means. A P value <0.05 was considered significant. The associations between social demographics characteristics and treatment outcomes were measured by Odds Ratios (OR).

Treatment History
All AYA in this study were on antiretroviral therapy (ART). The majority of 173 (69.2%) were on a 1 st line antiretroviral regimen, while 77 (30.8%) were on their 2 nd line regimen as per the national guidelines in Kenya. The most frequent ART regimen combination at ART initiation comprised of two nucleoside reverse transcriptase inhibitors (NRTIs); [zidovudine or abacavir or tenofovir or stavudine] plus lamivudine and one non-nucleoside reverse transcriptase inhibitor (NNR-TI), nevirapine or efavirenz. Only a third ( 33 (13.2%) were initiated on a protease inhibitor-based regimen. On the current ART regimen; majority 171 (68.4%) were on two nucleoside reverse transcriptase inhibitors (NRTIs); [zidovudine or abacavir or tenofovir plus lam-ivudine] and one non-nucleoside reverse transcriptase inhibitor (NNRTI), [nevirapine or efavirenz]. Seventy seven (30.8%) were on a protease inhibitor-based regimen while two (0.8%) were on an integrase inhibitor-based regimen. (This is shown in Table 1). A review of records revealed that 85 (34.0%) of study participants had a history of ART substitution.   Table 3).  Table 4). On factors associated with CD4 cell count at 6 months, early adolescents were more likely to have a CD4 cell count more than 500 cells/mm 3 (OR 7.3 (2.7-19.9 p<0.001) while young adults were more likely to have a CD4 cell count of less than 500 cells/mm3 (OR 0.2 (0.1-0.5 p<0.001). (The results are shown in Table 5).

Virologic Failure
At a reference point of 6 months ( 186 ) 76.9% of adolescents and young adults had a viral load <1000 copies/ml, while 93 (39.4%) had a viral load of >1000 copies/ml. There was a significant likelihood of virologic failure with higher age; with young adults more likely to have a viral load > 1000 copies/ml p<0.012. (The results are shown in Table 4). Although not statistically significant early adolescents were twice as likely to have virologic suppression compared to other age categories in the study OR 2.0 (0.8-4.8) p=0.126. (The results are shown in Table 6).

Discussion
Our study findings demonstrated that 76.9% of AYA in this cohort had achieved viral load suppression. Our study cohort viral load suppression was lower than the national average reported rate of 88.4% for patients 0-64 years 8 . The viral load suppression is slightly higher than the 67.1% found by KENPHIA among children aged 0-14 years but much lower than the 90.6% among those 15-64 years of age. The relatively low rates of viral load suppression in AYA has been revealed in several other studies 9,10 . In our study, young adults were less likely to achieve adequate virologic suppression, followed by late adolescents. Majority of early adolescents (87.9%) had achieved virologic suppression.
Median CD4 cell count was 650 cells/mm 3 , with 60% of adolescents having a CD4 cell count > 500 cells/ mm 3 . This immunological response is consistent to study findings in Asia and South Africa where the median CD4 cell count was fond to be 657cells/mm 3 and 686 cells/mm 3 respectively 11,12 . A similar trend of higher virologic suppression among early adolescents compared to mid, late adolescents and young adults was observed with CD4 cell count, where majority of early adolescents had a better immunological outcome compared to the other age groups. This might be explained by parental oversight to treatment for early adolescents. More than half of late adolescents and young adults had CD4 cell count of < 500 cells/mm 3 . This trend of a decrease in CD4 cell count and increased chances of virologic failure with advancing age was observed in a study by Mburugu and colleagues (2016) where mid and late adolescents had poor immunologic, virologic outcomes and poor adherence to treatment 13 . Similarly Geoffrey and colleagues 14 documented in a baseline characteristics study in South Africa that mid and late adolescents had lower median CD4 cell counts 17 . In the same study viral suppression was poorest in mid adolescents at any point time of ART. The poor immunological and virologic outcomes could be explained by less adherence support during transition to adult HIV programs leading to poor adherence to ART, less support from parents and guardians and complexities associated with the developmental stages of adolescence and young adulthood. At our last assessment at 6 months, a larger proportion of AYA in our study ( 79.4% ) were in WHO clinical stage I, while 11.5%,7.8% and 1.2 % were in WHO clinical stage II, III and IV respectively. Thus majority of the adolescents were stable with absence of recent opportunistic infections. However on further analysis, during the six months study period, there was significant deterioration in the clinical staging from WHO clinical stage I to IV. The finding of having a majority of HIV infected adolescents being stable was also found in a longitudinal study by Allison et al where they found 85% of adolescents in there cohort to be stable. 18

Limitations
This was a cross-sectional study that utilized routinely collected clinical outcome data from an electronic medical records system as part of a clinical trial. The study involved a single site which is a tertiary teaching hospital thus the results may have limited generalizability. However this study provides an insight into the treatment outcomes of HIV infected adolescents and identifies the age categories that need special attention during care and treatment for that age group.

Conclusion
Viral load suppression in this cohort of adolescents was suboptimal with late adolescents (18-19 years) and young adults (20-24 years) being more likely to have virologic failure. Interventions in this program should targets adolescents and young adults with special focus to late adolescents and young adults to improve treatment outcomes.