The objective of ART in the treatment of HIV disease is to inhibit viral replication and achieve and sustain viral suppression while promoting immune reconstitution and clinical improvement. There is increasing evidence from sub-Saharan Africa that under routine program conditions, HIV-infected children receiving ART can achieve improved immune function, sustained virological suppression and good clinical outcomes [25
]. Ours is the first study to report the early and medium-term immuno-virological outcomes of ART-naïve HIV-1 and HIV-2 infected children starting ART in the concentrated low prevalence setting of The Gambia – West Africa.
Despite a median baseline CD4 T cell percent of 13%, which indicates that the HIV-1 infected children in this cohort were at an advanced stage of immunosuppression at entry, the overall median CD4 T cell percent increased to 23.3% and 27% at six and 12
months respectively, with a ‘plateau’ at 18 and 24
months. By this measure, the MRC paediatric cohort demonstrated an immediate and significant response to therapy. These observed early- and medium-term responses are comparable to reports from a review of 28 paediatric cohorts in Sub-Saharan Africa which showed that CD4% increased within the first year of treatment with a plateau after 12 to 18
months of treatment [25
]. When stratified by age however, children
years of age at the start of ART in our cohort achieved a sustained increase in CD4% until 30
months of treatment, after which a decline in the median change was observed. This could be explained partially by the anticipated decrease in CD4% with age during the first 5
years of life [26
]. Likewise, children
years at start of ART had an initial increase in absolute CD4 count up to 12
months followed by a more blunted CD4 T cell recovery subsequently. These significant early immunological recovery followed by an apparent mid-term ‘plateau’ in CD4 T cell recovery which we observed in our cohort is comparable to reports in other settings [27
The proportion of HIV-1-infected children who achieved an undetectable viral load, defined as HIV RNA <100 copies/mL, was 63.2% at six months. By 12- and 24- months of therapy, 55.6% and 36.4% of the children in the cohort respectively had undetectable viral loads. However, the proportion increased to 58% at 36
months. Possible explanations for this include the fact that some children who failed therapy were switched on to second-line regimens as stated in the results section, as well as the effect of reinforced adherence counselling. Second-line therapy was however frequently delayed as a result of shortages of appropriate medicines and formulations.
Overall, 62.5% of all deaths in our cohort occurred within the first 90
days, with these early mortality predominantly among those with HIV-1 infection and aged <2
years. This is similar to reports elsewhere in the sub-region where early mortality (i.e. in the first 90
days of ART) is substantially higher compared to later months [29
]. The estimated 12
months survival probability on ART among the HIV-1-infected children of 89.9% falls within range of 12
month survival probability reported from the review of paediatric cohorts in sub-Sahara Africa mentioned earlier [25
]. Our findings show sustained improvement in survival experiences which is independent of the baseline CD4+
T cell percentage, therefore reinforcing the obvious benefit of ART in a routine program setting in a resource-limited setting and the need to further scale-up access to ART in children.
We found a very poor outcome among the six HIV-2 infected children after a median of 23
weeks on combination antiretroviral therapy, with three out of six being lost to follow-up and two dying. A review of the available laboratory parameters for treatment monitoring similarly showed poor immunologic and virologic responses in almost all of those children. A couple of factors could have contributed to the poor clinical and immuno-virologic outcomes we observed among the HIV-2 infected children in this report. First, all the children were at an advanced disease stage at diagnosis of their HIV-2 infection, and initiated antiretroviral therapy at a point when they were severely immunocompromised. Given the chronicity of HIV-2 infection, the long duration of infection before these children became eligible for ART at the same CD4 count threshold as HIV-1 infected children might have contributed to the poor immunological recovery. Secondly, despite the progress made in scaling-up ART in West Africa, treatment of HIV-2 infection still remains a challenge and the optimal treatment option for HIV-2 infection is still a subject of several ongoing research studies focussing only on adults. While available data from adult HIV-2 cohorts suggests a sub-optimal response to some of the protease inhibitors (PIs) commonly used in the West African sub-region [33
], very little is known about HIV-2 therapeutic approach in children. However, from our report, it is difficult to conclude that a sub-optimal response to LPV/r-based regimen that we use might have contributed to the observed poor outcome in view of the small number of cases.
Baseline co-morbidities such as tuberculosis, malnutrition, helminthiasis, anaemia as well as ART toxicity or immune reconstitution inflammatory syndrome (IRIS) particularly in those with advanced immunosuppression at ART initiation are well recognized risk factors for poor outcomes on antiretroviral therapy including early mortality [31
]. Most HIV-infected children in resource-limited settings present at an advanced stage of disease with marked immunosuppression mostly because of widespread lack of laboratory capacity for diagnosis of HIV-infection in children less than 18
months of age, who still harbour maternal HIV-antibodies. Although many of the children in our cohort were children of HIV-infected adults enrolled in a natural history cohort who would have benefitted from early diagnosis, a good number however, were referred from the MRC outpatients’ clinic and hospital ward as well as other health-care facilities due to chronic ill health and at an advanced HIV disease clinic stage.
Adherence to ART is another important determinant of viral suppression and treatment outcome [37
], and in children is strongly linked to the availability of treatment support. Although a limitation of this study is that adherence as measured by pill counting and care-giver/self report was not specifically analysed as a variable, the observed immuno-virological responses were objective evidence of the levels of adherence to medications within the cohort. Reddi et al.
] reported that having at least one HIV-positive caregiver provides a protective effect against mortality when compared with care-givers who were untested or HIV-negative especially where this care-giver receives treatment at the same site as such a care-giver. They hypothesize that such a care-giver may be more knowledgeable about symptom management and disease progression and therefore able to provide more informed treatment support. The majority of the children on ART in our cohort had both parents alive and at least one HIV-positive parent; we however did not analyze the effect of care-giver HIV-status on treatment outcomes.
As with most other paediatric HIV treatment programmes in resource-limited settings [38
], we resorted to using adult, generic fixed-dose combination tablets cut in half or quartered. In the absence of pill-cutters this usually resulted in unequal-sized fractions and could have inadvertently contributed to inappropriate, particularly sub-therapeutic, drug doses particularly with the LPV/r gel capsules [40
As more countries in the sub-region scale-up access to ART for HIV-infected children, paediatric treatment outcomes data can help guide programme implementation. The level of care provided at our HIV clinic is however not representative of paediatric HIV care in West Africa as patients had the advantage of frequent and regular care provided by highly qualified physicians and counsellors at essentially no cost to the patients, with routine viral load and CD4 cell counts and -percent monitoring. This, coupled with the small size of our paediatric cohort makes it difficult to generalize our findings to other settings in the sub-region. We however experienced similar programmatic challenges such as lack of appropriate paediatric formulations of almost all the antiretroviral drugs especially second-line drugs, and stock-out of at least one drug annually. Over 60% of the children on our cohort were severely immunocompromised at treatment initiation, and the need for multiple pre-ART counselling sessions and prior approval by the national Eligibility committee further contributed to delays in starting treatment which may have put these patients at increased risk of poor immunological recovery, treatment failure and mortality.