First and Second Line Highly Active Anti-Retroviral Therapy Failure in HIV Infected Nigerian Children at University of Abuja Teaching Hospital Gwagwalada, Nigeria

Globally, over 3.4 million children were living with HIV at the end of 2011, 91% of whom are in sub-Saharan Africa [1]. Highly active antiretroviral therapy (HAART) has markedly reduced the morbidity and mortality of patients with HIV/AIDS [2,3] and has a major role of achieving and maintaining undetectable viral load (VL) of <400 copies/ml after 6 months on therapy [4]. While the number of children receiving HAART continues to increase [5], access to optimal laboratory monitoring with VL testing remains limited. For this reason, most health care providers in resource limited settings make decisions about commencement of these drugs and subsequent switch over to 2^nd line regimen by the World Health Organization (WHO) immunologic and clinical guidelines without VL measurement [6]-[8], thus leaving HIV-infected children in those environment at risk of unrecognized virologic failure and the subsequent development of antiretroviral resistance. Switch to 3^rd line medication is practically non-existence in most areas were routine VL monitoring is not place.

The WHO [6]-[8], and Nigerian National Guidelines for Pediatric HIV and AIDS Treatment and Care [4], defined virological failure as VL not suppressed to undetectable levels (<400 copies/ml) after 6 months on antiretroviral therapy (ART), or a persistent increase in VL > 1000 copies/ml following a period of adequate suppression. The immunological indication for treatment failure is the development or return of CD4 cell count to <200cells/mm³ or % CD4+ of < 10% for a child > 2years to < 5 years of age or CD4 cell count of <100cells/ mm³ for a child 5 years of age or older, or return of CD4 cell count to pre-therapy baseline level, or count of below 50% decline from on-therapy CD4 cell peak level [4,6]-[8]. The clinical indication for treatment failure is the development of either new or recurrent WHO stages 3 or 4 events in a child on therapy [4,6]-[8]. The virological failure occurs as first event, followed by immunological and clinical failure usually months or years afterwards [9]. If patients do not achieve an undetectable VL, viral replication ensures, this often leads to development of spontaneous random mutation in HIV genome [9]. The most concern finding is that the WHO immunologic criteria for treatment failure detected virologic failure in 1 of 16 children (6%) [9]. In adults, the criteria have been shown to be insensitive measures for virologic outcomes [10,11], and a high prevalence of ARV resistance mutations has been documented in HIV-infected adults by the time they progress to meet the WHO criteria for ARV therapy switch [12]. A recent report showed same similarity in pediatric WHO guidelines, where it has been found to offer an insensitive measure of virologic failure [13]. This was confirmed in Ugandan study by Ruel et al. [9], and additionally demonstrated that the HIV-infected African children with ongoing undetected virologic failure develop ARV-mutations that could compromise future therapeutic options.

Immunologic and virologic measures are the state of the art for monitoring ART and should be made available in all settings where HIV infection is treated. In resource-rich countries, the standard of care for monitoring treatment in HIV-infected children is the routine laboratory monitoring of CD4+ T cell percentage or count, and VL [14]. In USA, CD4+ T cell percentage or count and VL are measured at the time of diagnosis of HIV infection and at least every 3-4 months thereafter [15]. Due to the lack of accessible, costly technical equipment, complex technology, expensive reagents, poor laboratory infrastructure, and prohibitive maintenance cost, this test is not routinely available in many resource-limited settings [16]. The available commercial assays for VL determination costs between $50 and $100 per test making it unaffordable to many centers in resource-limited countries [18]-[20]. Failure to 1^st line ART has been documented in 13% of HIV-infected Uganda children [9], 16% in South Africa [21], 16% in Thailand [22], 33% in Kenya [23], 44% in Mali [24], 9.1% in another South African study [25], and 0.8% from a multi-center study in Africa and Asia [26]. Adherence to ARV therapy has a major role of achieving good viral suppression on HIV patients. For patients failing 2^nd line therapy, treatment options are largely non-existent in most resource limited settings. The current WHO guidelines provide some guidance for treatment in the case of 2^nd line failure, but financial constraints will limit the adoption of 3^rd line options in countries with limited resources. Nigeria is one of the high HIV-burden countries in Africa, and has up to date made no provision for ART therapy beyond 2^nd line in its national guidelines, and no provision for routine use of VL in monitoring patients on HAART [4]. Thus, there is a need to document the rate/time of failure to 1^st and 2^nd line regimen using WHO immunological and clinical failure. Such information is needed to evaluate the effectiveness of WHO immunological and clinical criteria, and provide baseline information on children failing these drugs. Such information is needed for optimal ARV management in the population.

It was an observational descriptive retrospective study, conducted at Pediatric Outpatient Special Treatment Clinic (POSTC) for HIV/AIDS children at the University of Abuja Teaching Hospital (UATH), Gwagwalada from January 2006 to June 2015. POSTC is an out-patient clinical service where HIV infected children and exposed babies are followed up for treatment and monitoring. It has consulting rooms for the doctors, the nurses, and adherence counselors. Record clerks, pharmacists, and nutritionists are also at their disposal on week days (Monday-Friday, from 7.30 am to 4 pm.). UATH is a 350 bed capacity referral hospital, sub-serving the people of Federal Capital Territory (FCT) Abuja and five neighbouring states. Is one of the first centers to start offering free HIV/AIDS services in the country, through the President Emergency Plan for AIDs Relief (PEPFAR) since 2005.

The subjects were pediatric HIV infected patients' ≤18 years diagnosed by either by serological method or by deoxy-ribonucleic acid (DNA) polymerase chain reaction (PCR) test, started on 1^st line ARV therapy and later switched over to 2^nd line drugs. They all met WHO immunological and clinical criteria for commencement on 1^st line ARV therapy, and were started on them according to the national guideline [4]. Viral load is not routinely done in our center before commencement on 1^st line ARV therapy; it is only done at suspected immunological and clinical failure to 1^st line medications, and most recently one year after the commencement of 2^nd line medications or when there is immunological or clinical indication of non or slow response to 2^nd line medication. The medical folders of all the patients switched to 2^nd line ARV therapy were retrieved from the medical information department of the health institution and required information collected. They include: age of the patient at the commencement on 2^nd line drugs, sex, type of 1^st line and 2^nd line regimen started, CD4 cell count at initiation of 1^st line, CD4 cell count at subsequent periods of 6 months until the time of switch over to 2^nd line drug and thereafter, their VL before and after the switch to the 2^nd line regimen, and yearly subsequently. Their weight was also collected as their CD4 cell count. Also collected was adherence level as recorded by adherence counselors in their medical folders. Self-reporting by the parents/ caregiver of the patients, or the patients themselves in cases of adolescents, and keeping of appointment in the health facility were the methods used by adherence counselors in determining the level of adherence in our health institution, while the physicians uses weight, CD4 cell count and clinical parameters in following up their patients. Other methods such as pill count, and pharmacy refill were used by the pharmacists, while other methods eg mobile phone technologies, and electronic monitoring were not in use in our health facility. CD4 cell count was measured using automated Partec Cyflow easy count kit (Partec code no. 05-8401 Western Germany), VL measurement was with (Roche Smp /prep /cobs Taqman 96, USA), and Seca beam weighing scale accurate to the nearest 0.01kg was used for measuring their weight.

Ethical clearance was obtained from the Ethics Committee of the health institution before the commencement of the study. Data analysis was conducted using SPSS version 16.0 that produced frequencies, percentages, means, and standard deviations. The tests for associations and differences were done by student t-test, and statistical significance set at p value <0.05.

Table 1 depicts the characteristics of the study population at initiation of 1^st line HAART. A total of 82/514 (15.9%) patients, 48(58.5%) males and 34(41.5%) females were switched to 2^nd line medications, while 6/82 (7.3%), with 3/6(50.0%) males and 3/6 (50.0%) females required 3^rd line medication during the nine years review period. The average switch over rate for 2^nd line drugs was 9 persons/year, and 2 persons per every 3 years for those requiring 3nd line medications. Median switch over time for 2^nd line drugs was 31.3 months (IQR 26.4-36.2), and 42.6 months (IQR 40.1-44.3) from the time of switching to 2^nd line drug for those requiring 3^rd line regimen. Majority `of patients (90.4%) switched to 2^nd line drugs and those requiring 3^rd line medication were within the age bracket of 9 to 18 years, none was an infant and only 1(1.2%) under five. Their mean age, weight and CD4 cell count of the 82 subjects before commencement on 1^st line HAART were 11.3±2.1years, 18.5+5.3kg, and 101.58+29.32 cells/mm³, while their CD4 cell count and VL before 2^nd line HAART were 109.09±28.37 cells/mm³, and 180,887.29 ±54,628 copies/ml respectively. The 1^st line ARV drugs used were: 36(43.9%) for combination of AZT + 3TC + NVP, 27(32.9%) for D4T + 3TC + NVP, 19(21.2%) for D4T + 3TC + EFV, while the 2^nd line ARV drugs used were: 49 (59.8%) for ABC + didanosine (ddl) + LP/r, 24 (29.3%) for TDF + 3TC + LP/r, and 9(11.0%) for ABC + 3TC + LP/r, (data not shown).

CD4 cell count, and viral load of 6 subjects requiring 3^rd line HAART

Table 2 showed the CD4 cell count, VL and outcome of the 6 subjects requiring 3^rd line medications. There were 3/6 (50.0%) males with median switch over time of 42.6 months (IQR 41.1-43.3) from the start of 2^nd line medication. Their mean CD4 cell count was 89.42±28.67 cell/mm³, their VL which was also 237,337.5±64,619 copies/ml which has increased to over 590 folds at the time of immunological failure that occurred at mean period of 4.3 years of commencing 2st line HAART. With intensification of adherence there was marked reduction in VL of 2 (33.3%) subjects, 4/6 (63.3%) died inspite of improvement in adherence and non-availability/affordability of 3^rd line ARV drugs. It was also noted that the three deaths from failed 2^nd line regimen received tenofovir, and died from HIVAN (HIV Associated Nephropathy), chronic renal failure, and HIV encephalopathy. Unfortunately resistance testing was not done in any of the patients because of non-availability and accessibility to laboratory services.

Pattern of CD4 cell count, weight, and viral load while on 1^st and 2^nd line HAART

Table 3 shows the pattern of CD4 cell count, weight and viral load changes during the 9 year review period while on 1^st and 2^nd line HAART for the 82 subjects that failed 1^st line ARV therapy. The CD4 cell count doubled about the 6th month of initiation of 1^st line HAART, tripled/ peaked by the 1^st year before the gradual decline, and return to base line level between the 3^rd-4th years. Though the mean pre 1^st line HAART weight of 82 patients that failed 1^st line drugs was 18.5±5.3 kg below the 5th percentile for their age using NCHS growth chart [27], it however increased at initiation of HAART, but never showed a downward trend even when the CD4 cell count had started declining and returning to baseline level at the 3^rd and 4th year on medication. It is worth noting that at the time of immunological failure, the VL has increased to more than 450 folds above the 400 copies/ml recommended undetectable viral level after 6 months on ARV therapy, and this occurred at about 3.5 years after commencing 1^st line HAART.

CD4 cell count, VL and weight Vs adherence level before 2^nd Line HAART

Table 4 shows CD4 cell count, VL, and weight of patients before switch to 2^nd line HAART, and their adherence levels. There was statistical significant difference in the mean values of CD4 cell count and weight for those with adherence of ≥95% and those with ≤ 95% (p values were 0.024 for CD4 cell count, and 0.038 for weight). However no difference was seen in their viral load, p value 0.103. The mean VL of 187,676.72 +32,963.2 copies/ml of those with adherence level of ≥95% in this study was found to be over 450 fold greater than the recommended 400 copies/ml for adequate viral suppression when adherence is ≥95%.

The present study showed switch over rate to 2^nd line ARV drugs of 15.9% in our health institution over a 9 year review period. This was similar to the reported rate in some other resource limited settings were HIV RNA VL assays was not routinely used in the monitoring of patients on ART; 13% in Uganda children [9], 16% in South Africa [21], 16% in Thailand [22]], but fair much lower than 33% in Kenya [23], and 44% in Mali [24]. It was however higher than 9.1% reported from another South African study [25], and 0.8% from a multi-center study in Africa and Asia [26]. In the present study also, over 7.0% children had failed 2^nd line medication, they had a mean VL of 237,337.5±64,619 copies/ml with over 590 fold increase thus posing a serious cause for concern in view of non provision/availability/affordability of 3^rd line medications in Nigerian guideline and in most resource limited settings. This finding underscores the need for VL monitoring in patients on ARV therapy for early detection of virological failure and early switch of patients to 2^nd line drugs in order to prevent persistent viraemia which is a good predictor for rapid development of ARV resistance mutations [9]. Third line ARV medications should also be made available in resource limited countries as growing number of children have started failing 2^nd line medications. This is necessary to prevent untimely death and spread of résistance virus. If we continue relying on WHO recommended immunological failure before switch over to 2^nd line medications, which occurred at a VL of 180,480.29±35,303 copies/ml in this study, it might infer that many of the children were VL is not routinely use for monitoring of patients on ARV therapy will be left with undetected persistent viraemia and subsequent development of multiple TAMs resistant viral strain. This will not only jeopardize the activity of thymidine analog ARVs use in second line regimen, but foster other mutations that will threaten the efficacy and reliability of second line regimens [9]. Resistant testing at this point is also supposed to be done prior to switch to either 2^nd or 3^rd medication for this will help in the detection of resistance virus, and guide in the choice of 2^nd or 3^rd medication. In view of the above findings in this study, though the sample was not large enough for inference, the recommended Nigerian guideline for monitoring of patients on ART should possibly be reviewed to add VL measurement before commencement of 1^st or 2^nd line ARV therapy and 6monthly thereafter for monitoring purpose. Viral resistance testing should also be included in the national guideline before switch to 2^nd or 3^rd line medication. Inclusion of VL in the monitoring process, and resistance testing will not only ensure failure identification in patients at least within one year of starting ARV therapy but also will guide in the choice of 2^nd or 3^rd line. In the light of the above evidence, using VL monitoring and resistance testing will be cost-effective intervention that has proved successful in industrialized countries, and suppressed VL means less HIV transmission, and successful ART.

The clinical indication for treatment failure is the development of either new or recurrent WHO stages 3 or 4 events in a child on therapy [4,6]-[8]. With such development, one expects reduction in weight of the patients from chronic opportunistic infections. In the present study, even when immunological failure has occurred, weight reduction was not remarkable. Studies have shown that virological failure occur first, followed by immunological and clinical failure usually months or years afterwards [9,10]. Using only clinical failure and possibly weight loss which usually occurs months or years after immunological failures as the only criteria for switch to a 2^nd line regimen as practiced in some resource limited settings also implies that the HIV-infected children in these areas will be left with ongoing undetected virologic failure. The maintenance of such children and adults on these failed drugs for years before waiting for WHO stage 3 or 4 symptoms to appear with accompanying weight reduction is a great opportunity for accumulation of resistance mutation of virus that will not only compromise future therapeutic options, but also risk of transmission of these mutant virus in future.

WHO defines adherence as the extent to which a person's behaviour in taking medication, following a diet, and/or executing lifestyle changes, corresponds with agreed recommendations from a health care provider [28]. It remains the major tool in achieving VL suppression, prevent drug resistance, and maintain the health of HIV infected patients. An adherence level of ≥ 95% is the accepted level for achieving good viral suppression [29,30], and missing drugs for as little as 2 days in a month can lead to an increase in VL, and development of mutant viral strains [31,32]. Such high level of adherence is required because of the rapid replication and mutation rate of HIV [33]. Adherence is the second strongest predictor of progression to AIDS and death in a person living with HIV, the first being the CD4 count [34]. Patient self-reporting is one of the most common methods used in assessing adherence in resource poor settings [35]. It measures adherence via the use of self-administered or interviewer-administered questionnaires and interviews [35]. The reliability and validity of this method depend on patients or caregiver/parents memory recall, honesty, and social desirability [35,36]. The other methods include: pharmacy refill records, pill count, VL monitoring, CD4 cell count, appointment keeping, electronic monitoring, and mobile phone technologies [28,30],[35,36]. In this study, keeping of appointment with the clinic, self-reporting and CD4 cell count were majorly used by the adherence counselors in our health institution in the assessment of adherence level of the patients. There was statistical significant difference in CD4 cell count and weight of patients having good adherence of ≥ 95% and those with poor adherence of ≤ 95% in this study. This is similar with findings from other studies [35,36], but having no significant difference in VL in this two groups is in contrast to finding elsewhere [35,36], and could be because VL was not routinely used in the monitoring of patients in our study. The mean VL of 187,676.72 +32,963.2 copies/ml of those with adherence level of ≥95% was found to be over 450 fold greater than the recommended 400 copies/ml when adequate viral suppression occur with adherence of ≥95%. This finding in this study buttresses the inadequacy of using CD4 cell count and other adherence monitoring tools without VL in the monitoring/assessing of patients on ART. Because virology failure occur first, months or years before immunological/clinical failure, using other adherence monitoring tools without VL means that those failing ART will be detected much more latter inspite of their been recorded as having good adherence or doing well on their ARV drugs.

The importance of adherence in maintaining good viral suppression should be stressed tirelessly, before the introduction of the first line drugs as well as during every following visit. Adherence in adolescents has been reported to be poorer than in younger children, and older adults in both resourced rich and limited settings [18]. Reasons include adolescents' risk taking behavior, the stage of development when thought and reasoning is not fully developed, and regimen fatigue [18,38]. As perinatally infected HIV children who have been on ARVs grow into adolescence, adherence was found to decrease because of long period on medication [18,19],[36]-[38]. Some studies have reported adherence rates for long term prescription medications to be between 50–75% [19]. In the present study, adolescents between 9-18 years (90.4%) were switched to 2^nd line medications, thus bringing into focus the issue of adherence to medications among adolescent, and the need to find a lasting solution to this reoccurring decimal. Introduction of Direct observational therapy (DOTs) strategies should be considered as a very important option in mitigation of this ugly trend in adolescents.

WHO immunological and clinical criteria were found not to be adequate in determining when to switch patients to 2^nd line and 3^rd line medications. Six monthly VL monitoring and resistance testing should be introduced in Nigerian guideline for monitoring of HIV children on ART for better therapeutic and pharmaco-economic outcomes. Third line medication should also be made available for them, and intensification of adherence in adolescents is required to minimize the risk of non-adherence in this age group.

Owing to the small sample upon which the results were based, it is recommended that similar investigations be carried out in other centers in the country.