Cristina de Jesús Herrera-Castillo ·Angélica Maldonado-Rodríguez ·Othón Rojas-Montes ·Ana María Cevallos ·José Guillermo Vázquez-Rosales ·Rosalía Lira

According to World Health Organization (WHO) estimates,in 2020,there were 1.8 million human immunodeficiency virus (HIV)-infected children below 15 years old living with HIV and 150,000 children who acquired the infection[1].Every year,110,000 children die as a consequence of acquired immune deficiency syndrome-related causes [2].As in adults,optimal management of infected children depends on early diagnosis and treatment with antiretroviral drugs based on their viral genotype.

Since its establishment in 1983,the Mexican governmental program for epidemiological surveillance of HIV infection has detected a total of 5404 children less than 15 years old (52% males and 48% females) corresponding to 1.7% of all HIV-registered diagnosis [3].However,it is estimated that approximately half of infected pregnant women do not know their HIV status and therefore do not receive treatment to prevent vertical transmission [4].Furthermore,the prevalence of transmitted drug resistance (TDR) in Mexico is unknown.In the only published study on the topic,none of the nine Mexican children studied had TDRs [5],while the overall prevalence in Latin America varies between 0 and 27% [6].In this retrospective study,we evaluated the presence of drug resistance mutations (DRMs) in two naive groups of children with vertically transmitted HIV infection and analyzed the data according to their response to treatment.

In this study,samples from 32 children with HIV infection were selected from a plasma bank at Pediatric Hospital“Silvestre Frenk Freund”,Mexican Social Security Institute(IMSS).Children were sent for evaluation at the infectology department with diagnosis of HIV infection.None had received antiretroviral treatment.Children with confirmed HIV infection were treated at the clinic and followed up until the age of 18 years,and were tested before treatment and subsequently every 3—6 months to monitor response to treatment.Initial and subsequent plasma samples were routinely kept at -70 °C in case further analysis was required.The period of sample collection was from 1999 to 2011.We selected children with two different outcomes:17 with a virological response (VR) and 15 with virological failure(VF) to treatment (Fig.1).In case there was not enough plasma left for samples in a given child,one criterion for the selection of the cases was that the naive samples were available in the plasma bank with a viral load greater than 5000 copies/mL,and for the children with VF that at least 3 follow-up samples were available.According to guidelines in pediatric HIV infection [7],VR was defined as the suppression of HIV RNA to undetectable levels after 4—6 months of antiretroviral therapy and VF was defined as the detection of HIV RNA after 4—6 months of antiretroviral therapy [8].The study was approved by the IMSS Commission of Scientific Research of the Paediatrics Hospital Mexico City,Mexico in accordance with Good Clinical Practice (R-2011-3603-221).

Fig.1 Flowchart of sample selection process.HIV human immunodeficiency virus,VL viral load,VR virological response,VF virological failure

The RNA-HIV viral load (VL) of the plasma samples was detected using the Amplicor HIV-1 Monitor kit(HIV RNA detection cut off 200 copies/mL) to ensure the integrity of the viral RNA.For genotyping,plasma RNA was purified using the High Pure Viral RNA kit (Roche Molecular Systems).For amplification of the HIV-1

pol

gene an in-house protocol previously reported was performed with modifications [9].A 1084-bp fragment of the HIV-1

pol

gene encoding amino acids 6—99 of the protease region and codons 1—251 of the reverse transcriptase region was amplified and sequenced.DRMs were analyzed using the Stanford genotyping drug resistance interpretation algorithm v8.4 (http://sierr a2.stanford.edu/sierra/servl et/JSier ra).Mann—Whitney

U

test was used to determine differences regarding age,CD4 counts and VL,a

P

value of ＜0.05 was considered significant.Fisher’s exact test was used to determine difference regarding gender.A twotailed

P

value of ＜0.05 was considered significant.A total of 32 children were studied,17 with VR and 15 with VF (Table 1).Sex was the only demographic characteristic for which a statistically significant difference was found with more females in the VF group and more males in the VR group (Fisher’s exact test

P

=0.03).There were no differences in age,CD4 counts,or VL between the groups.All children were infected with HIV-1 subtype B according to the genotype (GenBank accession numbers:OM329944-48,OM366233-37,OM489471,OM418561,OM700184-199,OM801891-93).We identified 7 children with DRMs in their naive samples among our 32 (22%) studied cases,4 in the group of VR and 3 in the group of VF (Table 2).The drug resistance genotype of at least three follow-up samples per child with virological failure was analyzed.The length of follow-up varied from 6 to 46 months,with a median of 15 months.Of the 12 antiretroviral naive children without DRMs,3 (25%) did not develop any DRMs and 8 (75%) acquired them after treatment.In general,DRMs were maintained once they appeared (Table 3).The most frequent nucleoside reverse transcriptase inhibitor (NRTI)DRMs occurred at codon position 184 (

n

=7),followed by mutation at position 215 (

n

=3).The most frequent protease inhibitors (PI) mutations occurred at codon positions 54(

n

=5) and 82 (

n

=4) (Table 3).

Table 1 Demographics and characteristics of the study population at presentation

virological response, virological failure,viral load,25th—75th percentiles,standard deviation, Centers for Disease Control and Prevention.a Clinical data from one patient with VF were not available for staging; b not available in 2 VR and 3 VF patients; c Student’s test not significant

Table 2 Description of the HIV-1 drug resistance mutations identified in naive samples of vertically infected children with virological response and virological failure

human immunodeficiency virus,virological response, virological failure, viral load,protease inhibitors, nucleoside reverse transcriptase inhibitor, non-nucleoside reverse transcriptase inhibitor, male, female,diagnosis

Table 3 Genotypes and clinical information from children with virological failure

Table 3 (continued)

ritonavir,saquinavir,zidovudine, lamivudine, lopinavir, lopinavir/ritonavir,didanosine,stavudine,efavirenz,nevirapine,viral load, protease inhibitors,drug resistance mutation, nucleoside reverse transcriptase inhibitor, non-nucleoside reverse transcriptase inhibitor. Samples with vertically transmitted drug resistance

There is a lack of information about vertical transmitted drug resistance mutations in children from Mexico.In the only study that analyzed the presence of TDR in Mexican children,no DRMs were detected and the appearance of new mutations was not analyzed [5].In our retrospective study,we searched for the presence of drug resistance associated mutations in a cohort of naive vertically infected HIV-1 children in a tertiary care hospital.None of these children had a history of maternal screening for HIV infection but the possibility that their mothers had received anti-retroviral treatment cannot be discarded.We identified a frequency of 22% of the children with TDR mutations,which is higher than the 6.2% prevalence reported in adults [6,10].We cannot rule out the possible bias in the selection of the samples.However,these DRMs were probably vertically transmitted as none had been previously treated with antiretrovirals,and the age of most of them was 24 months old or less.We found that most of the children with VF were male.However,the analysis of 172 children with vertically transmitted HIV infection that responded to treatment,for which records were complete and available in our hospital,showed no significant difference between males and females (

P

=0.36,data not shown),suggesting that probably the significant difference in the VF group was due to sample selection bias.This is consistent with published data where gender has not been associated with treatment failure [11,12].

Analysis of data according to the response to treatment showed that the DRM profile of the children with TDR that did not respond to treatment had more mutations that confer resistance to more classes of drugs than children with VR.PI mutations were only identified in children with VF,raising the possibility that these mutations increase the risk of treatment failure.During the follow-up,of the 12 children with VF that did not have DRMs in the naive sample,75% acquired DRMs.The DRMs identified did not have a clear pattern of acquisition,except that NRTI DRMs appeared faster than PI mutations as has been previously reported [13,14].Three children did not acquire new DRMs (ID-334,343,350),probably due to lack of adherence to treatment.This was a frequent problem in children due to the poorly palatable regimens,complicated administration techniques,a higher pill burden,and frequent dosing changes according to age [15].

Antiretroviral treatment schemes were similar in adults and children.Recommended first line regimens consisted in either two NRTIs (abacavir,lamivudine,tenofovir,emtricitabine,zidovudine) and one PI (lopinavir/ritonavir or saquinavir,fosamprenavi,atazanavir) or two NRTIs and one non-nucleoside reverse transcriptase inhibitor (efavirenzor nevirapine).It is not surprising that the mutations found in naive children reflect the treatment regimens used at that time.

Genotyping was not generally performed and therefore,treatment was either not modified or changed as new effi-ciency studies or different drugs became available.Therefore,most of these children with transmitted DRMs did not receive optimum treatment according to their genotype.Identification of specific genotypes would have helped physicians to differentiate between appearance of DRMs and lack of adherence to treatment,a frequent problem in this population [15].

In Mexico,mother-to-child transmission (MTCT)has significantly decreased due to the establishment of institutional programs for detection and treatment in pregnant women [16].In our hospital,an unpublished evaluation of the success of the local MTCT program identified a reduction of transmission to 2% (Guillermo Vazquez-Rosales,personal communication).Unfortunately,MTCT programs are not widely available in all clinical settings with poor areas without access to this program.

In conclusion,this is the first time that TDR has been reported in children in Mexico.In our study,22% of naive children with vertically transmitted HIV infection had DRMs.Empirical treatment of these children was associated with an increased risk of treatment failure.The use of low cost in-house drug resistance genotyping assay could significantly improve the management of these children.

Acknowledgements

The authors are grateful to Ma.Teresa Alvarez-Mu?oz,for the historical bank of plasmas.

Author contributions

HCCdJ contributed to conceptualization,funding acquisition,project administration,formal analysis,reviewing and editing.MRA and RMO contributed to methodology,investigation,formal analysis,data curation,reviewing and editing.CAM contributed to formal analysis,data curation,writing of the original draft,reviewing and editing.VRG contributed to conceptualization,formal analysis,and data curation,writing of the original draft,reviewing and editing.LR contributed to conceptualization,project administration,methodology,supervision,visualization,formal analysis,writing of the original draft,reviewing and editing.All the authors approved the final version of the manuscript.

Funding

This work was supported by Health Research Funding,Mexican Social Security Institute (IMSS) (grant number:FIS/IMSS/PROT/G11-2/1030).

Data availability

The data of the nucleotide sequences generated and analyzed during the current study are available in the GenBank NCBI.All genotyping and clinical data analyzed during this study are included in this article.

Declarations

Ethical approval

The study was approved by the Ethics Committee of the Pediatrics Hospital of the Mexican Social Security Institute (IMSS)in Mexico City,Mexico,in accordance with Good Clinical Practice(R-2011-3603-221).

Conflict of interest

No financial benefits have been received from any party related directly to the subject of this article.The authors have no conflict of interest to declare.