Levels of bone markers in a population of infants exposed in utero and during breastfeeding to tenofovir within an Option B+ programme in Malawi

Abstract

Introduction

The initiation of lifelong ART in all HIV-infected pregnant women, irrespective of CD4 count, with maintained infant breastfeeding up to 12 months (Option B+, usually with a fixed combination of tenofovir, lamivudine and efavirenz), represents a WHO-recommended strategy for the prevention of mother-to-child transmission of HIV,¹ which is being adopted by a rapidly growing number of countries, and it is likely to represent standard of care in the near future.^1,2 It is commonly accepted that the potential risks of this approach are clearly outweighed by its multiple benefits, represented by simple implementation and management, immediate reversal of immunological decline, maintenance of the positive effects of breastfeeding and prevention of HIV transmission to uninfected partners.² Nonetheless, there are still some uncertainties concerning the safety of the drugs being used, particularly for the potential effects that intrauterine and early postnatal exposure through breastfeeding might have on infant health. Registry data and systematic reviews have not confirmed the early concerns of an increased risk of birth defects associated with prenatal exposure to efavirenz,^3,4 and most of the attention is therefore now focused on the potential effects of pre- and postnatal tenofovir exposure on infant growth, renal function and bone metabolism. Tenofovir is known to affect bone turnover, bone mineral density (BMD) and renal function in adult HIV-infected patients,^5–8 has a significant transplacental passage⁹ and is found in meconium and in breast milk, although at low concentrations.^10–12 It is therefore necessary to investigate potential effects of the drug on bone mineralization and renal function in infants exposed during pregnancy and breastfeeding. Most of the published studies have shown no impact of prenatal exposure to tenofovir on infant growth,^13–15 although recent studies showed that newborns exposed to tenofovir during pregnancy had slightly reduced values in height and head circumference measurements at 1 year of age¹⁶ and a lower mineral content in the first month of life.¹⁷ Overall, the potential role of tenofovir in inducing bone abnormalities responsible for growth impairment is still uncertain and no data are available on bone metabolism markers in infants exposed to this drug during both pregnancy and breastfeeding. Given the rapidly growing number of infants exposed to tenofovir in the prevention of mother-to-child transmission programmes, and the need to collect more information on this subject, we investigated bone metabolism markers in the first year of life in infants from mothers who received tenofovir, lamivudine and efavirenz during pregnancy and 12 months of breastfeeding in a national Option B+ programme in Sub-Saharan Africa.

Methods

We used data from an observational study conducted in Malawi within the Drug Resource Enhancement against AIDS Malnutrition (DREAM) programme of the Community of S. Egidio, an Italian faith-based non-governmental organization.

The study (tenofovir/efavirenz study) aimed to evaluate safety aspects of the Option B+ programme that was introduced on a national basis in Malawi in 2011.¹⁸ In short, according to this programme, all pregnant women, irrespective of CD4 levels, started tenofovir, lamivudine and efavirenz in pregnancy and were assigned to continue indefinitely the ART after delivery, with a 2 year recommended duration of breastfeeding. All infants (tenofovir-exposed and -unexposed) received nevirapine from birth to 6 weeks (10 or 15 mg once daily, based on birthweight below or above 2500 g, respectively).

Ethics

Study approval was obtained by the National Health Sciences Research Committee of Malawi in 2011 (approval no. 905) and women provided informed consent for themselves and their babies.

Study population

The population analysed included all the infants who had at least one serum sample collected and available for analysis of bone markers. Infant height (recumbent), weight and main laboratory data (haemoglobin, creatinine, AST, ALT) were routinely collected within the DREAM programme at 1 month, 6 months and 12 months of infant age. Based on infant age, gender, height and weight, Z-scores of weight for age, height for age, weight for height and BMI for age were calculated, using as reference the WHO Child Growth standards.¹⁹

Laboratory evaluations

Serum samples were collected for all infants in the morning, stored at −80°C soon after collection, shipped in dry ice and thawed only at the time of testing. Serum levels of bone-specific alkaline phosphatase (BAP; marker of bone formation) and of C-terminal telopeptide of type I collagen (CTX; marker of bone resorption) were centrally analysed in batches at the Istituto Superiore di Sanità in Rome, using commercial immunoassays: MicroVue BAP EIA (Quidel Corporation, San Diego, CA, USA) and Serum CrossLaps ELISA (IDS Ltd, Boldon, Tyne & Wear, UK). Sensitivity thresholds for the two assays were 0.7 U/L for BAP and 0.02 ng/mL for CTX; inter-assay variation (CV%) was 5.0%–7.6% for BAP and 2.5%–10.9% for CTX, and intra-assay variation was 3.9%–5.8% for BAP and 1.7%–3.0% for CTX. For quantification of BAP values, samples were diluted 10-fold to adjust for higher concentrations found in infants and compared with reference values of 62–146 U/L for male and of 84–160 U/L for female infants (age: 0–2 years, mean ± 2SD), according to Tsai et al.²⁰ and to MicroVue BAP EIA Product Information.²¹ For quantification of CTX, samples were analysed undiluted and compared with reference values of 0.202–2.311 ng/mL according to Crofton et al.²² When paired samples of the same subject at different times were available, these were usually analysed in a single assay run. Following analysis of bone markers in children exposed to tenofovir, a small number of random samples from tenofovir-unexposed newborns from the same clinical centre was also analysed for bone markers to have a control group of tenofovir-unexposed children of the same age and from the same setting. Briefly, these children were also followed within the DREAM programme, but were born to mothers receiving a triple drug combination of either zidovudine or stavudine plus lamivudine and nevirapine from week 25 of pregnancy, continued until 6 months post-partum or indefinitely if their CD4 cell count was <350/mm³.²³

Data analysis

The general characteristics of the population studied were summarized as medians with IQR, means with standard deviations, and percentages, and the distribution of variables was tested for normality using the Kolgomorov–Smirnov test. Qualitative variables were compared using the χ² test or the Fisher test, and quantitative variables using Student's t-test or the Mann–Whitney U-test, according to the characteristic of the distribution. Pearson's correlation coefficient was used to evaluate correlations between quantitative variables. P < 0.05 was considered statistically significant. All the analyses were performed using SPSS software, version 22 (released 2013; IBM Corp, Armonk, NY, USA).

Results

Overall, 136 tenofovir-exposed infants (female 68, male 68) from 133 mothers (three twin pregnancies) were analysed. The general characteristics of infants and mothers are reported in Table 1. Most mothers were ART-naive and received a median of 16 weeks (IQR 10–19) of ART during pregnancy. A high proportion of infants showed significant growth restriction, particularly with respect to height for age, which was <10th percentile in almost half of the cases at 12 months. With respect to bone markers, no infant had at either timepoint CTX values above the upper reference limit of 2.311 ng/mL (range at 6 months: 0.10–1.13 ng/mL, range at 12 months 0.09–1.85 ng/mL). At 6 and 12 months, respectively, 106 of 114 infants (93%) and 70 of 72 infants (97.2%) had levels of BAP above the upper reference limit for the age range. No differences were found between infants with normal or elevated BAP levels at 6 months with respect to BMI, weight for height, height for age and infant linear growth between 6 and 12 months (data not shown). Infant BAP levels were substantially unchanged between 6 and 12 months (362 and 354 U/L, respectively, P = 0.759), while CTX levels increased significantly, although always within normal range (from 0.55 ng/L at 6 months to 0.84 ng/L at 12 months, P < 0.001). The two bone markers showed no significant differences by infant gender (Table 2) and were not associated with growth restriction (height for age Z-score or weight for age Z-score <10th percentile at 6 or 12 months) (Table 3) or to height, height for age, height increase between 6 and 12 months (Table 4). Duration of maternal ART in pregnancy, however, showed a significant positive correlation with height and height for age at 6 and 12 months (Table 4). In the small control group of tenofovir-unexposed children (n = 40), mean levels of bone markers at 6 months were similar to those of tenofovir-exposed children of the same age (CTX: 0.62 versus 0.55 ng/mL, P = 0.122; BAP: 384 versus 362 U/L, P = 0.631).

Discussion

We evaluated a population of infants from Malawian mothers enrolled in an Option B+ programme for the prevention of mother-to-child transmission of HIV, reporting for the first time, to our knowledge, bone marker levels in children postnatally exposed through breastfeeding to tenofovir-based regimens until 12 months of life. The main limitations of this study are represented by the lack of a control group of HIV-unexposed infants from the same population, by the absence of reference values for bone markers in African infants and by the impossibility to perform a more detailed evaluation of other potentially relevant variables, such as vitamin D levels and nutritional intakes.²⁴ All the above limitations clearly derive from general problems of scarcity of data available for low-resource settings (particularly for markers uncommonly used in clinical practice), further complicated by the difficulty to define infant reference values in the context of high prevalence of malnutrition and potential comorbidity. In the absence of reference values for African infants, we used assay-specific normal ranges defined in other populations (Asian for BAP and European for CTX) that were, however, specific for the age of the infants studied. Reference to ‘normal’ values should therefore be considered keeping into account the presence of several concomitant factors, such as different ethnicity, suboptimal nutritional status and possible comorbidities that might affect bone mineral metabolism.

Despite these limitations, we were able to evaluate levels of bone markers with respect to growth indexes, presence of malnutrition, creatinine and other biochemical parameters: bone markers showed no differences by gender or growth restriction, and no correlation with growth indexes or haematochemistry parameters. Most importantly, for both markers, levels were similar to those of tenofovir-unexposed children followed in the same context, suggesting no significant effect of tenofovir on the two bone markers considered. This finding is particularly relevant, because tenofovir-exposed and -unexposed cases were taken from the same population, differing essentially for the profile of drug exposure. The levels of CTX, although significantly increasing between 6 and 12 months, were consistently within normal limits, while levels of BAP were consistently elevated at both timepoints. The presence of normal values of the bone resorption marker (CTX) for all tenofovir-exposed infants, even with the above caveats, is reassuring. The interpretation of the elevated levels of BAP should take into account different issues. High levels of BAP, coupled to normal levels of the CTX bone resorption marker, might not have negative implications, reflecting primarily skeletal growth. BAP levels, although generally representing a marker of bone formation, are also related to bone turnover. Few studies have been conducted in the paediatric population and information is limited. In infants and children, levels of bone markers are usually much higher than in adults and reflect not only bone remodelling but also rapid skeletal growth. Season of birth, intrauterine growth restriction and maternal conditions such as diabetes have been reported to affect bone marker levels in infancy, further complicating the interpretation of findings.^25–27 In this study, irrespective of tenofovir exposure, high BAP values were found in the majority of children, with levels that were roughly 3-fold higher compared with age-matched paediatric reference values published for Asian²⁰ and European populations.²⁸ Levels were also higher with respect to two published Italian studies that evaluated children from HIV-infected mothers, with or without prenatal exposure to tenofovir.^13,29 Growth restriction as a consequence of inadequate nutritional intake was quite common in the population studied; it is, however, uncertain whether malnutrition may represent a possible cause of the observed high BAP levels, because other studies in the context of limited resources have reported an opposite direction of the effect of malnutrition on BAP levels: malnourished infants had significantly lower BAP levels than healthy European controls, with a significant effect of nutritional intervention that restored BAP levels similar to those of healthy, age-matched European children.³⁰ Among specific nutrients, vitamin D intake and levels may be particularly relevant. Data in African infants are limited and show in the first year of life lower levels compared with later life.³¹ Controversial data have, however, been reported regarding the association between vitamin D deficiency and levels of bone markers. In European paediatric populations, vitamin D and CTX levels were inversely correlated,³² but vitamin D supplementation had little to no impact on markers of bone turnover, with the exception of osteocalcin.³³ No firm conclusions can be drawn, because no studies have concomitantly evaluated and discriminated in infants the independent roles of ethnicity, growth, malnutrition and vitamin D deficiency on bone metabolism and mineral turnover.

Interestingly, a longer duration of maternal ART in pregnancy (reflecting an earlier start in pregnancy of ART) was positively correlated with some growth indexes in the infants. This probably does not represent a direct positive effect of maternal ART on infant growth, but, more likely, an indirect effect of the benefits of an earlier entry in antenatal care. Although the absence of significant differences between tenofovir-exposed and -unexposed children is reassuring, indicating no negative impact of tenofovir on the levels of the two bone markers studied, this does not exclude potential negative effects of exposure to tenofovir during late pregnancy and breastfeeding on bone mineralization. We were unfortunately unable to verify this hypothesis through an analysis of BMD in our population. Reduced BMD has been described in HIV-infected children receiving tenofovir-based salvage regimens.^34,35 In one of these studies a limited number of HIV-infected children showed, following switching to tenofovir-based HAART regimens, an increase in levels of osteocalcin and bone-specific alkaline phosphatase, accompanied by absolute decreases in BMD.³⁴ In a comprehensive study that simultaneously analysed bone markers, infant growth and quantitative bone measurements, prenatal exposure to tenofovir showed no negative impact on bone health and infant growth.¹³ In this study, however, compared with ours, infants were older (range 11.8–77.9 months) and did not receive breast milk, being therefore unexposed in postnatal life through breastfeeding. In the recent study from Siberry et al.,¹⁷ exposure to tenofovir in pregnancy was associated with a lower bone mineral content in the first month of life, although tenofovir-exposed and -unexposed newborns showed no differences in anthropometric measures. Available evidence indicated that tenofovir penetration in milk is scarce,¹¹ and recent data on tenofovir plasma levels in breastfeeding infants confirm this finding, showing limited infant exposure in the presence of maternal treatment with tenofovir.³⁶ Although these data are reassuring, the safety of long-term infant exposure to tenofovir-based regimens during prolonged breastfeeding should be confirmed examining bone mineralization status in the first years of life. Longitudinal studies should investigate simultaneously the bone markers, bone mineralization status and growth velocity in children enrolled in Option B+ programmes in countries with limited resources.

Funding

This work was supported by grants from: Esther-Italy 2009-2010, Rome, Italy (no. 9M34); and Ministry of Health, Rome, Italy (no. 3C04/1).

Transparency declarations

S. V. received honoraria from ViiV, Gilead and Merck for scientific board membership. All other authors: none to declare.

Author contributions

G. L., M. A., M. C. M., E. B., P. S., S. M., S. V., M. G. and L. P. designed the study and finalized the manuscript. M. F. was responsible for statistical analysis and drafted and finalized the manuscript. G. L., H. J. and J.-B. S. substantially contributed to clinical activities, to acquisition of data and to critical revision of the manuscript. C. M. G. and R. A. performed laboratory analyses and critically revised the manuscript. All the authors gave approval to the final version to be published.

Acknowledgements

We wish to thank Alessandra Mattei for administrative help and Marco Mirra, Massimiliano Di Gregorio, Stefano Lucattini and Luca Fucili for IT support.

References