Preventing Perinatal Transmission of HBV: An Australian Perspective
- First Online:
- Cite this article as:
- Levy, M.T. Curr Hepatitis Rep (2012) 11: 206. doi:10.1007/s11901-012-0144-4
- 75 Views
Hepatitis B infection is best controlled by prevention rather than treatment, as chronic infection is usual once infected at a young age. Infant immune-prophylaxis is highly efficacious, although in the setting of high maternal viral load, breakthrough infection still occurs in almost 10 % of babies. Ante partum antiviral therapy for the purpose of preventing mother to child transmission in this group is important to consider. This article provides an up-to-date account of the available evidence of the safety and efficacy of ante partum antiviral therapy options and a management plan is proposed. Mothers with HBV infection and high viral load should have the opportunity to consider this evidence. Post-partum HBV flares are common but usually mild. It is unclear whether post partum flares are best ignored till they settle or if they represent an opportunity for intervention to increase the chance of HBV clearance.
Despite the relatively low prevalence of hepatitis B virus (HBV) infection in Australia, high rates of migration from South East Asian Countries have resulted in higher prevalence of chronic HBV in some geographic areas.
Standard active-passive immune prophylaxis is a success story, reducing transmission from 70 to 90 % to 5–10 %. [2–4]. For many years this was considered sufficient and in Australia mothers with HBV infection were reassured and babies were vaccinated, but not generally followed up for HBV testing. Times are changing.
We know that HBV infection in infancy results in a >90 % chance of lifelong infection with significant clinical, financial and personal consequence . With the advent of direct anti-virals and growing experience with their use in pregnancy in the setting of HIV infection , attention has turned to the problem of failed immunoprophylaxis and perinatal HBV transmission. Preventing HBV transmission to all babies of infected mothers has become the goal. Identification of those most at risk of perinatal transmission for consideration of additional strategies is now recommended by guidelines and in review articles [7–9]. Which strategies and for whom are still not clearly defined.
This article will review key evidence guiding the management decision points in the prevention of perinatal transmission of HBV.
Perinatal Transmission Despite Immunoprophylaxis
Despite overall efficacy of immunoprophylaxis, varying and sometimes alarming rates of perinatal HBV transmission from 7 % to as high as 27 % from HBV DNA positive mothers have been reported [3, 10–12].
The rate of transmission of HBV despite immunoprophylaxis has now been clarified by consistent reports from a number of recent studies. A large retrospective case–control study of 869 infants from HBsAg positive mothers at a single centre in Beijing was recently reported by Zou, et al. [13•]. Maternal and infant factors associated with immunoprophylaxis failure were examined. All babies in whom immunoprophylaxis failed were born to HBeAg positive mothers. The risk of transmission from HBeAg positive mothers was 5.6 %, similar to the rate our group reported (6.6 %) and similar to that of other studies [14, 15•]. Transmission from HBeAg negative mothers was generally not seen. No association was found between immunoprophylaxis failure and infant prematurity, premature rupture of membranes, premature birth, low birth weight, poor Apgar scores or meconium staining [13•]. Breast feeding is not thought to be a risk factor for transmission of HBV, unlike HIV infection .
Maternal HBV DNA levels allow further risk stratification. In the study by Wiseman et al., transmission was seen only from mothers with viral load ≥ log 8 copies/ml observed in 8.5 % [15•]. Zou et al. identified transmission in 3.2 % of babies when maternal viral loads were between 6 and 6.99 log copies/ml, 6.7 % for 7–7.99 log copies/ml and 7.6 % for ≥ 8 log copies/ml [13•]. Transmission rates of 8 % and 8.6 % from high viral load mothers were reported in the control arms of two recent studies utilising telbivudine to prevent transmission[17••, 18••].
An algorithm was recently proposed suggesting that antiviral therapy be considered when maternal viral load is ≥ 6 log copies/ml. Extrapolating cut-offs from studies using RT-PCR assays reported in copies/ml without reference to standards is not ideal. To exemplify this problem, only 14 % of HBeAg positive mothers had viral loads ≥ 8 log copies/ml in the Zou study, whereas 76 % of our HBeAg positive cohort had viral loads ≥ 8 log copies/ml[13•, 15•]. These differences are more likely to reflect assay variation than true difference. In the absence of studies reporting transmission risk according to HBV DNA in IU/ml, guidelines should acknowledge variability between PCR assays in different centres.
Currently we use a cut-off of ≥ 7 log IU/ml (approximately equivalent to ≥ 8 log copies/ml in our lab) to identify women for whom a discussion about perinatal transmission risk and consideration of additional strategies is appropriate.
Efficacy of Direct Acting Antiviral Therapy on Transmission Rates
The first data to emerge on the use of antepartum antiviral therapy utilised lamivudine and showed a promising effect compared to historical controls . A randomised trial of lamivudine versus placebo in 2009 was hampered by disproportionately large numbers lost to follow–up in the placebo group (13 vs 31 %). Analysis excluding those lost to follow up failed to demonstrate a statistically significant benefit of lamivudine. An analysis combining cohorts (albeit published mostly in the Chinese literature) concluded that antepartum lamivudine was effective in reducing transmission to 13 % from 23.7 % in those untreated . However, the rates of infection in the lamivudine arms approached what we see in untreated mothers. From 2009 to 2011 we offered lamivudine 100 mg to mothers with viral loads above 7 log IU/ml from 32 weeks pregnancy continuing till 2–4 weeks after birth. To date, we have not seen any perinatal HBV transmission from lamivudine treated mothers in 38 babies tested at 9 months of age (0/38). In contrast, we continue to see transmission despite immunoprophylaxis from mothers who chose not to take lamivudine (2/10) (unpublished observations).
What is the virological effectiveness of short term lamivudine 100 mg in this setting? The initial dose finding reports using lamivudine for HBV suggested that 100 mg was markedly inferior to 300 mg at the 4 week time point although by 12 weeks there was no significant difference in viral suppression . Is lamivudine 100 mg starting at 32 weeks able to suppress virus sufficiently? We examined the virological responses in 25 mothers with viral load >7 log IU/ml after a mean of 52 days of lamivudine . Maternal viral load at the time of birth remained above log 7 IU/ml in 20 % of cases. Why is lamivudine effective in preventing transmission even if the virological responses after 8 weeks are relatively modest? Perhaps a foetus that has sufficient lamivudine exposure, as a result of placental transfer, is directly protected against HBV in their blood stream during birth reducing the chance of immunoprophylaxis failure .
In view of the suboptimal virological responses to lamivudine, in combination with studies suggesting incomplete protection from transmission, more potent antiviral therapy may be preferred.
A tested alternative to lamivudine is telbivudine, whose efficacy has been reported recently in two prospective case–control open-label opt-in trials both conducted at the Second Affiliated Hospital of the South East University of Nanjing China. The first offered telbivudine 600 mg commencing at week 20–32 to 229 mothers with viral loads > log 7 copies/ml and normal ALT (135 telbivudine, 94 control) [17••]. The second trial directed toward mothers with elevated ALT offered telbivudine 600 mg earlier, commencing week 12–30 mothers with viral loads > log 6 copies/ml and ALT ≥ 40 U/L (53 telbivudine, 35 controls) [18••]. This study was directed to women with abnormal ALT so the Telbivudine was started earlier and with lower viral load threshholds. Telbivudine was effective in both studies, with no cases of transmission in the treatment arms compared with 8 %(7/88) and 9 % (3/35) in the control arms.
It would appear that direct acting anti-viral therapy, particularly telbivudine is effective in the prevention of perinatal transmission. It may be expected that other potent anti-virals such as tenofovir and entecavir could be effective, but as yet the efficacy of tenofovir is not proven and entecavir is not recommended in pregnancy. This will be discussed further below.
Safety of Direct Acting Antiviral Therapy in Pregnancy and Breastfeeding
Registration studies for HBV antiviral therapy have specifically excluded pregnant or breast feeding women, thus these therapies were not licensed for use in this population. As a consequence, there is limited experience of the use of these agents in pregnancy. To determine safety for the developing foetus, information about placental transfer and foetal drug exposure, studies of teratogenicity and fertility in animals as well as humans are required. Animal experiments (type of animal and drug exposure) do not appear to be standardised which makes comparisons between drugs somewhat difficult. Human data has largely come from the use of these agents in HIV infection, for which information is prospectively collected in an Antiretroviral Pregnancy Registry established in 1989[24, 25]. Only a minority of subjects in this register have HBV mono-infection. The registry is primarily designed to detect congenital anomalies, teratogenticity usually consequent on early pregnancy drug exposure, rather than more subtle changes that might be a concern in last trimester exposure. The registry size enables power to detect a ≥ 2 fold increase in overall birth defects for lamivudine and tenofovir. Dr Giles and her colleagues have recently reviewed this topic thoroughly .
Lamivudine is a FDA pregnancy category C drug as animal studies have shown adverse foetal effects, particularly embryo lethality in rabbits but not rats. There are no adequate well controlled studies in humans. Lamivudine crosses the placenta by simple diffusion in a 1:1 ratio. Despite this FDA category, the largest experience in pregnancy is with lamivudine. There are over 6,000 third trimester exposures recorded with no increases in the rate of foetal malformations. Lamivudine is excreted into breast milk although it has been calculated that the infant would be exposed to a relatively small dose compared with standard oral dosing [28, 29].
The significant safety issue with lamivudine is for the mother since in other settings lamivudine has been superseded because of its low barrier to resistance . Selection and archiving of resistance mutants by exposure to lamivudine might compromise a mother’s responsiveness to anti-viral therapy in the future [31, 32]. Initially it was thought that resistance might only be a concern after 6 months exposure, but using more sensitive assays, resistance mutations emerged in 16 of 30 subjects after only three months of lamivudine . We examined a group of 24 lamivudine treated mothers and using population sequencing the multidrug resistant HBV variant rtA181T was noted to emerge in one woman. With subsequent pregnancies it might be anticipated that repeat lamivudine exposure will expand these resistance mutant populations
Telbivudine is a more potent nucleoside analogue than lamivudine, with a reduced but still significant incidence of antiviral resistance [35–37]. Telbivudine has a FDA pregnancy category B listing in the USA as no animal toxicity has been shown in mice and rats. Human data are limited, as telbivudine has not been used in HIV infection . Toxicity was not reported in either of the two recently published studies [17••, 18••]. It is difficult to be certain about the role of telbivudine in this setting. Certainly it is effective, but further study is required to determine if selection of drug resistance mutants particularly with repeated exposure is a concern. There is no information to guide advice about breastfeeding with Telbivudine.
Tenofovir disoproxil fumarate (TDF) is an orally absorbable pro-drug of tenofovir, which has activity against HBV and HIV and is also a pregnancy FDA category B drug. Tenofovir has a high barrier to resistance and superior potency and thus is a first line therapy for HBV infection. Tenofovir is used with increased frequency in the setting of pregnancy in HIV. Reports of over 500 exposures in second and third trimester in the Antiretroviral Drug Registry are reassuring with no evidence of congenital malformations.
Reports of animal toxicity with tenofovir have raised concerns about safety in humans. High dose administration (30 mg/kg) to gravid Rhesus macaques resulted in lower foetal crown rump length, body weight and slightly smaller adrenal glands, as well as lower foetal circulating insulin-like-growth-factor levels [38–40]. Some of this toxicity was thought to be due to maternal renal tubular toxicity and consequent phosphate depletion rather than a direct effect on the growing foetus . When more comparable doses were administered to gravid Rhesus macaques, growth restriction was not observed, suggesting that the effects may be dose dependant. A recent study of tenofovir in human pregnancy did not show any effect on infant birth weight or small for gestational age measurements. Despite the highly reassuring information from this report, it was noted that despite exposure only in-utero, the infants at 12 months age were observed to have a slight reduction in head circumference (0.32 cm) and length (0.41 cm) compared to matched controls[41••, 42]. The mechanism of this effect is not known, although late effects attributed to anti-retroviral exposure in-utero have been reported in other studies including an increase in rare CNS cancers in older children after in-utero exposure, thus follow up of infants at later time points is important.
There may be some biological plausibility that suggests toxicity is reduced in the foetus. The renal toxicity of tenofovir is related to cumulative renal tubular intracellular concentrations. Infant macaques did not have any proximal renal tubular toxicity even after high dose maternal tenofovir exposure and relatively high foetal plasma drug concentrations . It is postulated that as foetal proximal tubules are poorly developed they have low uptake of tenofovir which prevents accumulation to toxic intracellular concentrations. This theoretically would then protect the foetus from bone effects secondary to renal tubular toxicity and phosphate wasting. A study of tenofovir in healthy HIV infected children ages 6–17 years reported no effects on bone mineral accrual after 12 months therapy [44, 45].
In mothers taking the oral formulation of tenofovir (the bioavailable pro-drug TDF), breast milk is expected to contain the poorly absorbed parent compound tenofovir . A study in lactating rhesus macaques showed that tenofovir concentrations in breast milk corresponded to 20 % of the serum values . After maternal tenofovir injection, human breast milk concentrations of tenofovir are very low and it has been calculated that nursing infants would be exposed to less than 2 % of the oral infant therapeutic dose. Overall these findings are very reassuring and should be discussed with a mother in the context of her wishes about breastfeeding and the plans to continue tenofovir post-partum.
Post-partum Hepatitis Flares
A further concern for the mother is the impact of antiviral therapy on her own liver disease. Post partum hepatitis flares in HBV are thought to be the result of immune reconstitution after delivery [48–52]. Co-incidental virological rebound following withdrawal of antiviral therapy may exacerbate immune reconstitution flare. A study by ter Borg and co-workers showed that post partum hepatitis flared occurred in 62 % of their lamivudine treated cohort compared to 36 % in their untreated historical control. This study was limited by small numbers (n = 13) in the treatment group and by unmatched cohorts (85 % of the lamivudine group but only 52 % of controls were HBeAg positive). In contrast, a lower rate of post partum flares in lamivudine treated subjects was described in the Xu study . In our clinic, we have observed that those taking ante-partum lamivudine who cease within 2 weeks after birth commonly have asymptomatic post-partum ALT flares, seen in 14/44 (32 %) of lamivudine treated and 3/16 (18 %) untreated mothers, although the difference was not statistically significant. Importantly, the majority of these flares settle by 6 months . HBeAg seroconversion was seen in 8.7 % of our cohort but was not more common in those treated with lamivudine or in those who had a post partum hepatitis flare .
In a study of telbivudine in mothers with elevated baseline serum ALT values and high viral load, HBeAg seroconversion was observed in 15 % of the telbivudine treated mothers, most of whom continued drug post-partum. HBeAg seroconversion was not observed in the control cohort[18••]. This most interesting observation needs further examination in a larger study.
In our experience post-partum hepatitis flares are common, asymptomatic and settle spontaneously in the majority within the first 6 months. With increased monitoring in the post-partum period, hepatitis flares will be frequently recognised but is treatment always necessary? In the absence of significant underlying liver disease, particularly for mothers who were in the immune tolerant phase prior to commencing therapy, we usually watch LFTs for 3–6 months before instituting therapy which as for all HBV, is likely to be required for many years or indefinitely.
Recommendations for the prevention of perinatal transmission and post-partum care
1. First trimester
Assess HBsAg positive women for HB eAg, HBV DNA
Assess for liver disease and treat if indicated
Provide HBV education
2. Second trimester
If viral load ≥ log 7 IU/ml discuss antiviral therapy
(If your lab reports copies/ml, adjust threshold)
Use Lamivudine, telbivudine or tenofovir
Start by 30–32 weeks
Instruct on intended stopping date
3. Birth, breastfeeding and beyond
Timing of cessation will depend on drug choice and baseline maternal factors
Caution against prolonged therapy with lamivudine and telbivudine to avoid selection of resistance mutation HBV
Monitor ALT, HBV serology and HBV DNA post-partum.
Reassess liver disease.
Post partum hepatitis flares are common and usually settle
Discuss breast feeding data relating to drug regime chosen with patient
Breast feeding is not thought to be a route of HBV transmission
Check infant for HBsAb (vaccine success) and HBsAg at 9 months age or 3 months after last vaccination.
HBcAb not useful, maternal antibodies can persist for ≥ 12 months
No potential conflict of interest relevant to this article was reported.