Tenofovir Disoproxil Fumarate for Prevention of Vertical Transmission of Hepatitis B Virus Infection by Highly Viremic Pregnant Women: A Case Series

Abstract

Background

Despite appropriate immunoprophylaxis, up to 10% of infants born to highly viremic hepatitis B virus (HBV–DNA C 7 log IU/mL) mothers are infected with HBV. Use of TDF to prevent vertical transmission (VT) by such mothers has not been evaluated.

Purpose

To evaluate the efficacy and safety of TDF in preventing VT from highly viremic HBV-infected mothers.

Methods

Data were collected retrospectively from HBV mono-infected, hepatitis B e antigen (HBeAg) positive, pregnant women between 6/2008 and 11/2010. Cases enrolled were HBV mono-infected mothers who received TDF (300 mg orally once a day) in the third trimester. Those with pregnancy complications or an abnormal fetus on sonography were excluded from use of TDF. All infants received hepatitis B immunoglobulin and vaccination at birth and subsequently.

Results

Eleven Asian mothers received TDF at the median gestational age of 29 (28–32) weeks and the median duration of TDF use before delivery was 10 (7–12) weeks. A significant reduction in serum HBV–DNA was achieved at delivery compared with baseline (mean 5.25 ± 1.79 vs. 8.87 ± 0.45 log₁₀ copies/mL, respectively; p <0.01). Three had serum ALT levels more than 1.5 times the upper limit of normal and two of these normalized before delivery. The 11 infants were born with no obstetric complication or birth defects. Five infants were breastfed. All infants were hepatitis B surface antigen negative 28–36 weeks after birth.

Conclusion

Our preliminary data suggest that TDF use in the third trimester is safe, and effectively prevents VT of HBV from high viremic HBeAg-positive mothers.

Introduction

Globally, chronic hepatitis B virus (HBV) infection is a leading cause of chronic hepatitis, cirrhosis, and hepato- cellular carcinoma. In HBV-endemic regions, particularly the Asia–Pacific region, infection is most commonly acquired by either vertical transmission (VT) or exposure during early childhood [1]. The risk of progression to chronic HBV infection is inversely proportional to the age at which the infection is acquired. Without immunoprophylaxis, infants born to mothers positive for both hepatitis B surface antigen (HBsAg) and hepatitis B e antigen (HBeAg) have a 40–90 % risk of VT, and subsequently 85–95 % of infected infants develop chronic HBV infection [2, 3], whereas infection acquired during adulthood is associated with <5 % chance of chronicity [4]. Therefore, prevention of VT is an important strategy to control HBV infection worldwide.

Utilization of HBV immunoprophylaxis has been effective at reducing VT of HBV; however, despite appropriate use of HBV vaccine and HBV immunoglobulin (HBIG), 8–10 % of infants born to HBeAg-positive mothers with HBV–DNA >6–8 log₁₀ copies/mL develop chronic HBV infection [5–8]. Use of antiviral treatment in late pregnancy to prevent or reduce the occurrence of VT has been well established in human-immunodeficiency virus (HIV) infection [9]. However, there is limited data on whether or not highly viremic pregnant women with HBV should receive antiviral therapy to prevent VT. Several clinical studies, including two recent relatively large, prospective controlled studies of lamivudine and telbivudine, have demonstrated that antiviral treatment in the third trimester significantly reduced the occurrence of VT to 0–10 % compared with transmission of 9–39 % for HBeAg-positive untreated mothers with high viral load [6, 7, 10].

Among the five United States Food and Drug Administration (FDA) approved oral anti-HBV agents, tenofovir disoproxil fumarate (TDF) and entecavir are the most effective agents in terms of potency and resistance profile, and as such are regarded as first-line therapy for HBV [11, 12]. According to the FDA drug category for pregnancy, TDF and telbivudine are classified as category B (no evi- dence of risk to humans: either animal findings indicate risk, but human findings do not; or, if no adequate human studies have been conducted, animal findings are negative) for use in pregnancy, whereas lamivudine, entecavir, and adefovir are category C (risk cannot be ruled out: human studies are lacking, and animal studies are either positive for fetal risk, or lacking also. However, potential benefits may justify the potential risk). In addition, TDF has been used in over 600 HIV mono-infected and HIV/HBV co-infected mothers and favorable efficacy and safety profile have been demonstrated [9, 13, 14]. Despite several theoretical advantages of using TDF to prevent VT, to date there have been no studies of the use of TDF in this setting. Therefore, the purpose of this retrospective analysis was to evaluate the efficacy, tolerability, and safety of TDF in preventing VT in highly viremic HBV-infected mothers.

Patients and Methods

Data were collected retrospectively from pregnant women at Elmhurst Hospital, New York Downtown Hospital, a private practice Hepatology clinic in New York, and the Hospital of the University of Pennsylvania from June 2008 to November 2010. Eligibility criteria for inclusion in this analysis were:

1. pregnant women;

2. positivity for serum HBsAg and HBeAg for a period of at least 6 months;

3. HBV–DNA levels C6 log₁₀ copies/mL before initiation of TDF; and

4. therapy with TDF 300 mg orally once a day (Viread; Gilead Sciences, CA, USA) beginning in the third trimester.

Mothers with HIV co-infection, pregnancy complications, or an abnormal sonographic examination were excluded from use of TDF. The study was approved by the local Institutional Review Board at each site. Tenofovir was chosen, as opposed to other antiviral agents, because of a combination of advantages of potent antiviral effects and favorable safety profile.

Baseline demographic data and virological characteristics (age, race, HBeAg, and history of prior HBV therapy) of pregnant women were recorded. Serum alanine aminotransferase (ALT), creatinine, and HBV–DNA levels were assessed before initiation of TDF and at delivery. Serum HBV–DNA levels were determined by polymerase chain reaction (PCR) assay (Quest Diagnostics, USA or Lab- Corp, USA), with a lower limit of detection of 169 copies/ mL. Data upon delivery of the infant (mode of delivery and complications) were also recorded. A significant change in serum creatinine was defined as a change in serum creatinine of C0.5 mg/dL from baseline as defined in the Phase 3 trial. The decision to stop or to continue TDF after delivery was made by the treating physician and on the basis of the desire of the patient. For those mothers for whom TDF was discontinued, serum ALT and HBV–DNA levels were monitored periodically for at least 24 weeks to detect hepatitis flare. Characteristics of the neonate at birth (Apgar score, anthropometry, and birth defects), and history of immunoprophylaxis of infants were reviewed and recorded. All infants born in these centers were scheduled to receive a single 0.5 mL (200 IU/mL) dose of HBIG (HyperHEP B solvent/detergent treated; Talecris Biotherapeutic, NC, USA), intramuscularly within 12 h of birth and then 0.5 mL (10 mcg/0.5 mL) of HBV vaccine (Recombivax HB; Merck Sharp and Dohme, NJ, USA) intramuscularly within 24 h of birth and at 1 and 6 months of age. Infants were assessed for serum HBsAg at 28–36 weeks of age. Vertical transmission was evaluated by HBsAg testing in the peripheral blood of infants at 28–36 weeks of age.

Statistical Analysis

Categorical variables were summarized as number or per- centage. Continuous variables were summarized as mean (± standard deviation) or median (range). HBV–DNA levels were logarithmically transformed for analysis. To compare two variables, SPSS 15.0 software (SPSS, Chicago, IL, USA) was used for statistical analysis. Student’s t test was used for continuous variables of in-between group comparisons, expressed as means (± standard deviation); for categorical variables, v² tests were used for in-between group comparisons. The a-level was set at 0.05.

Results

Maternal Characteristics

During the period from June 2008 to November 2010, twelve highly viremic pregnant women received TDF in the third trimester to prevent VT; one mother had pre- treatment HBV–DNA <6 log₁₀ copies/mL, and was excluded from the analysis. Finally, 11 pregnant women met all inclusion criteria. The key characteristics of each patient at baseline and delivery are shown in Table 1. The median maternal age was 29 (23–45) years and all were Asian Americans (9 were born in China, 1 was born in Nepal, and 1 was born in the United States). Two of the husbands were HBsAg positive, 7 were negative and status was unknown for two. Three women had antiviral therapy before pregnancy for reason of elevated serum ALT levels; one received entecavir for a few months before pregnancy, one woman received lamivudine for 1 year before preg- nancy, and one received TDF for 7 months before pregnancy. For these three women, all antiviral agents were discontinued before pregnancy. Tenofovir was started at the median gestational age (GA) of 29 (28–32) weeks. The median duration of TDF use before delivery was 10 (7–12) weeks. The median HBV–DNA level before initiation of TDF was 9.00 (7.70–9.40) log₁₀ copies/mL. Serum ALT levels were elevated to more than 1.5 times the upper limit of normal (ULN) in three patients before initiation of TDF. Serum creatinine levels were within the normal ranges in all patients before initiation of TDF.

Table 1.

Characteristics of 11 pregnant women with HBV on tenofovir during pregnancy

ID no.	Age (years)	Prior Rx	At initiation of TDF				Duration of TDF before delivery (weeks)	At delivery						Rx duration after delivery (weeks)
			GA (weeks)	HBV-DNA (log c/mL)	ALT (U/L)	Cr (mg/dL)		GA (weeks)	Mode	HBV-DNA (log c/mL)	ALT (U/L)	Cr(mg/dL)	Complications
1	30	-	30	9.00	25	0.5	9	39	VA	8.81	20	0.6	-	2
2	45	ETV	29	7.70	35	0.6	12	41	CS	3.40	17	0.6	-	0
3	23	-	28	9.01	18	0.4	11	39	CS	6.02	20	0.5	-	0
4	30	LAM	30	8.81	37	0.4	9	39	VA	4.59	57	0.4	-	Ongoing
5	23	-	32	8.76	20	0.5	7	39	CS	6.72	50	0.7	-	0
6	23	-	28	9.01	18	0.4	11	39	CS	3.02	20	0.5	-	0
7	28	-	31	9.40	73	0.4	10	41	CS	3.96	37	0.6	-	6
8	31	-	29	9.40	723	0.5	12	41	CS	3.63	19	0.6	-	Ongoing
9	28	-	28	8.72	448	0.4	9	37	VA	6.75	338	0.5	MSAF	12
10	29	TDF	32	8.81	15	0.5	8	40	VA	6.24	12	0.6	-	Ongoing
11	32	-	28	9.00	42	0.7	11	39	CS	4.60	62	0.8	-	Ongoing

Rx, treatment; ETV, entecavir; LAM, lamivudine; TDF, tenofovir; GA, gestational age; VA, vaginal; CS, caesarian section; HBV, hepatitis B virus; ALT, alanine aminotransferase; Cr, creatinine; MSAF, meconium-stained amniotic fluid

Maternal Outcomes

A significant reduction in serum HBV DNA level was achieved at delivery compared with baseline (mean 5.25 ± 1.79 vs. 8.87 ± 0.45 log₁₀ copies/mL, respectively; p <0.01). Six of 11 mothers (55 %) had a serum HBV DNA level \10⁶ log₁₀ copies/mL at delivery. Serum ALT levels became normal in two of the three patients with an elevated serum ALT at baseline and remained elevated at >1.5 times ULN before delivery in one patient. No significant change in creatinine levels during TDF treatment was observed. The median GA at delivery was 39 (37–41) weeks and, on the basis of obstetric indications, seven women had caesarian section whereas four delivered vaginally. There were no significant obstetric complications during delivery; one woman had meconium-stained amniotic fluid without adverse consequence to the infant. In eight cases, TDF was discontinued soon or within a few months after delivery (range 0–12 weeks), whereas three woman continued TDF until the end of the study period (two of these women had HBV viremia with elevated ALT before pregnancy). After TDF discontinuation, no ALT flares (>5 times of baseline or >10 times ULN) were observed during our study period. For six mothers for whom follow-up HBV–DNA levels were obtained, HBV– DNA rebound of >2 log₁₀ copies/mL was observed in all.

Infant Characteristics and Outcomes

The characteristics of eleven infants are summarized in Table 2. All infants had good Apgar scores (range 9–10) and did not have birth defects or congenital malformations.

Table 2.

Characteristics of 11 infants born to mothers with HBV

ID no.	Gender	BW (g)	Apgar scores at 1/5 min	Birth defect	Breast feeding	Immunoprophylaxisa	HBsAg after 28–36 weeks of age
1	F	2,721	9/9	None	No	Yes	Negative
2	F	3,220	9/9	None	Yes	Yes	Negative
3	M	3,005	9/9	None	Yes	Yes	Negative
4	M	3,828	9/9	None	Yes	Yes	Negative
5	M	3,850	9/9	None	Yes	Yes	Negative
6	M	3,005	9/9	None	Yes	Yes	Negative
7	F	3,669	9/9	None	NA	Yes	Negative
8	F	3,121	9/9	None	No	Yes	Negative
9	M	3,401	9/10	None	No	Yes	Negative
10	F	2,948	9/9	None	No	Yes	Negative
11	F	3,266	9/9	None	No	Yes	Negative

F, female; M, male; BW, birth weight; HBsAg, hepatitis B surface antigen

^aHepatitis B immunoglobulins 200 IU intramuscular within 12 h after birth and three doses of hepatitis B vaccine at 0 (within 24 h after birth), 1, and 6 months

Discussion

All infants were HBsAg negative at 28–36 weeks of age. All infants received appropriate immunoprophylaxis with HBIG and hepatitis B vaccine. In this paper, we report retrospective results from use of TDF to prevent VT by pregnant women with high viremia HBV infection. Although there is much data on use of TDF to prevent VT by HIV mono-infected and HIV/HBV co- infected mothers, as far as we are aware, this is the first report in HBV mono-infected mothers. In accordance with previous studies [6, 7], a combination of early post-natal immunoprophylaxis and third trimester antiviral therapy was safe, well-tolerated, and effectively prevented VT of HBV by highly viremic pregnant women.

Vertical transmission is believed to occur during delivery [15]. A correlation between intrauterine serum HBV– DNA levels and perinatal transmission of HBV by pregnant women has been clearly demonstrated [5–8]. Immunoprophylaxis with immediate HBIG and HBV vaccine after delivery effectively prevents VT in most instances. Failure of immunoprophylaxis is almost always because of high maternal viral load [5, 16]. A prospective, observational study from Australia reported that, despite appropriate immunoprophylaxis, VT occurrence was 7 % from HBeAg-positive mothers, and 9% from mothers with very high HBV DNA levels (C10⁸ copies/mL), whereas trans- mission risk was 0% when HBV DNA levels were <10⁸ copies/mL [5]. In addition, a recent analysis of 869 HBV- infected mother–infant pairs from China revealed that serum HBV DNA level was an independent risk factor (odds ratio 1.88, 95 % CI 1.07–3.30) for failure of immunoprophylaxis and that this occurred for HBeAg-positive mothers with HBV DNA levels C6 log₁₀ copies/mL [8]. These findings are similar to reported experience at our center. Among 101 mothers, nine (9 %) had immunoprophylaxis failure; six of nine mothers were tested for HBeAg and all were positive, and five mothers who were evaluated for HBV–DNA had viremia of >7 log₁₀ copies/ mL [17]. Therefore, as a corollary to the strategy for pregnant women with HIV infection, initiation of an anti- viral agent during pregnancy to suppress HBV replication and to reduce HBV–DNA to low or undetectable levels before delivery would be a reasonable approach to prevent VT by highly viremic pregnant women. Recently, this hypothesis has been supported by prospective studies; a randomized double-blind placebo-controlled study (N =141) using lamivudine from GA 32 weeks to week 4 postpartum significantly reduced the occurrence of VT compared with placebo (18 vs. 39 %, respectively; p = 0.014) [6]. Notably, mean HBV–DNA during delivery in the lamivudine treated group was 51.4 mEq/mL (approximately 10⁸ copies/mL) and this study had relatively high drop-out in both groups (13 and 31 %, respectively). More recently, a prospective, open-labeled study (N = 229) using a more potent nucleoside analog, telbivudine, starting earlier, from GA 20 weeks, achieved a mean HBV–DNA level of 2.42 log₁₀ copies/mL before delivery and was associated with 0 % VT compared with 8 % for controls (p = 0.002) [7].

The threshold for HBV–DNA levels and optimum time- point before delivery for intervention are questions that are not well answered. One could however speculate that lower HBV–DNA levels (for example C10⁶ copies/mL) and sooner than 3 months from term would be better thresholds for intervention, because they would further reduce the risk of VT. However, intervention in the first and second trimesters of gestation, during which time fetal organ development occurs, may be a challenge. Tenofovir, one of the most potent antiviral agents available [11, 12], results in faster viral decline, and its efficacy is superior to that of adefovir and lamivudine for patients with HBeAg-positive chronic HBV, although not in head-to-head studies of the latter [12]. Its antiviral effect is clearly observed from weeks 4 to 8 of treatment and approximately 7 log₁₀ copies/mL reduction is achieved at week 24 of treatment [18]. In our study, approximately 4 log₁₀ copies/mL reduction of HBV–DNA was achieved during the median TDF treatment period of 10 weeks in the third trimester, which is comparable with that observed for non-pregnant HBeAg- positive patients [18] In addition, no VT was observed, as compared with 9 % VT for immunoprophylaxis alone in our historical cohort [17]. However, a limitation of our study is that we do not have a control group of untreated pregnant mothers with high viremia and, therefore, a larger prospective study is needed to reveal whether intervention with anti-viral prophylaxis is effective is preventing maternal–fetal transmission. Nevertheless, this pilot study observation, which is the first to be undertaken with tenofovir and with no safety signals, should serve as an impetus to systematically study it, to address the question of prevention of maternal–fetal HBV transmission.

It should be noted that the potential benefit of this strategy should be weighed against the risk of adverse effects of the drug on mother and fetus, for example teratogenicity, long-term effects on infant bone development, post-treatment ALT flare, and induction of HBV-resistant mutations. Tenofovir is both safe and well tolerated in the non-pregnant setting. The most frequent adverse reactions in patients receiving TDF in clinical trials were mild to moderate gastrointestinal symptoms [18, 19]. Although renal events have been observed with use of TDF for patients with HIV infection, predominantly patients with preexisting renal disease, no nephrotoxicity was observed in a three-year period of use of TDF for chronic HBV patients with preserved baseline renal function [18, 19]. In terms of pregnancy safety, TDF is classified by FDA as category B. In-vivo safety data on antiviral agents during pregnancy are mainly derived from the Antiretroviral Pregnancy Registry (APR), which is an international, vol- untary-registered, prospective cohort study of women exposed to an antiviral agent, most of whom were HIV infected [20]. Among anti-HBV agents, lamivudine and TDF are the agents with which there is most experience in the first trimester; birth defects have been reported for 3.1% (118 from 3,864 live births) and 2.4 % (26 from 1,092 live births), respectively, similar to the occurrence in normal pregnancy [20]. In uncontrolled observations, one case of birth defect (from 25 live births) has been reported after use of entecavir, and no birth defects were reported after use of adefovir (from 42 live births) and telbivudine (from 7 live births) [20]. More evidence supporting the safety of anti-HBV agents for mothers during the second/ third trimester have been reported in several clinical studies [6, 7, 10, 21] and the APR [20].

Apart from teratogenicity, there has been a concern regarding TDF and fetal growth, particularly bone development. In the setting of pregnancy, a study of monkeys given TDF doses approximately twice as high as those for human therapeutic use resulted in a variety of reduced fetal growth and reduction in fetal bone porosity within 2 months of starting maternal therapy [22]. Although the clinical significance is unknown, TDF use in HIV-infected children has been reported to be associated with decreases in bone mineral density [23, 24]. In contrast, other long-term safety data for infants perinatally exposed to TDF indicated there was no abnormal bone metabolism or growth impairment in the children, or of renal impairment [25, 26]. In our study, TDF seemed to be safe for both mothers and their infants. Serum creatinine levels were stable and within the normal range during TDF treatment of all mothers, and we did not encounter any adverse pregnancy outcomes and/or birth defects. It should, however, be noted that assessment of infant bone development was by clinical observation and not periodic radiologic bone survey.

There is no convincing evidence that HBV is transmitted through breast milk [15, 25] and a current guideline suggests that breastfeeding is not contraindicated for mothers with HBV who are not receiving antiviral therapy and while infants are receiving immunoprophylaxis [12]. However, because of scant data on antiviral agent secretion in human breast milk and the potential long-term effect on infants, breastfeeding is not recommended for mothers on antiviral therapy [15, 25]. Although there are no human data, secretion of TDF has been documented in animal models [27]. However, TDF is converted to tenofovir before it is secreted in the milk and tenofovir is not absorbed by the gastrointestinal tract. Thus, the effect of TDF through breast feeding may be minimal, if any. Regardless of this, in our study, despite instructions from physicians, five mothers decided to proceed with breast- feeding and no short-term adverse outcome to their infants was observed.

For mothers who had active liver disease, either before or during the pregnancy, long-term treatment after delivery may be required to prevent viral rebound and ALT flare. A major advantage of TDF over other anti-HBV agents used in this context (lamivudine and telbivudine) is its superior resistance profile. For HBeAg-positive patients, resistance to TDF was 0 % after 3 years of treatment [19] whereas resistance to lamivudine and telbivudine was 40 and 25 %, respectively, after 2 years [28].

In conclusion, our preliminary results suggested that TDF use in the third trimester was safe, well-tolerated, and effectively prevented VT by HBeAg-positive HBV mono- infected pregnant women. Thus, TDF may be an effective option, in addition to immunoprophylaxis, for preventing VT from highly viremic HBV-infected mothers. In addition, TDF has a low resistance profile and thus enables long-term continuation of anti-HBV therapy for these mothers, if such therapy is believed to be medically indicated. Prospective randomized, controlled, and larger studies are warranted to confirm these findings.