Tenofovir alafenamide use in pregnant and lactating women living with HIV

Abstract

Introduction:

Tenofovir alafenamide (TAF)-containing fixed-dose drug combinations (FDCs) are increasingly being used in managing pregnant women living with HIV. However, TAF is not currently recommended during pregnancy due to limited pharmacokinetic and safety data. TAF, a newer nucleotide phosphonamidate prodrug of tenofovir (TFV), achieves high levels of tenofovir-diphosphate in lymphoid cells and hepatocytes, and 90% lower systemic concentrations of TFV compared to tenofovir disoproxil fumarate (TDF), thereby maximizing TAF’s antiviral efficacy, potency and clinical safety.

Areas covered:

This review discusses the currently available information on the pharmacology of TAF in pregnant women living with HIV. Pharmacokinetic studies with TAF during pregnancy have yielded varying results compared to postpartum, but TAF exposures during pregnancy have been within the range of those typically observed in non-pregnant adults. The efficacy and safety of TAF in treatment-naïve pregnant women living with HIV is currently being evaluated in the VESTED study, a phase-III NIH randomized clinical trial.

Expert opinion:

Initial pregnancy data suggest that TAF-based FDCs have high efficacy and low risk of adverse effects during pregnancy. TAF is likely to become part of first-line regimens for use in pregnant women living with HIV once additional pregnancy data from phase III trials are available.

1. Introduction

The use of antiretroviral (ARV) medications in pregnant women living with human immunodeficiency virus (HIV) continues to be of critical public health importance [1,2]. Without treatment, about 15–40% of pregnant women living with HIV are at risk of transmitting the virus to their fetuses [3]. The 1994 landmark Pediatric AIDS Clinical Trials Group (PACTG) 076 study, the first study of antiretroviral safety and efficacy in pregnancy, showed that the administration of zidovudine (ZDV) to pregnant women living with HIV and to their neonates after birth, decreased the risk of perinatal HIV transmission by 68% (from 25.5% to 8.3%) [4]. Since 1994, combination ARV therapies with multiple potent HIV drugs have proven to be more effective than ZDV alone at preventing perinatal HIV transmission, changing the trajectory for the treatment of HIV during pregnancy and reducing perinatal transmission rates to negligible numbers, while improving maternal health immensely.

The World Health Organization (WHO) [5], United States Department of Health and Human Services (DHHS) [1] and the European AIDS Clinical Society (EACS) [6] HIV perinatal guidelines all recommend that pregnant women living with HIV receive a combination of two nucleoside reverse transcriptase inhibitors (NRTIs) with a third ARV from another class. Currently recommended regimens for most pregnant women include a two NRTI backbone of abacavir/lamivudine, tenofovir disoproxil fumarate (TDF)/lamivudine or TDF/emtricitabine, plus an integrase strand inhibitor (raltegravir or dolutegravir) or a boosted protease inhibitor (darunavir/ritonavir or atazanavir/ritonavir) [1]. Although TDF is commonly included in the NRTI backbone of ARV regimens currently used in pregnant women living with HIV [7], its use was associated with a higher risk of adverse pregnancy outcomes in comparison to ZDV-based regimens [8], though other studies and systematic reviews have deemed its use as safe [9,10]. TDF is associated with declines in renal function and greater bone density loss in non-pregnant adults in comparison to tenofovir alafenamide (TAF), especially when coadministered with ritonavir or cobicistat [11]. Despite these advantages, TAF is not currently recommended during pregnancy due to limited pharmacokinetic (PK) and safety data.

TAF is a newer TFV prodrug that is increasingly being used by women living with HIV across the United States and Europe [12-14]. TAF achieves high levels of tenofovir-diphosphate (TFV-DP) in lymphoid cells and hepatocytes, and ~90% lower systemic concentrations of TFV compared to TDF [15]. The ability of TAF to selectively concentrate in target cells and its greater affinity and distribution into lymphoid tissue maximizes its antiviral efficacy, potency and clinical safety [16]. These pharmacologic properties are critical, as the lower plasma concentrations of TFV from TAF are associated with a reduced risk of decline in glomerular filtration rate, renal tubular toxicity, and decreased bone mineral density with prolonged use compared to TDF [14,17].

Fixed dose combinations (FDCs) of ARVs provide single pill once daily dosing regimens that improve HIV management due to their convenience, enhanced safety profile and reduced cost, leading to improved adherence and reduced risk of HIV drug resistance and transmission [18-20]. Since FDCs that include TAF are increasingly being used in pregnancy to treat HIV infection and prevent perinatal transmission, there needs to be a clear understanding of its pharmacology, safety and efficacy in this population. The aim of this article is to provide a review of the currently available information on the pharmacology, clinical efficacy, and safety of TAF containing ART combinations during pregnancy.

2. TAF fixed dose combinations

Tenofovir must be administered as a prodrug to facilitate its absorption through the gastrointestinal tract. Tenofovir was first licensed as the prodrug TDF in 2001, and the newer prodrug TAF became available in 2016. TAF is more stable in the plasma compared to TDF, and is administered at approximately one-tenth the TDF dose, resulting in lower systemic TFV exposures [15]. TAF may be administered at a strength of 10 mg in combination with cobicistat for boosting or 25 mg either unboosted or with cobicistat or ritonavir boosting. TAF is available co-formulated in fixed dose combinations with several other ARV medications – Table 1. The once daily FDC of RPV/TAF(25 mg)/FTC (Odefsey^®, Gilead Sciences) [21] is indicated for treating HIV-1 infections without known resistance mutations to NNRTIs, tenofovir or emtricitabine, and with a baseline viral load ≤ 100,000 copies/mL [21] The tablet must be taken with food. Since lower exposures of rilpivirine were observed in prior PK studies involving pregnant women on a rilpivirine based regimen [22,23], pregnant women on TAF/FTC/RPV should be monitored very closely for viral breakthrough. The BIC/TAF(25 mg)/FTC fixed dose combination (Biktarvy^®, Gilead Sciences) [24] is approved for the treatment of adults and adolescents living with HIV-1 (Table 1), but this product is not recommended for use in pregnancy because there are currently no pregnancy PK or safety data. TAF comes co-formulated with the booster cobicistat as DRV/c/TAF(10 mg)/FTC (Symtuza^®, Janssen) [25] and EVG/c/TAF(10 mg)/FTC (Genvoya^®, Gilead Sciences) [26], but these products are not recommended for use in pregnancy [27-29] due to the failure of cobicistat to adequately boost protease or integrase inhibitors [30,31]. TAF is also available as the 2 drug combination TAF (25 mg)/FTC (Descovy^®, Gilead Sciences) and as the single agent TAF (25 mg) (Vemlidy^®, Gilead Sciences) for use in combination with additional ARVs.

Table 1.

Currently available Tenofovir Alafenamide (TAF) Fixed dose drug combinations and their use during pregnancy.

Drug Name	Formulation	Company	FDA Approval date	Relevant Clinical PK studies in pregnancy	Use in Pregnancy
TAF/FTC/RPV (Odefsey)	TAF – 25 mg FTC – 200 mg RPV – 25 mg	Gilead Sciences	March 2016	*IMPAACT P1026 s (Momper et al.) [63]	Insufficient data to use during pregnancy. Monitor renal function in pregnancy
TAF/FTC (Descovy)	TAF – 25 mg FTC – 200 mg	Gilead Sciences	April 2016	*IMPAACT P1026 s (Brooks et al.) [64]	Insufficient data to use during pregnancy. Monitor renal function in pregnancy
TAF/FTC/EVG/COBI (Genvoya)	TAF – 10 mg FTC – 200 mg EVG – 25 mg COBI – 150 mg	Gilead Sciences	November 2015	IMPAACT P1026 s (Momper et al) [63] #PANNA (Schalkwijk et al.) [36]	Not currently recommended for use during pregnancy due to low cobicistat exposures during the second and third trimester of pregnancy [33–37]
TAF/FTC/DTG	TAF – 25 mg FTC – 200 mg DTG – 50 mg	Mylan	February 2018	IMPAACT 1026s	Insufficient data to use during pregnancy.
TAF/FTC/DRV/COBI (Symtuza)	TAF – 10 mg FTC – 200 mg DRV – 800 mg COBI – 150 mg	Janssen	July 2018	IMPAACT P1026	Not currently recommended for use during pregnancy due to low cobicistat exposures during the second and third trimester of pregnancy[33–37]
TAF/FTC/BIC (Biktarvy)	TAF – 25 mg FTC – 200 mg BIC – 50 mg	Gilead Sciences	March 2019	IMPAACT 2026 s and Gilead GS-US-380-5310 (NCT03960645)	Insufficient data to use during pregnancy. Data to be studied in ^IMPAACT P2026s. Monitor renal function in pregnancy

^*International Maternal Pediatric Adolescent AIDS Clinical Trials – IMPAACT P1026 s, an ongoing, non-randomized, open-label, multi-center study of antiretroviral PK in pregnant women living with HIV in the United States, Brazil, Thailand, and Africa.

^{^}International Maternal Pediatric Adolescent AIDS Clinical Trials – IMPAACT P2026 s, a non-randomized, open-label, multi-center study of antiretroviral PK in pregnant women living with HIV in the United States, Brazil, Thailand, and Africa. The PK study arms will open in 2020.

^#Pharmacokinetics of newly developed ANtiretroviral agents in HIV-positive pregNAnt women (PANNA), an ongoing, non-randomized, open-label, multi-center study of antiretroviral PK in pregnant women living with HIV in Europe.

3. Chemistry, pharmacokinetics profile of TAF

3.1. Chemistry and mechanism of action

TAF (GS-7340; C₂₁H₂₉N₆O₅P) is an acyclic purine nucleotide phosphonate prodrug of TFV [32]. Following oral administration and absorption, TAF enters the systemic circulation as a prodrug, and undergoes very little hydrolysis to TFV in plasma due to its unique molecular properties that confer stability to its chemical structure [33]. TAF subsequently enters cells passively, including lymphoid tissue and PBMCs [34], where it is hydrolyzed to TFV and phosphorylated by intracellular kinases to tenofovir-monophosphate (TFV-MP), and then its active metabolite, TFV-DP [35] – Figure 1. Cathepsin A (CatA), a lysosomal carboxypeptidase, is the key enzyme responsible for the first hydrolysis step of TAF in PBMC [36,37]. TFV-DP inhibits reverse transcriptase by competing with endogenous nucleotides (2’-deoxyadenosine triphosphate – dATP) for inclusion into HIV viral DNA, causing premature DNA chain termination. TFV-DP also has minimal effects on DNA polymerases alpha, beta, and mitochondrial DNA polymerase, and thus is associated with less mitochondrial toxicity than older NRTIs [36,38].

Figure 1.

Tenofovir Alafenamide (TAF) absorption, metabolism and elimination.

3.2. Absorption, volume of distribution, drug transport, distribution and half-life

TAF is 80% protein bound to plasma proteins [39], with an apparent volume of distribution of over 100 liters [40] and a blood-to-plasma ratio of 1.0 [39]. TAF is rapidly absorbed from the gastrointestinal tract after oral administration, reaching peak concentrations between 0.5 to 2 hours post-dose [39]. TAF is a substrate of P-glycoprotein (P-gp) and breast cancer resistance protein (BCRP) [41]. Therefore, drugs that inhibit P-gp or BCRP (e.g., cobicistat or ritonavir) increase TAF absorption. Similarly, medications that induce P-gp or BCRP can decrease TAF absorption, resulting in decreased plasma concentrations and potential loss of therapeutic effect or resistance. After absorption, TAF is taken up into the liver via hepatic uptake organic anionic transporters family 1B1 and 1B3 (OATP1B1 and OATP1B3), and circulates largely intact in the blood where it loads PBMCs [42]. The half-life of TAF in plasma is ~30 minutes [15], while the half-life of its active anabolite, TFV-DP, in PBMCs is approximately 4 days [43,44]. TAF achieves ~2-7-fold higher intracellular TFV-DP concentrations in PBMCs than TDF due to its increased stability in blood and higher lipophilicity in comparison to the parent tenofovir form [14,16,45].

3.3. Metabolism and excretion

Over 80% of orally administered TAF is metabolized intracellularly in PBMCs and hepatocytes [12,46]. The cytochrome P450 enzymes (mainly CYP3A) minimally metabolize TAF. TAF is converted to tenofovir intracellularly in the liver by carboxylesterase type 1 (CES1) and in white blood cells (PBMCs and macrophages) by CatA, and then undergoes additional phosphorylation steps to TFV-DP [33]. TAF is excreted primarily in the feces (32%). TAF is not a substrate for the renal transporters OAT1 and OAT3, unlike tenofovir, and thus less than 1 percent is excreted via the kidneys [47]. Lower plasma TFV concentrations are achieved with TAF, and parent tenofovir is excreted through the kidneys. The lower plasma tenofovir levels achieved with TAF are responsible for the reduced renal and bone complications when compared to TDF [14]. Due to its low renal excretion and low plasma TFV exposures, TAF has been used safely in patients with chronic renal disease with glomerular filtration rates of <30 mL/min [48-52]. In addition, declines in renal function are less common with TAF than TDF [53].

4. Pharmacodynamics – antiviral activity of TAF and barrier to drug resistance

Pharmacodynamic variables of importance in measuring dose-response and viral susceptibility to TAF include concentrations that produce 50% or 90% of maximal antiviral activity (EC₅₀ or EC₉₀) [54]. The established plasma EC₅₀ for HIV-1 for TAF and TDF are 0.005 μM, and 0.05 μM respectively [15,34], demonstrating that TAF is 10 times more potent than TDF [34]. These pharmacodynamic parameters help predict and define TAF’s potency, efficacy and barrier to resistance. Another way to assess TAF’s potency is by measuring the logarithmic decline in HIV-RNA levels (viral decay) when TAF is used in the management of HIV. Ascending TAF doses demonstrated median declines in HIV-1 RNA of 1.08 log₁₀ copies/mL, 1.46 log₁₀ copies/mL, and 1.73 log₁₀ copies/mL from baseline to day 11 after treatment with 8 mg, 25 mg, or 40 mg of TAF, respectively, as compared with a viral decay of 0.97 log₁₀ copies/mL with 300 mg TDF. The enhanced viral decays with TAF 25 mg and 40 mg doses was due to the higher intracellular concentrations of TFV-DP achieved with TAF versus TDF [16]. While TAF’s effective plasma concentration (EC₅₀ or EC₉₀) and viral decay as measures of potency are critically important, the duration of drug effect and barrier to antiviral resistance are also very crucial measures of antiviral effect. TAF was shown to have a more rapid attainment of protective levels in PBMC and, a longer duration of effect above the EC₉₀ following drug discontinuation in comparison TDF (16 vs. 10 days, respectively) [55]. TAF has a higher genetic barrier to treatment-emergent resistance mutations such as K65 R, Q151 M, T69-insertion complex, and multiple thymidine analog resistance mutations (TAMs) in comparison to TDF, again owing to the high TFV-DP concentrations attained in PBMCs [56].

5. Physiological changes during pregnancy relevant to TAF disposition

Several physiologic changes during pregnancy may impact the PK of TAF through alterations in drug absorption, distribution, metabolism and elimination compared to the non-pregnant state (Figure 1). TAF absorption may be affected by increased residence time of food in the gastrointestinal tract and increased gut pH due to the smooth muscle relaxant effect of progesterone production by the placenta during pregnancy [57]. TAF is transported across the gastrointestinal wall by passive diffusion, but is also a substrate for the efflux transporters, P-gp and BCRP, which can decrease its absorption [41]. Inhibition of P-gp by cobicistat or ritonavir reduces P-gp–mediated TAF efflux, thereby increasing the fraction of TAF absorbed. While the absorption and absolute bioavailability of TAF have not been studied in pregnant women, these parameters may differ during pregnancy due to these collective physiologic changes.

Due to the increased plasma volume and reduced albumin and alpha-1-acid glycoprotein concentrations during pregnancy [58], the protein-binding of TAF is expected to be reduced during pregnancy, as approximately 80% of TAF is bound to plasma proteins, increasing the free fraction available for clearance. TAF is a substrate for hepatic OATP1B1 and OATP1B3 [42]. While the expression of OATP1B3 decreased in women with intrahepatic cholestasis of pregnancy relative to normal pregnancy, the expression of OATP1B1 was relatively unchanged [59]. Following hepatic uptake, TAF is hydrolyzed in hepatocytes by CES1 and metabolized by CYP3A to a smaller extent. Hence, the increased activity of CYP3A enzymes that occurs during pregnancy is not expected to have a significant effect on the clearance of TAF or its metabolites. Pregnancy PK studies of drugs metabolized by CES1 have demonstrated that CES1 activity is unchanged during pregnancy [60].

Expression of CatA, the enzyme that activates TAF in PBMCs and macrophages to TFV-DP, increases during pregnancy and postpartum [61]. Increased CatA levels during pregnancy would likely increase the conversion of TAF to TFV-DP. Renal elimination of TFV occurs through a combination of glomerular filtration and tubular secretion via uptake through OAT1 and OAT3 renal transporters [62,63] and efflux into urine via MRP-4 [64]. Pregnancy decreases plasma TFV concentrations with TDF in the second and third trimesters through increases in plasma volume, and increased glomerular filtration rates owing to increases in renal blood flow. In contrast to TFV, TAF is not eliminated by OAT1 and OAT3 kidney transporters. Hence, TAF does not concentrate within the proximal tubules of the kidneys [65]. However, TAF is the predominant moiety circulating in blood and loading PBMCs, and so the relevance of pregnancy-related increased TFV clearance is unclear. Placental expression of P-glycoprotein on the microvilli of syncytiotro-phoblasts decrease with increasing gestational age from placental studies, and this has implications for maternal to fetal transfer of TAF [66].

6. TAF pharmacokinetic studies during pregnancy

6.1. IMPAACT P1026 s PK study of TAF 25 mg unboosted and TAF 10 mg boosted with COBI [67]

The International Maternal Pediatric Adolescent AIDS Clinical Trials (IMPAACT) P1026 s study was a non-randomized, open-label, multi-center study that evaluated the PK of ARV medications in pregnant women living with HIV. The first PK data from TAF in pregnant women came from this study [67]. Steady state plasma PK profiles of TAF following once-daily dosing of either TAF/FTC/RPV (25/200/25 mg, Odefsey^®) or TAF/FTC/EVG/COBI (10/200/150/150 mg, Genvoya^®) were obtained during the 2nd and 3rd trimesters of pregnancy, and 6–12 weeks postpartum. Data were available in 31 women taking TAF 25 mg without boosting, and 27 women taking TAF 10 mg boosted with cobicistat. For the 25 mg unboosted arm, TAF exposures were 43% lower during the second trimester versus postpartum [(GMR 0.57 (90% CI 0.34–0.98)], and 34% lower during the third trimester compared to postpartum [(GMR 0.66 (90% CI 0.54–0.82)]. Of note, these postpartum exposures were higher than typical non-pregnant adult values – Table 2. For the 10 mg TAF cobicistat-boosted arm, TAF exposures were 21% lower during the second trimester versus postpartum [(GMR 0.79 (90% CI 0.50–1.27)], and 14% lower during the third trimester compared to postpartum [(GMR 0.86 (90% CI 0.66–1.12)], but these differences were not statistically significant – Table 2 [67]. Though pregnancy and postpartum differences in TAF levels were observed between the 2 dosing regimens, plasma TAF exposures during pregnancy and postpartum with both regimens were within the range of those typically observed in non-pregnant adults. Furthermore, 10/11 women (91%) in the TAF 25 mg arm (without boosting) and 24/27 women (89%) in the TAF 10 mg with cobicistat arm had suppressed viral loads (<50 copies of HIV-1 RNA/mL) at the time of delivery [67].

Table 2.

Steady state Pharmacokinetics of Once-Daily Oral Administration of Tenofovir Alafenamide (TAF): ^aTAF (25 mg unboosted) in pregnancy, ^bTAF (10 mg boosted) in pregnancy, ^cTAF(25 mg boosted) in pregnancy, ^dTAF (25 mg unboosted) in non-pregnant adults, and ^eTAF (10 mg boosted) in non-pregnant adults.

	a25 mg unboosted			b10 mg with COBI			c25 mg with RTV or COBI
Dose	2nd Trimester	3rd Trimester	Postpartum	2nd Trimester	3rd Trimester	Postpartum	2nd Trimester	3rd Trimester	Postpartum	d25 mg unboosted (Non-pregnant adults)	e10 mg with COBI (Non-pregnant adults)
TFV AUCtau (ng•h/mL)	162 (153–184)	184 (147–350)	390 (188–461)	197 (145–354)	209 (158–284)	213 (169–304)	133 (128–720)	335 (192–549)	507 (221–693)	273 (227–319)	271 (213–329)
TFV Cmax (ng/mL)	69.7 (61.0–87.7)	91 (50–105)	157 (92–241)	80.4 (57.5–121.0)	93 (51–136)	96 (74–110)	44 (41–219)	101 (78–119)	164 (107–337)	7.9 (6.4–9.5)	7.4 (6–8.8)
TFV CL/F (L/hr)	154 (136–163)	136 (72–170)	64 (55–133)	127 (71–172)	120 (88–158)	118 (83–148)	188 (35–195)	75 (46–130)	49 (37–123)	-	-
TFV T1/2 (hours)	0.25 (0.24–0.27)	0.28 (0.23–0.53)	0.32 (0.30–0.43)	0.36 (0.25–0.52)	0.35 (0.27–0.48)	0.30 (0.27–0.44)	0.20 (0.20–0.33)	0.32 (0.26–0.61)	0.31 (0.25–0.66)	0.40 (0.36–0.43)	0.46 (0.38–0.53)

^*Summary PK reported as median (IQR).

Source:

^{a, b}Momper JD, Best B, Wang J, et al. Tenofovir Alafenamide Pharmacokinetics With and Without Cobicistat in Pregnancy. 22nd International AIDS Conference; July 23–27, 2018, 2018; Amsterdam, the Netherlands.

^cBrooks K, Pinilla M, Shapiro D, et al. Pharmacokinetics of tenofovir alafenamide 25 mg with PK boosters during pregnancy and postpartum. Oral abstract presented at 20th International Workshop on Clinical Pharmacology of HIV, Hepatitis, and Other Antiviral Drugs; 14–16 May 2019, 2019; Noordwijk, the Netherlands

^{d, e}Yamaha H, Yonemura T, Nemoto T et al. Pharmacokinetics of Tenofovir Alafenamide, Tenofovir and Emtricitabine following coformulated Emtricitabine/Tenofovir Alafenamide in Healthy Japanese Subjects. Clin Pharmacol Drug Dev, 2019; 8(4):511–520

Key to Acronyms: AUC_tau = area under the curve over the dosing interval (i.e., 24 hours); CL/F = apparent oral clearance; C_max = maximum plasma concentration; T_1/2 = half life; TAF = tenofovir alafenamide

6.2. IMPAACT P1026 s PK study [68] of TAF 25 mg boosted with ritonavir (RTV) or COBI

The PK and safety of TAF 25 mg with ritonavir or cobicistat boosting were also examined in a separate arm of the P1026 s study [68]. Seventeen women were enrolled during the second or third trimesters of pregnancy, and were concomitantly taking either darunavir/cobicistat, darunavir/ritonavir, atazanavir/cobicistat, or atazanavir/ritonavir. PK data were available from six, fourteen, and eight women during the second trimester, third trimester, and postpartum respectively. TAF exposures did not significantly differ during the third trimester compared to postpartum [(GMR 0.94 (90% CI 0.38–1.33)] – Table 2 [68], and exposures were comparable to or higher than historical data in adults receiving TAF 10 mg with cobicistat (Table 2). Importantly, only two women had paired data available during the second trimester and postpartum, so GMR comparisons were limited to third trimester versus postpartum in a total of eight women while awaiting additional data. A total of 16/17 women (94.1%) had suppressed HIV-1 viral loads at the time of delivery. No major safety concerns were noted. The full manuscripts and publications of the three TAF in pregnancy PK studies are in progress.

7. Randomized clinical trials of TAF-based FDC during pregnancy and lactation

7.1. VESTED-IMPAACT 2010 randomized clinical trial [69]

The Virologic Efficacy and Safety of combined antiretroviral therapy with TAF/TDF, EFV, and DTG (VESTED) trial [69] is an ongoing National Institute of Health (NIH)/IMPAACT funded phase III randomized controlled trial (NCT03048422). IMPAACT 2010 is comparing the virologic efficacy and safety of three ARV regimens (TAF/FTC/DTG, TDF/FTC/DTG, and TDF/FTC/EFV) in over 600 ARV-naïve pregnant women living with HIV [69]. It is the first phase III clinical trial to randomize pregnant women specifically to a TAF-based regimen to study HIV virologic suppression. The study will also compare the safety of these ARV regimens among pregnant women and their infants. At study entry, pregnant women are randomly assigned 1:1:1 to receive one of three regimens: TAF/FTC/DTG (arm 1), TDF/FTC/DTG (arm 2), or TDF/FTC/EFV (arm 3) during pregnancy and through one year postpartum. The primary outcome of the trial is the proportion of pregnant women living with HIV with viral loads < 200 copies/mL at the time of delivery. The study will also compare rates of adverse pregnancy outcomes, including maternal and infant adverse events across all three arms. The current clinical sites include the United States, Botswana, Brazil, India, South Africa, Tanzania, Thailand, Uganda, and Zimbabwe. Estimated completion date is July 31^st, 2020.

8. TAF drug-drug interactions

TAF plasma concentrations can be altered by concomitant administration of other medications. TAF is a substrate of BCRP and P-gp transporters, and absorption of TAF increases when it is co-administered with inhibitors of these efflux transporters (e.g., cobicistat). Conversely, use of TAF with a BCRP/P-gp inducer such as rifampicin may reduce TAF concentrations. Concomitant use of other ARVs can also affect the PK of TAF in non-pregnant adults [41]. TAF and TFV concentrations are unaffected after co-administration with RPV, and ~20% higher with DTG. Co-administration with BCRP/P-gp inhibitors such as cobicistat or ritonavir (as part of boosted protease inhibitor regimens) can result in marked increases in TAF and TFV exposure. However, the net effect of boosted PIs on TAF varies between the specific booster and PI used. Therefore, drug interactions between TAF and other ARVs, as well as drugs that induce or inhibit BCRP/P-gp are important to assess, as these drug interactions may further alter TAF exposures in this population.

9. Cobicistat boosting of TAF versus cobicistat boosting of protease/integrase inhibitors containing FDCs

Although cobicistat effectively boosts TAF exposure during pregnancy (section 6), it does not effectively boost exposures of the integrase inhibitor elvitegravir or the protease inhibitors darunavir or atazanavir (section 2) during pregnancy. There is a unique difference between the site and mechanism of action of cobicistat boosting of TAF and cobicistat boosting of protease and integrase inhibitors [28,29]. Cobicistat boosts TAF plasma exposures by inhibiting efflux transporters (P-gp and BCRP) in gut enterocytes, enhancing bioavailability. However, cobicistat boosts protease and integrase inhibitors mainly by selectively inhibiting CYP3A4 metabolism in the liver and intestinal tract, and by inhibiting efflux transporters (to a smaller extent) [70]. Cobicistat inhibition of CYP3A4 is dependent on plasma cobicistat concentrations, which are reduced in pregnancy. This difference in site and mechanism of action likely explains the difference in effectiveness of cobicistat boosting of TAF in pregnancy compared to protease and integrase inhibitors. Since TAF 10 mg with cobicistat is available only in FDCs with either darunavir, atazanavir or elvitegravir, the ability of cobicistat to effectively boost TAF 10 mg during pregnancy is not clinically relevant, as use of these FDCs in pregnancy is not recommended due to the low exposures of darunavir, atazanavir and elvitegravir during pregnancy. The IMPAACT P1026 s data suggest that the TAF exposures are adequate when TAF is used in pregnancy as 25 mg with or without boosting and 10 mg with cobicistat.

10. Intracellular TFV-DP data in pregnancy are limited

Intracellular TFV-DP data during pregnancy are currently limited to dried blood swab (DBS) assessments with TDF. DBS levels of TFV-DP with TDF were lower in pregnant women on HIV pre-exposure prophylaxis (PrEP) in comparison to the postpartum period [71]. A separate study in pregnant women taking TDF/FTC for PrEP under directly observed therapy is currently being conducted to establish DBS thresholds associated with 100% adherence in this population [72]. The Promoting Maternal Infant Survival Everywhere (PROMISE) [73] randomized clinical trial did not identify associations between higher TDF exposure, as measured by maternal TFV-DP concentrations in DBS, and adverse maternal, fetal and neonatal outcomes [73]. DBS levels measured in PROMISE were also lower than those measured in other studies in non-pregnant adults living with HIV [74]. Several studies have consistently demonstrated lower plasma TFV [14,15] and higher intracellular TFV-DP concentrations in PBMCs in non-pregnant adults on TAF-based regimens compared to TDF [14,16,55,75-77]. There are currently no data describing TFV-DP intracellular levels with PBMCs or DBS in pregnant, postpartum and lactating women on TAF. However, these assessments are planned in IMPAACT 2026. As discussed above (sections 6.1 and 6.2), although plasma TAF concentrations are lower during pregnancy than postpartum, they remain close to those found in non-pregnant women [67]. While the three different TAF dosing arms of IMPAACT 1026 s in pregnant women showed acceptable plasma TAF levels, viral load data at the time of delivery are sparse and challenging to interpret as they reflect efficacy of the entire ARV regimen and not just TAF. The majority of women – 91% (50/55) had HIV-1 viral suppression across all three P1026 s arms [67,68], which suggests that adequate TFV-DP levels in PBMCs are likely achieved. However, additional studies to establish intracellular TFV-DP concentrations in PBMCs and DBS with TAF in pregnant women are needed to inform efficacy, safety, and adherence assessments in this population.

11. TAF use and the risk of congenital anomalies and metabolic complications

There is a dose-response relationship between most drugs and the risk of adverse effects and congenital anomalies [78-80]. TDF exposure during pregnancy has been associated with reduced neonatal whole-body bone mineral density, decreased mean length-for-age Z-scores, and lower head circumference-for-age Z scores at one year of age in children enrolled in the Surveillance Monitoring for ART Toxicities (SMARTT) cohort [81], but these findings were most likely of uncertain significance. Hence, TDF is considered safe in pregnancy and still one of the ARV back-bones used during pregnancy. While TAF results in lower systemic concentrations of TFV, it produces higher intracellular concentrations of TFV-DP than TDF. It remains uncertain if higher intracellular TFV-DP concentrations would increase the risk of congenital anomalies and adverse pregnancy outcomes. The HIV Antiretroviral Pregnancy Registry (APR) is a project established to monitor prenatal ARV exposures and detect potential increases in the risk of teratogenicity [82]. While results from the APR are increasing bit by bit, there remains a paucity of data on TAF use during pregnancy and lactation, making this Registry an essential component of the ongoing program of epidemiologic studies of the safety of TAF-based FDCs [83]. In the IMPAACT P1026 s TAF 25 mg and 10 mg boosted with cobicistat study arms, congenital anomalies considered possibly related to study drugs included a ventral septal defect (VSD) in one infant and congenital pseudo-arthrosis of the left clavicle and neonatal compartment syndrome in another infant. Since this study arm involved a small number of women receiving TAF FDCs, it is difficult to conclude if these congenital anomalies were related to TAF or other ARVs in the FDCs, or were incidental findings. The number of cases related to TAF reported in the APR is insufficient to draw any reasonable conclusions on the association between TAF and any congenital anomalies at the current time. As more data become available, additional information on possible adverse effects and risk of teratogenicity will be gathered.

Recent studies in non-pregnant adults have linked TAF with increases in body weight and metabolic abnormalities compared to TDF [84,85]. Data from the AMBER randomized trial with cobicistat-boosted darunavir showed that participants randomized to TAF-based regimens had a greater increase in body weight compared to TDF-based regimens [86]. Other studies have shown similar findings of increased weight gain with TAF-based FDCs. Early data from the ADVANCE randomized controlled trial suggest that exposure to TAF-FDCs led to progressive increases in weight gain, serum lipid parameters (elevated triglycerides), low high-density lipoprotein (HDL) levels, hypertension, and hyperglycemia (metabolic syndrome) in patients during a 96-week period [87]. TAF-FDC associated metabolic syndrome was demonstrated in 9% of participants versus 3–5% of individuals taking TDF-based FDCs. This weight gain was more pronounced in women compared to men, with an average of 6 kg weight gain in men compared to 9 kg weight gain in women[87]. The impact of these findings and the implications during pregnancy remain unknown.

12. Conclusions

Plasma TAF exposures during pregnancy are within the typical range of those in non-pregnant adults taking similar doses, but higher than expected plasma exposures of TAF were noted postpartum for unclear reasons. No pregnancy data describing intracellular concentrations of TFV-DP, the active moiety of tenofovir, are available. Pending further PK, safety and efficacy studies, the US DHHS Perinatal Guidelines do not recommend TAF for use in pregnancy [7]. Prospective, comparative virologic response, intracellular PK, and safety data are needed to establish the role for this drug in pregnancy.

13. Expert opinion

Early pregnancy data from the PK studies as described in sections 6.1 and 6.2 above are suggestive that TAF based FDCs hold promise as part of first-line regimens for the treatment of pregnant women living with HIV and the prevention of perinatal HIV transmission due to their limited reports of adverse effects. Cobicistat (a P-gp inhibitor) may explain some of the differences in findings between boosted and unboosted TAF PK during pregnancy. There are several research gaps that, if filled, could aid our understanding of TAF use in pregnancy – the population PK of TAF during pregnancy (including sources of variability and transfer into and biotransformation in cells), placental drug efflux transporters in TAF disposition, and the utility of DBS and PBMCs to serve as measures of medication adherence and site-of-action drug exposure, respectively, during pregnancy.

Population PK of TAF use during pregnancy has not been described, as current TAF population PK models exclude pregnant women [88,89]. Pop-PK models are important because they incorporate data from sparse PK sampling and can explore the effects of multiple covariates (for example, maternal height, age, maternal and fetal weight, disease status, serum creatinine, hemoglobin concentrations, genetic polymorphisms and gestational age) on inter-individual and residual variability unexplained by non-compartmental pharmacokinetics [90]. A two-compartment population PK model of TDF in pregnant women demonstrated that only gestational age and serum creatinine significantly influenced plasma tenofovir disposition [91]. TAF population PK modeling in non-pregnant adults identified body weight and protein-binding as significant covariates on plasma TAF disposition [89]. A separate population PK study in non-pregnant women also linked plasma TDF and TAF with TFV-DP concentrations in various compartments, including PBMCs [88]. Just as the population PK models in non-pregnant adults demonstrated the effects of between-subject and unexplained residual variability on the PK of TAF and TFV-DP, similar approaches should be attempted in future pregnancy population PK models. The TDF population PK model in pregnant women (as with many pregnancy population PK models) was limited because it did not include the placental or fetal compartments, as fetal plasma drug concentrations can only be collected safely from the umbilical cord of term fetuses at the time of birth (and not earlier in pregnancy). In addressing these knowledge gaps, a future population PK study on TAF disposition during pregnancy, with appropriate covariate–parameter relationships and covariate stratification to prevent hidden biases, would be critically important to the prediction of TAF disposition across the different trimesters of pregnancy and postpartum. Gestational changes in placental drug transporter expression and activity remain an area of intense study [92], yet very little is known regarding their regulation during pregnancy. To address limitations of population PK in predicting in utero drug exposures, data on the expression of drug transporters in the placenta in combination with cord blood samples at birth could be incorporated into physiologically-based pharmacokinetic (PBPK) models [93]. It is of critical importance to further understand the role of placental P-gp, BCRP, as well as other drug transporters in regulating TAF, TFV and TFV-DP transfer to the developing fetus, and how these transporter functions are altered during pregnancy.

Given the recent data pointing toward metabolic concerns with TAF-based FDCs in non-pregnant adults, TAF use during pregnancy and the potential for metabolic changes need to be examined. Pregnancy is known to be associated with weight gain, especially during the second and third trimesters. It is unknown whether TAF use in pregnancy results in additional weight gain above the current Institute of Medicine (IOM) recommendations for pregnancy or an increased incidence of metabolic syndrome, which could increase adverse pregnancy outcomes in the short and long term. Understanding whether TAF-induced metabolic syndrome is modified during pregnancy by age, gestational age, parity, and race are critical research questions, especially in women with class III obesity (body mass index of ≥40 kg/m²). Answers to these questions are necessary before we can be confident that TAF can be used safely and effectively in pregnant women.

There remains a critical, unmet need for objective, quantitative measures of ARV adherence and drug levels at the site of action (i.e., PBMC) among women living with HIV in pregnancy. Non-pharmacologic adherence measures, such as patient self-report, calculation of proportion of pill days covered, pillbox checks, and use of electronic pill boxes/apps, have limitations and may over-estimate or under-estimate antiretroviral adherence [74]. To address these shortcomings, multiple pharmacologic adherence measures have been developed to objectively quantify antiretroviral medication adherence [94,95]. One approach is the use of DBS. TFV-DP has a long half-life in RBCs (~17 days) [43,96], and thus reflects cumulative medication adherence over the previous 2–3 months of therapy. TFV-DP in DBS has been used in several HIV pre-exposure prophylaxis (PrEP) [97–99] and treatment [100-103] studies to better understand adherence and several other treatment outcomes in the HIV field. Additionally, examining TFV-DP concentrations in PBMCs will provide critical insights as to whether changes in plasma levels of TAF or TFV result in clinically relevant changes in TFV-DP levels at the site of action. There are currently no data on TFV-DP concentrations in PBMCs during pregnancy. Intracellular TFV-DP data are limited to DBS concentrations with TDF in pregnant women living with HIV [73] and on PrEP [71,72]. Assessments of intracellular TFV-DP levels in DBS and PBMCs with TAF during pregnancy are developments likely to be clinically important in the future for understanding relationships with medication adherence and drug levels at the site of action.

Article Highlights.

• There is a unique difference between the site and mechanism of action of cobicistat boosting of TAF and cobicistat boosting of protease and integrase inhibitors. This difference in site and mechanism of action likely explains the difference in effectiveness of cobicistat boosting of TAF in pregnancy compared to protease and integrase inhibitors.

• Plasma TAF exposures during pregnancy are within the typical range of those in non-pregnant adults taking similar doses, but higher than expected plasma exposures of TAF were noted postpartum for unclear reasons.

• TAF has a higher genetic barrier to treatment-emergent resistance mutations such as K65R, Q151M, T69-insertion complex, and multiple thymidine analog resistance mutations (TAMs) in comparison to TDF, owing to the high TFV-DP concentrations attained in PBMCs

• TAF was shown to have a more rapid attainment of protective levels in PBMC and, a longer duration of effect above the EC90 following drug discontinuation in comparison TDF.

• TAF, due to its potency and good safety profile, might form the cornerstone for management of pregnant women living with HIV

This box summarizes key points contained in the article.