Placental Transfer of Rilpivirine in an Ex Vivo Human Cotyledon Perfusion Model

Abstract

Placental transfers of the HIV nonnucleoside reverse transcriptase inhibitor rilpivirine were investigated in 8 term human cotyledons perfused with rilpivirine (400 ng/ml) in the maternal-to-fetal direction. The mean fetal transfer rate (FTR) (fetal/maternal concentration at steady state from 15 to 90 min) was 26% ± 8% (mean ± standard deviation), and the clearance index (rilpivirine FTR/antipyrine FTR) was 61% ± 20%. This shows that rilpivirine crosses the placenta at a relatively high rate, suggesting that the fetus is exposed to the compound during treatment of the mother.

TEXT

Antiretroviral drugs are remarkably effective for preventing vertical transmission of HIV-1, resulting in a decline in the rate of transmission, which is now less than 0.5% in France (1, 2). Current guidelines recommend antiretroviral therapy for all HIV-infected pregnant women for their own health and to prevent transmission (3, 4). Rilpivirine (4-[[4-[[4-[(E)-2-cyanoethenyl]-2,6-dimethylphenyl]amino]-2-pyrimidinyl]amino]benzonitrile monohydrochloride), a diarylpyrimidine analog, is a nonnucleoside reverse transcriptase inhibitor (NNRTI) which is increasingly used to treat HIV infection and in recent guidelines is considered a first-line treatment. It is available in a single-tablet formulation in combination with tenofovir and emtricitabine (Complera-Eviplera). Rilpivirine is classified as a pregnancy category B drug by the Food and Drug Administration (4), which means that animal studies have failed to demonstrate a risk to the fetus, but there have been no adequate and well-controlled studies in HIV-1-infected pregnant women. To date, only one report has been published on two cases of rilpivirine use in pregnant women (5), in which the plasma concentrations of rilpivirine were lower than would be expected in nonpregnant women. Placental transfer was studied in one of these two cases. Currently, there are no data on the safety of rilpivirine use during pregnancy.

Determining fetal exposure is important for all new medications in order to estimate the potential for preexposure prophylaxis and the risk for toxicities to the fetus. A number of adverse events have been reported following perinatal exposure to antiretroviral medications, including mitochondrial diseases (6). Since data from animal studies are difficult to extrapolate to humans due to differences in placental physiology, other preclinical approaches are required.

The ex vivo human cotyledon is an accepted model used to study and interpret placental transfer (7, 8). The purpose of this study was to investigate the placental transfer of rilpivirine in an ex vivo perfused human cotyledon.

Placentas were collected after uncomplicated pregnancy with full-term delivery (≥37 weeks gestational age) in mothers who were seronegative for HIV and hepatitis B and C viruses and had taken no medication, except for vitamin supplements and perimedullar analgesia. They were collected in a single center, the maternity ward of the Hôpitaux Universitaires Paris-Nord Val de Seine (Colombes, France), and were rapidly perfused on site. Written informed consent was obtained from each of the women who donated a placenta, according to French bioethics guidelines (Article L1211-2 of the Public Health Code).

Rilpivirine was obtained by crushing commercial tablets (25 mg per tablet; Janssen, Issy-les-Moulineaux, France), after which the powder was weighed and prepared in tubes for each experiment. Antipyrine–phosphate-buffered saline and Bradford reagent were purchased from Sigma-Aldrich (Saint Quentin Fallavier, France), other reagents were purchased from Invitrogen (Cergy-Pontoise, France), and albumin was purchased from Baxter Laboratories (Deerfield, IL, USA).

Placentas were perfused according to a method adapted from Schneider et al. (9), as previously described (10–12). Each perfusion was started within 20 min after delivery. After selecting an intact isolated cotyledon, a fetal chorionic plate artery and its associated vein were cannulated. A lack of difference in pressure between the fetal artery and vein was required to confirm that the cotyledon was viable. The maternal side was then examined for whitening of the perfused area, and the selected cotyledon was placed in the perfusion chamber with the maternal side up. The intervillous space on the maternal side was perfused using two needles piercing the basal plate. Each circuit was pumped separately by a peristaltic pump (Minipuls 3; Gilson Medical Electronics, Villiers le Bel, France). The fetal and maternal flow rates were 6 and 12 ml/min, respectively. The system reached steady state after 15 to 30 min. The pH values of the maternal and fetal solutions, prepared with Earle medium plus 2 g/liter of human albumin, were adjusted to 7.4 ± 0.1 and 7.2 ± 0.1, respectively, and the temperature was maintained at 37°C. Rilpivirine and antipyrine were dissolved into the infusion reservoir on the maternal side. The rilpivirine concentrations were targeted to be within the range of clinical peak plasma levels (maximum concentration of drug in serum [C_max]). Antipyrine, an inert molecule which freely diffuses through the placenta, was added at 20 mg/liter as a marker to validate the cotyledon's viability throughout the experiment (13).

Samples were collected every 5 min to determine the concentrations of rilpivirine and antipyrine in the fetal and maternal compartments. They were then stored at −20°C until analysis.

Rilpivirine concentrations in maternal and fetal samples were determined using ultraperformance liquid chromatography coupled with tandem mass spectrometry as previously described (14). The lower limit of quantification for rilpivirine was 5 ng/ml. Antipyrine concentrations were determined by high-performance liquid chromatography with UV detection at 290 nm after liquid-liquid extraction. The analytic column for antipyrine separation consisted of an octadecylsilyl Nova-Pak (3.9 mm by 150 mm). The mobile phase comprised 0.05 M phosphate buffer (pH 3)–methanol-tetrahydrofuran (75:25:0.9 [vol/vol/vol]). Standard curves for antipyrine concentrations ranged from 0.5 to 20 mg/liter. The lower limit of quantification for antipyrine was 0.01 mg/liter. Maternal-to-fetal transfer parameters were calculated at steady state according to the formulas of Challier et al. (13). The fetal transfer rate (FTR) is the ratio of fetal to maternal concentrations, and the clearance index (CLI) is the ratio of the FTR of rilpivirine over the FTR of antipyrine. An antipyrine FTR of greater than 20% was required to validate each experiment.

Eight procedures were validated, and the mean (± standard deviation [SD]) FTR for antipyrine was 48% ± 14%. The results of the 8 experiments at steady state are presented in Table 1.

TABLE 1.

Placental transfer of rilpivirine in ex vivo-perfused human cotyledons

Perfusion time (min)	Placenta 1a				Placenta 2				Placenta 3				Placenta 4				Placenta 5				Placenta 6				Placenta 7				Placenta 8
	CM (ng/ml)	CF (ng/ml)	FTR (%)	CLI (%)	CM (ng/ml)	CF (ng/ml)	FTR (%)	CLI (%)	CM (ng/ml)	CF (ng/ml)	FTR (%)	CLI (%)	CM (ng/ml)	CF (ng/ml)	FTR (%)	CLI (%)	CM (ng/ml)	CF (ng/ml)	FTR (%)	CLI (%)	CM (ng/ml)	CF (ng/ml)	FTR (%)	CLI (%)	CM (ng/ml)	CF (ng/ml)	FTR (%)	CLI (%)	CM (ng/ml)	CF (ng/ml)	FTR (%)	CLI (%)
15	388	40	10	52	544	37	7	13	262	67	26	38	423	44	10	44	387	42	11	41	438	53	12	27	438	53	6	18	420	36	9	31
30	217	45	21	64	399	62	16	26	249	82	33	47	400	92	23	47	406	250	62	134	399	131	33	50	399	131	14	33	408	66	16	40
45	622	76	12	31	463	86	19	29	375	78	21	26	400	162	41	71	418	101	24	52	451	129	29	48	451	129	16	34	401	159	40	71
60	598	50	8	23	410	105	26	45	243	79	33	42	420	96	23	42	350	291	83	223	430	168	39	106	430	168	22	58	403	189	47	72
75	210	65	31	139	404	86	21	33	314	77	25	29	387	90	23	101	398	103	26	70	413	153	37	60	413	153	21	58	418	196	47	93
90	341	75	22	61	387	118	30	59	304	106	35	58	340	110	32	225	799	113	14	33	440	147	33	63	440	147	28	98	358	191	53	84
Mean	396	58.5	17	62	434.50	82.33	20	34	291.17	81.50	29	40	395	99	25	88	459.67	150	37	92	428.5	130.17	31	59	428.5	130.17	18	50	401.33	139.5	35	65
SD	179.78	15.60	9	41	59.75	29.26	8	16	50.32	13.03	6	12	30.16	38.09	10	71	167.85	97.49	29	74	19.19	40.48	10	26	19.19	40.48	8	28	22.59	70.41	18	24

^aCM, rilpivirine concentration in the maternal compartment; CF, rilpivirine concentration in the fetal compartment; FTR, fetal transfer rate; CLI, clearance index.

The concentration (mean ± SD) in the maternal compartment was 401 ± 31 ng/ml. The concentration in the fetal compartment was 101 ± 38 ng/ml. This concentration was well above the 50% effective concentration (EC₅₀) against wild-type HIV-1, which is 0.27 ng/ml (15). The FTR of rilpivirine was 26% ± 8%, and the clearance index (CLI) was 61% ± 20%.

With the ex vivo model, this study indicates that maternal rilpivirine significantly crosses the human placenta. This finding is consistent with that of the only case reported to date, where the cord blood/maternal blood ratio at delivery in a single patient was 0.74 (5). Similar cord blood/maternal blood ratios have been reported for other NNRTIs, including 0.67 for nevirapine (16) and 0.51 for etravirine (17). There are considerable differences in placental transfer rates among the various antiretroviral drugs, which are partly but not entirely related to the class, defined as the site at which they inhibit viral replication. The CLI values reported using this ex vivo placental perfusion model were 40.3% for darunavir (18) and 73% for lopinavir at a 2% serum albumin concentration (19), whereas there was nearly no placental transfer for saquinavir (10). For the nucleoside reverse transcriptase inhibitors, the fetal concentrations were equal to or higher than the maternal concentrations (20), whereas the CLI of the CCR5 inhibitor maraviroc was 26%, and the fusion inhibitor enfuvirtide did not cross the placenta (21).

This ex vivo model does have important limitations. First, it evaluates placental transfer at term and not during the entire pregnancy, which is also true for clinical cord blood data at delivery. Furthermore, ex vivo experiments do not entirely reproduce in vivo conditions, particularly regarding binding to serum proteins, which is important for compounds such as rilpivirine, which are highly protein bound (19). Nonetheless, the procedure is carefully monitored and standardized to approach mimicking normal physiology, and the FTR of antipyrine is monitored in order to control the integrity of the placental barrier. The results obtained with this ex vivo placental perfusion model have been found to be consistent with in vivo findings for a number of compounds, including antiretroviral agents (7, 8).

In conclusion, rilpivirine placental transfer appears sufficient to expose the fetus to drug concentrations near the therapeutic range. This may have some effect as preexposure prophylaxis but also carries a potential for adverse effects. Clinical studies are required to determine whether rilpivirine has a more favorable safety and efficacy profile than other NNRTIs for use in pregnant women.