| |
Atazanavir plus low-dose ritonavir in pregnancy:
pharmacokinetics and placental transfer
|
| |
| |
AIDS:Volume 21(18)30 November 2007p 2409-2415
Ripamonti, Diegoa,b; Cattaneo, Darioe; Maggiolo, Francoa,b; Airoldi, Monicaa,b; Frigerio, Luigic; Bertuletti, Pierangeloc; Ruggeri, Mauriziod; Suter, Fredya
From the aDivision of Infectious Diseases, Ospedali Riuniti, Bergamo, Italy
bAntiviral Therapy Unit, Ospedali Riuniti, Bergamo, Italy
cDivision of Gynecology and Obstetrics, Ospedali Riuniti, Bergamo, Italy
dDivision of Pediatrics, Ospedali Riuniti, Bergamo, Italy
eMario Negri Institute for Pharmacological Research, Bergamo, Italy.
"As pregnancy does not appear to alter plasma exposure to atazanavir, no dose adjustment is required in pregnant women. Pharmacokinetic results suggest that a standard boosted atazanavir dose is a reasonable component of HAART during pregnancy..... We obtained a maternal blood/umbilical cord ratio of 0.10 that showed a clear positive linear correlation with maternal plasma levels. Consequently, even considering maternal Ctrough levels, the fetal exposure to atazanavir would fall into a therapeutic range approximately fourfold higher than wild-type HIV IC90. Whether significant plasma levels of antiretroviral drugs in the fetus could effectively decrease the risk of vertical transmission deserves further investigation. On the other hand, a limited, but still therapeutic, placental transfer of atazanavir may protect the fetus against the potential toxic effects of the drug. In this respect, we point out that none of the newborns presented with a clinically significant neonatal jaundice requiring phototherapy. These findings are consistent with recent data [15] in 33 women exposed to atazanavir/ritonavir during pregnancy, supporting its safety in this setting..... All women responded well to HAART...At the antepartum evaluation, as an example, the total bilirubin mean level was 2.58 mg/dl (SD ± 1.64), whereas the same postpartum value was 2.02 mg/dl (SD ± 1.04)..... None of the newborns presented with neonatal jaundice that required phototherapy for the treatment of hyperbilirubinemia. At birth all infants presented with an HIV-RNA level of less than 50 copies/ml and again tested negative 3 months after birth."
Abstract
Background: Adequate antiretroviral exposure during pregnancy is critical to prevent the vertical transmission of HIV and for maternal health. Pregnancy can alter drug kinetics. We assessed the pharmacokinetics of atazanavir/ritonavir (300/100 mg a day) during pregnancy.
Methods: An intensive steady-state 24-h pharmacokinetic profile of atazanavir was performed in the third trimester of pregnancy and postpartum. Maternal and umbilical cord blood samples were obtained at delivery. We measured atazanavir by reverse-phase high-performance liquid chromatography.
Results: Seventeen women completed the study. Antepartum, the atazanavir geometric mean area under the plasma concentration-time curve from 0 to 24 h (AUC0-24) was 28 510 ng„h/l, the maximum observed plasma concentration (Cmax) was 2 591 ng/ml and the 24-h postdose concentration (Ctrough) was 486 ng/ml. The same postpartum parameters were 30 465 ng„h/l, 2 878 ng/ml and 514 ng/ml, respectively. The antepartum to postpartum ratio for AUC0-24 was 0.94 and for Ctrough was 0.96, indicating equivalence, whereas Cmax values were slightly although not significantly lower. The ratio of cord blood/maternal atazanavir concentration in 14 paired samples was 0.13.
Conclusion: Atazanavir exposure during the third trimester of pregnancy is similar to that observed in the non-pregnant period. Over the whole dosing interval, therapeutic drug concentrations well above the wild-type HIV 90% inhibitory concentration are maintained. Atazanavir crosses the placenta, potentially providing further protection for the newborn. As pregnancy does not appear to alter atazanavir exposure, no dose adjustment is required in pregnant women. Results suggest that atazanavir is a reasonable component of HAART during pregnancy.
Discussion
This is the first study describing the pharmacokinetics of atazanavir in pregnant women. The results show that overall atazanavir exposure is similar in the pregnant and non-pregnant states after a standard boosted dose.
The antepostpartum within-patient comparison demonstrated no difference in the atazanavir AUC0-24 and Ctrough values. During pregnancy, however, a slightly lower Cmax value was observed. Several physiological changes occurring during pregnancy, including increased gastric emptying time, intestinal transit time, increased gastric pH and increased blood volume, may explain this difference [3,5]. We showed that during the third trimester of pregnancy mean Tmax values were slightly higher and much more variable than those observed postpartum. This observation suggests that delayed drug absorption may be the reasonable explanation for the difference in Cmax values.
Even though peak concentrations were lower during pregnancy and were not equivalent to those observed postpartum, they were not significantly different from the latter. Furthermore, this slight blunting of Cmax during pregnancy was not accompanied by a decrease in the extent of oral absorption, as indicated by the equivalent AUC0-24 values, and it seems unlikely to be clinically relevant.
The major pharmacokinetic problem faced by PI during pregnancy is reduced overall exposure. When compared with postpartum results, lower AUC and Ctrough values have been reported for boosted lopinavir (lopinavir/ritonavir) [5], boosted saquinavir (saquinavir/ritonavir) [11], indinavir [12] and nelfinavir [4]. Although the clinical implications of these altered concentrations in terms of the magnitude and durability of maternal virological response and the prevention of MTCT of HIV have been not completely addressed, the major risk is to expose pregnant women to subtherapeutic drug levels. Considerable data indicate that there is a significant association between antiretroviral drug concentrations and virological response, particularly for PI [13,14]. These relationships mean that antiretroviral drug exposure, especially the trough concentrations, should ideally be maintained above a defined threshold throughout the entire dosing interval in order to prevent viral replication and the development of resistant viral strains. Achieving adequate PI plasma levels would be particularly important in those patients who are first diagnosed with HIV infection during pregnancy and who may have on that occasion a very high viral load. Moreover, antiretroviral-experienced pregnant women may certainly benefit from higher drug plasma levels also in the third trimester to overcome some degree of viral resistance.
The mean atazanavir Ctrough of our women, in the third trimester of pregnancy, was 486 ng/ml, which was similar to the values reported in 16 pregnant women (373 ng/ml, range 71-1136 ng/ml) in the same time period [15]. The latter study did not, however, measure the Ctrough in the postpartum, so we cannot compare the mean atazanavir Ctrough (514 ng/ml) measured in our subjects at the postpartum evaluation. This latter measure postpartum is lower than expected for individuals on boosted atazanavir. In a study with both boosted and unboosted atazanavir, the mean Ctrough of 94 patients was 663 and 125 ng/ml, respectively [16]. Also another study reported a mean Ctrough of 862 ng/ml [17] for individuals on atazanavir boosted with ritonavir.
A number of reasons may provide an explanation for the lower than expected values found in our women. Seventeen subjects represent a relatively small sample size, and a large interpatient variability has been described for almost all PI, including atazanavir [17,18]. All that suggests the potential role of additional covariates in atazanavir pharmacokinetics. Genetic factors, such as polymorphisms at MDR1-3435, may significantly affect atazanavir plasma concentrations, even using ritonavir boosting [19]. In addition, ethnicity may have played a role as nine out of the 17 women in our group were of African origin. Possible differences in drug clearance between men and women have been described for saquinavir (approximately 50% decrease in saquinavir clearance in women) [11], although both sex and age failed to show any effects in atazanavir pharmacokinetics in previous reports [18]. In addition, although unlikely, we cannot rule out the potential bias caused by unreported co-medications, which may have negatively affected atazanavir plasma concentrations.
Finally, in our patients we did not measure ritonavir plasma levels, which are known to influence the pharmacokinetics of atazanavir greatly, as well as for almost all PI. One study [11] reported lower levels of ritonavir during gestation compared with the postpartum period in 13 women receiving boosted saquinavir (800/100 mg twice a day). Another study [5] in 11 pregnant women on lopinavir/ritonavir showed that ritonavir AUC and the 12-h postdose concentration were similar ante and postpartum, suggesting that the lower lopinavir concentration in late pregnancy is not secondary to reduced ritonavir exposure. Similarly, ritonavir Ctrough values were equivalent, during and off pregnancy, in seven women on lopinavir/ritonavir [20].
In our study Ctrough atazanavir levels, although presenting a quite high interpatient variability, were comparable in the pregnant and non-pregnant period and, most important, invariably allowed the calculation of an inhibitory quotient several fold greater than the atazanavir 90% IC for wild-type HIV, minimizing the risk of residual viral replication and the selection for resistant viral strains. All treated women achieved and maintained an undetectable viral load throughout the whole study period. Although limited in number, we obtained evidence that atazanavir was also effective in preventing MTCT.
From a pharmacokinetic point of view, this result may partly be explained by the transplacental passage of atazanavir. Several data have confirmed an overall low level of placental passage for PI. In line with the considerable intersubject variability seen for maternal levels [17,18,21], umbilical cord/maternal blood ratios vary substantially in different studies, mostly limited by the small sample size, and with possible differences in threshold limits or the methodological approach. In one study [22] involving 13 maternal cord blood sample pairs, the reported ratios were extremely low for all PI (nine patients on nelfinavir, two on ritonavir, one on saquinavir and one on lopinavir), most of them being below the limit of detection. Data from the PACTG 1026 study team [5] on lopinavir/ritonavir in pregnancy reported a ratio equal to 0.2 for the umbilical cord to maternal blood lopinavir concentration in 10 paired detectable samples. For three additional women, however, maternal delivery lopinavir concentrations were below the limit of detection. No measurable PI concentrations in cord blood have been reported in six mother/infant pairs receiving nelfinavir and in six receiving lopinavir/ritonavir [23]. In another study [24] cord blood concentrations were below the detectable threshold in 10 of 40 samples (25%), 25 of 40 (62%), nine of 11 (81%), four of six (66%) and five of six (83%) for nelfinavir, M8 metabolite, ritonavir, indinavir and saquinavir, respectively. Other authors [21] reported cord blood concentrations of lopinavir between 5 and 25% of the maternal plasma levels, averaging 16.7%. Only one single case [25] has been published on human placental transfer of atazanavir/ritonavir in a pregnant woman. At delivery, the cord blood atazanavir level was 362 ng/ml, whereas the maternal atazanavir level was 1515 ng/ml (ratio 0.23). As PI do not cross the placenta to an appreciable degree, the extent to which they can provide direct protection for the newborn is uncertain and probably modest. Reasons for the limited placental transfer are probably the high protein binding capacity of PI and their large molecular size [26].
We obtained a maternal blood/umbilical cord ratio of 0.10 that showed a clear positive linear correlation with maternal plasma levels. Consequently, even considering maternal Ctrough levels, the fetal exposure to atazanavir would fall into a therapeutic range approximately fourfold higher than wild-type HIV IC90. Whether significant plasma levels of antiretroviral drugs in the fetus could effectively decrease the risk of vertical transmission deserves further investigation. On the other hand, a limited, but still therapeutic, placental transfer of atazanavir may protect the fetus against the potential toxic effects of the drug. In this respect, we point out that none of the newborns presented with a clinically significant neonatal jaundice requiring phototherapy. These findings are consistent with recent data [15] in 33 women exposed to atazanavir/ritonavir during pregnancy, supporting its safety in this setting.
In summary, atazanavir overall exposure at steady state during the third trimester of pregnancy is similar to that observed in the non-pregnant state. Over the whole dosing interval, therapeutic drug concentrations are maintained well above the wild-type HIV IC90. Atazanavir crosses the placenta and this may potentially contribute to reduce the risk of vertical transmission further.
As pregnancy does not appear to alter plasma exposure to atazanavir, no dose adjustment is required in pregnant women. Pharmacokinetic results suggest that a standard boosted atazanavir dose is a reasonable component of HAART during pregnancy.
Introduction
In the developed world, HAART is the standard of care for the treatment of HIV infection. In pregnancy this strategy is employed to meet two goals: preservation of the health status of the woman and prevention of mother-to-child transmission (MTCT) of HIV. Combination regimens of nucleoside/nucleotide analogue reverse transcriptase inhibitors and either a non-nucleoside reverse transcriptase inhibitor or a protease inhibitor (PI) are commonly used in pregnant women. Which regimen should be preferred is, however, still unknown. Non-nucleoside reverse transcriptase inhibitor are either at risk of mutagenic effects [1] or have been correlated with severe adverse events in women [2]. More recently, significant reductions of drug concentrations of different PI have been reported during the third trimester of pregnancy [3-5]. This is a relevant piece of information with respect to PI because of their known concentration-response relationship. If pregnancy can significantly alter the levels of exposure to these drugs, it is critical to understand the effect of pregnancy on the absorption and disposition of different molecules of this drug class, in order to ensure the best therapeutic choices and that adequate concentrations are achieved in the blood of pregnant women, to prevent the development of viral resistance and vertical transmission of the infection.
The primary purpose of this study was to evaluate if a standard boosted atazanavir (atazanavir/ritonavir) dose produces adequate drug exposure during pregnancy compared with the non-pregnant state. A secondary objective was to compare atazanavir concentrations in plasma from cord blood with those in maternal plasma at the time of delivery. The cord/maternal blood ratio provides an estimate of the amount of drug that can cross the placenta, which may be potentially useful in further reducing the risk of vertical transmission.
Methods
Mother-to-child transmission programme
At our institution a structured MTCT programme has been implemented. All HIV-infected pregnant women receive HAART according to their clinical condition. HAART is started as soon as the HIV status is known (or continued in the case of a previously treated woman) if the woman's immunological status requires immediate treatment, otherwise HAART is delayed until the end of the first trimester of pregnancy. All women undergo elective caesarean resection. At birth the children receive one month zidovudine prophylaxis (2 mg/kg every 6 h). Breast feeding is avoided.
Within this programme we performed a pharmacokinetic, efficacy and safety evaluation of the use of atazanavir during pregnancy.
Patients
Consecutive patients were asked to participate in the trial. A fixed HAART combination was administered. It was based on the association of zidovudine plus lamivudine (as a fixed dose combination 300 mg plus 150 mg) twice a day plus a single daily dose of atazanavir (300 mg) with ritonavir boosting (100 mg). All patients presenting active viral replication before starting HAART performed a baseline genotype test.
All women gave their informed consent before starting the study, and the trial was approved by the bio-ethical committee of our institution.
Clinical and laboratory monitoring
Historical, demographic, clinical and laboratory data were collected at enrollment. The safety and efficacy of HAART were evaluated throughout the study period by means of clinical visits and laboratory tests performed every 2 months. Laboratory tests included alanine aminotransferase, aspartate aminotransferase, total and unconjugated bilirubin, creatinine, blood urea nitrogen, albumin and haemoglobin. CD4 cell count and plasma HIV-RNA concentrations were measured at the same timepoints. Infant gestational age and birth weight was recorded at delivery. Infants were monitored according to usual clinical practice. A HIV-RNA test was performed at birth and 3 months after birth.
Abnormal findings were characterized according to the Division of AIDS Standardized Toxicity Table for Grading Severity of Adult Adverse Experiences (August 1992; http://rcc.tech-res-intl.com ).
Sample collection
The physiological changes occurring during gestation and affecting drug disposition appear to be greater in late pregnancy. In order to compare atazanavir pharmacokinetics between the third trimester and the postpartum period, along with the umbilical cord/maternal blood atazanavir ratio, plasma collections were planned accordingly.
Plasma samples (5 ml) for pharmacokinetic evaluation were collected at three evaluation times: antepartum (during the third trimester of pregnancy), at delivery, and postpartum (between 1 and 6 months after delivery). After delivery the women continued to receive their HIV therapy until the following pharmacokinetic evaluation. This second evaluation time has been planned for all the women after at least 4 weeks from delivery in order to give sufficient time for the pregnancy-associated physiological changes to return to normal.
Patients were receiving a stable antiretroviral regimen for at least 3 months before any pharmacokinetic sampling. Patients were instructed to take their ritonavir-boosted atazanavir dose at the same time each day for the 3 days before the day of the ante and postpartum pharmacokinetic evaluation. Nine plasma samples were drawn at both the ante and postpartum pharmacokinetic evaluation visits, starting immediately before the daily dose and at 0.5, 1, 2, 3, 4, 8, 10, and 24 h after the witnessed dose. To assess transplacental passage, on the day of caesarean resection, women were given their usual morning dose and atazanavir concentrations were measured on two concomitant samples obtained from maternal and umbilical cord blood at delivery. All samples were frozen within an hour and stored (-80C) until further analysed.
Analytical and pharmacokinetic methods
The concentration of atazanavir in plasma were measured using a validated high-performance liquid chromatography method [6,7]. Briefly, 600 ml of each plasma sample were mixed with 50 ml of internal standard (diazepam, 450 mg/ml) and 100 ml methanol. Subsequently, the solution was mixed-vortexed for 30 s and then centrifuged at room temperature. After centrifugation, the clean supernatant was loaded onto a Bond-elute cartridge (C18, 200 mg, 3 ml; Varian, Lein, Italy) on a Vac Elut 20 Manifold (Varian). The cartridges were washed with a solution of methanol/water (5/95). Atazanavir and internal standard were eluted with 2 ml methanol. The eluate was taken to dryness and reconstituted in 150 ml methanol. Each sample was then transferred to polypropylene vials. Thirty millilitres of each sample were injected into the chromatograph. The separation of atazanavir from internal standard and endogenous components was achieved using an isocratic elution on an octyl column. The detector was set at 260 nm. The method showed a linear relationship between peak height ratios and blood concentrations in the range of 10-10 000 ng/ml (r2 = 0.9991). The observed intra and interday assay imprecision ranged from 2.2 to 14.7% (at the lower limit of quantification), whereas inaccuracy varies between 1.0 and 14%. The limit of quantification was set at 20 ng/ml.
The calculated pharmacokinetic parameters for atazanavir were the trough plasma concentration (Ctrough; defined as the 24 h concentration after the observed dose), the maximum observed plasma concentration (Cmax), the area under the plasma concentration-time curve from 0 to 24 h (AUC0-24), the time to reach the maximum plasma concentration (Tmax) and the elimination half life (T1/2). All these parameters were calculated using actual blood sampling times and non-compartmental modelling techniques. The AUC was calculated using the trapezoidal rule. The pharmacokinetic analyses were performed using NONMEM software (Globomax LLC, Hanover, Maryland, USA).
Statistical analysis
Descriptive statistics are presented as means ± standard deviation (SD), medians and range, percentages and 95% confidence intervals (CI). For pharmacokinetic parameters geometric least-squares means were also calculated. The magnitudes of difference in median or mean values of the pharmacokinetic parameters of interest at each study period were assessed using the Wilcoxon signed-rank test with the Student's t-test for paired data.
Antepartum and postpartum atazanavir measurements from each woman were compared using a repeated measure design. These comparisons were made at the within-subject level, using geometric mean ratios and 90% CI. A no effect window for the 90% CI was defined as 0.8-1.25. Therefore, if the entire 90% CI for the mean geometric ratio (the antilog of the true mean of the log ratios) was within these range limits, the pharmacokinetic exposure for the pregnant and non-pregnant conditions were considered equivalent [8].
Results
Patient characteristics
Eighteen women agreed to participate in the study. One women was withdrawn because the baseline genotypic test revealed the presence of a mutant viral strain (M184V mutation). The remaining 17 women completed the study follow-up. Their mean age was 30.7 years (SD ± 5.2 years), nine of them were of African origin, seven were Caucasian and one was Asian. Nine women had already been receiving HAART and five of them were currently on antiretroviral drugs. Eight women were naive to HAART, either because their immunological status did not require treatment or because they were diagnosed at routine pregnancy screening. As a consequence, HAART was started in all patients between weeks 0 and 26 (mean 12.3 ± 8.3 weeks) of pregnancy. At treatment initiation the mean CD4 cell count was 420 ± 169 cells/μl and the mean HIV-RNA blood level was 11 308 SD ± 17 194 copies/ml.
Pharmacokinetics
An antepartum pharmacokinetic evaluation was performed 38 ± 22 days (mean and SD) before delivery. At the time the evaluation was performed the mean body mass index (BMI) of the 17 women was 26.39 (SD ± 0.7), significantly (P < 0.0001) higher than that observed at the time of postpartum evaluation (BMI 24.02 ± 0.6) performed 71 ± 45 days after delivery.
The pharmacokinetic parameters were not significantly associated with the calculated BMI (data not shown).
A summary of pharmacokinetic ante and postpartum parameters is provided in Table 1. In no case was the difference between the ante and postpartum parameter statistically significant (Fig. 1). The geometric mean within-subject ante/postpartum AUC0-24 ratio was 0.94 with a 90% CI between 0.80 and 1.11. The CI value for this parameter ratio fell completely within the limits of 0.8 and 1.25, showing that the ante and postpartum AUC0-24 values were equivalent. The ante/postpartum Cmax ratio was 0.94 with a 90% CI from 0.74 to 1.11. As CI values for Cmax did not fall within the predefined limits (0.8-1.25), this did not support an equivalent Cmax value before and after partum. As far as the Ctrough is concerned, the geometric mean ratio was 0.96 with a 90% CI between 0.82 and 1.13, again indicating equivalence. Most important, when Ctrough values were compared with the 90% inhibitory concentration (IC90) for atazanavir against wild-type HIV (14 ng/ml) [9,10], they were constantly several fold higher. Therefore, the antepartum mean inhibitory quotient was 40-fold higher than the IC90 (95% CI 33-47) and the mean postpartum inhibitory quotient was 47-fold higher (95% CI 37-57).
Maternal plasma samples at delivery and umbilical cord blood samples were available for 14 mother/infant pairs. Detectable atazanavir concentrations were present in all umbilical cord plasma samples, and the mean ratio of cord/maternal concentration was 0.13 and showed a low variability (95% CI 0.10-0.16; Fig. 2).
Clinical monitoring and tolerability
All women responded well to HAART. HIV-RNA levels were invariably below the detection limit (50 copies/ml) for all subjects at antepartum, delivery and postpartum evaluations. HAART was well tolerated during pregnancy and postpartum with no women stopping/changing the regimen or experiencing adverse events. At the antepartum evaluation, as an example, the total bilirubin mean level was 2.58 mg/dl (SD ± 1.64), whereas the same postpartum value was 2.02 mg/dl (SD ± 1.04).
All 17 infants (nine boys, eight girls) were live born. Their mean gestational age was 37.2 weeks (SD ± 0.56) and their mean birth weight was 2.912 kg (SD ± 0.195). None of the newborns presented with neonatal jaundice that required phototherapy for the treatment of hyperbilirubinemia. At birth all infants presented with an HIV-RNA level of less than 50 copies/ml and again tested negative 3 months after birth.
|
|
| |
| |
|
|
|
