Participants

Women (n = 17) were recruited from Magee-Womens Hospital (Pittsburgh, PA) and an outlying clinic. Participants were recruited and enrolled from June 1, 2014, through Nov. 30, 2014. All participants were receiving twice-daily buprenorphine maintenance therapy, as prescribed for clinical purposes by their respective caregivers, and were expected to be at steady state on their current dose prior to each study visit. The protocol was approved by the University of Pittsburgh’s Institutional Review Board, and all participants underwent the informed consent process using institutional review board-approved consent documents.

Procedures

Demographic details, baseline laboratory parameters, medication, and/or substance use history and obstetric history were collected for all participants. Eligibility criteria for these women included the following: (1) pregnant and on a stable, twice-daily dose of buprenorphine for at least 7 days, (2) ≥18 years of age, (3) able to willingly consent, and (4) willing to have urine samples screened for the presence of alcohol, barbiturates, opiates, cocaine (or metabolites), benzodiazepines, synthetic opioids, PCP, and concurrent medications or substances.

Exclusion criteria included the following: (1) hepatic or renal dysfunction, (2) sickle-cell disease and on active treatment, (3) HIV and on active treatment (potential drug interactions), (4) hypersensitivity to opioids, (5) comorbid dependence on benzodiazepines, (6) concurrently taking monoamine oxidase inhibitors, neuroleptics, or disulfiram, and (7) concurrently taking medications known to be CYP3A inducers (such as rifampin, phenobarbital, phenytoin, or carbamazepine) or CYP3A inhibitors (such as azole antifungals, macrolide antibiotics, or HIV protease inhibitors).

Up to 3 studies were performed in each participant. PK-2 studies were performed in the second trimester between 18 and 25 weeks (n = 7), PK-3 studies were performed in the third trimester between 31 and 37 weeks (n = 11), and PK-P was performed 4-18 weeks postpartum (n = 10). All PK study visits were identical and conducted in the Clinical and Translational Research Center of the University of Pittsburgh Medical Center’s Montefiore Hospital.

Participants arrived early in the morning after a ≥8 hour fast with their medication log, underwent a predose blood draw (BD Vacutainer glass blood collection tubes with sodium heparin; Becton Dickson Inc, Franklin Lakes, NJ) and oral fluid collection (Sarstedt salivette cotton swab for saliva collection; Sarstedt Inc, Newton, NC) for trough buprenorphine concentrations and clinical laboratory parameters; provided a urine sample for laboratory tests and toxicology; had oral pH recorded (Hydrion urine and saliva pH paper, range 5.5–8.0; Micro Essential Lab, Brooklyn, NY); and then took their prescribed sublingual buprenorphine dose under direct supervision by study staff.

Participants were instructed to allow the tablets to dissolve under their tongue, with minimal swallowing, until there was no visible residue remaining. Dissolution time was recorded after visual inspection of the sublingual area by study staff. Serial blood and oral fluid samples were subsequently collected over 1 dosing interval at 0 (before the dose), 0.25, 0.5, 0.75, 1, 1.5, 2, 4, 8, 10, and 12 hours after the dose. All blood samples and oral fluid salivettes were centrifuged for 15 minutes at 15,000 rpm to obtain plasma and oral fluid, respectively. Samples were immediately frozen at –80°C until analysis.

All spontaneously voided urine was collected throughout the entire study period of 12 hours. Participants were not allowed to eat until 2 hours after the dose to allow for adequate drug absorption, and all food consumed during the study day was reported on a dietary log.

Assay methodology

Buprenorphine (BUP) and its 3 active metabolites (norbuprenorphine, buprenorphine glucuronide, and norbuprenorphine glucuronide), along with the deuterated internal standards for BUP, norbuprenorphine, and norbuprenorphine glucuronide (buprenorphine glucuronide was not available), were extracted from plasma samples by solid-phase extraction methods.

Plasma concentrations were determined using high-performance liquid chromatography with tandem mass spectrometric detection. The peaks of interest were well separated and the overall run time for each sample was 7 minutes. Detection was accomplished utilizing ion spray tandem mass spectrometry in positive ion multiple reaction monitoring mode. The lower limit of quantification was 0.05 ng/mL for BUP, and calibration curves were linear, ranging from 0.05 to 50 ng/mL for BUP with coefficients of determination (r2) greater than 0.99.

Both the intraday and interday precisions were evaluated and the coefficient of variation values were less than 15% for low, medium, and high controls. The accuracy as measured by bias was less than 5%. Cumulative urine and serial oral fluid samples were also collected and concentrations measured (data not shown).

Statistical analysis

Noncompartmental analysis was performed and various PK parameters were calculated using Microsoft Excel (Richmond, WA). Maximum BUP plasma concentrations (Cmax), trough BUP concentrations at time zero (C0), BUP concentrations at 12 hours (C12), and time to maximum BUP concentrations (Tmax) were observed values using each participant’s plasma concentration-time profile.

In each of the participants, the area under the BUP plasma concentration-time curve for the 12 hour dosing interval (AUC _0→12 ) was calculated from time 0 to 12 hours using the trapezoidal rule. The mean values were then used to compare cohorts. Given the reportedly long half-life of buprenorphine (approximately 37 hours) and the shorter dosing interval (12 hours) used in the study participants, it was not possible to calculate the terminal half-life.

Univariate descriptive statistics were used to summarize sample demographic and clinical characteristics at each study visit. Means and standard deviations were reported for continuous variables, and frequencies and percentages were reported for categorical variables. BUP pharmacokinetic data were similarly summarized descriptively at each study visit using means and standard deviations. Statistical analyses were then performed to identify significant differences in BUP PK parameters between (1) PK-2 and PK-3 and (2) pregnancy and postpartum, without focus on a specific trimester of pregnancy.

The distributions of sample PK data were investigated, and a natural logarithmic transformation was applied to each of the PK variables to produce approximately normally distributed data for statistical analysis. Univariate linear mixed models were then fit to the log-transformed PK data to make the 2 comparisons of interest because such modeling techniques utilize data from every participant and do not require that participants have complete data at all 3 study visits.

Time (PK-2, PK-3, and PK-P) was treated as a categorical fixed effect in all models, and a random effect was included in the models to account for repeated measures over time from the same participants. For the random effect of time, a variance component covariance structure was assumed. Models were fit through maximum likelihood estimation in SAS version 9.4 (SAS Institute, Cary, NC), and convergence criteria were met without problems.

The fitted models were then used to compare PK parameters between the second and third trimesters as well as between pregnancy and postpartum, using Student t tests. Specifically, the first comparison tested whether the log-transformed PK parameter differed significantly between PK-2 and PK-3. For the second comparison, a formal hypothesis test was performed to determine whether the log-transformed mean at PK-P differed significantly from the average of the log-transformed means at PK-2 and PK-3. The 2 P values were adjusted for multiple comparisons using the Sidak adjustment method. A value of P < .05 indicated a statistically significant difference.