* Dr J. E. Norman, University Department of Obstetrics and Gynaecology, Glasgow Royal Infirmary, 10 Alexandra Parade, Glasgow, G31 2ER, UK.
Vaginally administered nitric oxide donors such as isosorbide mononitrate have been used to ripen the uterine cervix in pregnancy. The pharmacokinetics of isosorbide mononitrate following vaginal administration are unknown. Serum levels of isosorbide mononitrate were determined at baseline and 60, 180 and 360 minutes after vaginal administration of 20 or 40 mg isosorbide mononitrate to pregnant women scheduled for induction of labour at term. Serum levels of isosorbide mononitrate continued to rise up to 360 minutes after isosorbide mononitrate insertion, with mean (SD) final levels of 337 (94) μg/L following isosorbide mononitrate 40 mg and 144 (47) μg/L following isosorbide mononitrate 20 mg, P < 0.01.
We and others have demonstrated the efficacy of vaginally administered nitric oxide donors for cervical ripening in the first trimester1. In contrast to prostaglandins, nitric oxide donors do not induce uterine contractions. Nitric oxide donors may therefore have advantages over prostaglandins for cervical ripening prior to induction of labour in the third trimester. A recently published report compared the nitric oxide donor glyceryl trinitrate with prostaglandin E for pre-induction cervical ripening2; other studies including our own are currently under way.
In the majority of clinical studies using nitric oxide donors for cervical ripening, the formulation of nitric oxide donor used was designed for the purposes of oral administration. Details of the pharmacokinetics following vaginal administration of these agents are unknown. Data on the serum profile of vaginally administered nitric oxide donors will inform the optimum dose and dosage interval for pre-induction cervical ripening. This report describes serum levels of isosorbide mononitrate measured using gas chromatography–mass spectrometry (GC-MS) after administration of either isosorbide mononitrate 40 mg or isosorbide mononitrate 20 mg to the posterior vaginal fornix of pregnant women at term.
Primigravid women scheduled for induction of labour were recruited from the labour ward at Glasgow Royal Maternity between August 1998 and July 1999. Approval was granted by the local research ethics committee and all subjects gave written informed consent. The women recruited to this study were those participating in a recently published study determining the effects of isosorbide mononitrate on maternal and fetal haemodynamics3. Participants were randomised to receive either isosorbide mononitrate 40 mg (n= 11) or isosorbide mononitrate 20 mg (n= 12) inserted into the posterior vaginal fornix. Treatment options were randomised in groups of 12 (block randomisation) by one of us (JN) using random number tables. Treatment allocations were placed into opaque envelopes numbered sequentially with a unique study number and sealed. After recruitment of each patient to the study, the envelopes were opened in order by an independent observer, who then administered the treatment. Information on treatment allocation was then discarded by the independent observer, and the unique study number was used to code the patient's blood samples. The code was broken after isosorbide mononitrate concentration was measured in each of the samples.
Isosorbide mononitrate was administered in the form of either one or two tablets each of 20 mg isosorbide mononitrate (Elantan, Schwarz Pharma, Chesham, Bucks, UK). These tablets are formulated for oral administration, but we have previously found them to be effective when administered vaginally for cervical ripening in the first trimester1.
The sampling schedule was informed from data on the pharmacokinetics of isosorbide mononitrate following oral administration, where peak plasma levels are achieved within 2 hours and have a half-life of approximately 5 hours4. Accordingly, a venous blood sample was taken from the antecubital fossa at baseline, immediately prior to isosorbide mononitrate insertion, and at 60, 180 and 360 minutes after drug treatment. Blood was collected into a plain vacutainer tube and allowed to coagulate. The blood sample was centrifuged within 15 minutes and the serum was aspirated and frozen at −20°C until analysis.
Isosorbide mononitrate concentration was determined in serum using GC-MS. Samples were analysed blind to the treatment dose. Duplicate aliquots of the serum samples (1 mL) were spiked with an internal standard (200 ng isomannide mononitrate) and extracted with 4 mL ethylacetate by vortex mixing. The samples were centrifuged for 5 minutes at 3000 rpm (Labofuge 400, Heraeus Instruments, Hanau, Germany) and the organic phase transferred into a clean tube. The ethyl acetate was evaporated to dryness under a stream of nitrogen gas at room temperature. The extracts were then derivatised by addition of 50 μL bis(trimethylsilyl)trifluoroacetamide. A 1-μL portion of the derivatised extract was injected into the GC-MS. The GC-MS was operated in Negative Chemical Ionisation mode using methane as the reagent gas. The base peak ion of m/z 217 was selectively monitored for both isosorbide mononitrate and the internal standard. The chromatographic separation was carried out on a non-polar cross linked methyl silica column (HP-1, 12 m × 0.22 mm × 0.33 μm film thickness). The GC oven was heated from 60°C to 320°C at a rate of 20°C/minute with the injector and transfer line maintained at 220°C. A calibration graph was constructed for each analyses using 1 mL blank female serum samples spiked with 200 ng isomannide mononitrate and (0–300 ng) isosorbide mononitrate.
The effects of dose were analysed on SPSS (Version 8.0, www.spss.com) using a general linear repeated measures statistics model. Within each dose group, analysis of variance, with Scheffe's test as a post hoc test, was used to explore differences in isosorbide mononitrate levels at different times.
Mean (SD) age of the women was 27.4 years (5.3) in the isosorbide mononitrate 20 mg group and 26.4 years (5.8) in the isosorbide mononitrate 40 mg group. Mean (SD) gestation was 41 weeks 1 day (two days) in the isosorbide mononitrate 20 mg group and 41 weeks 2 days (one day) in the isosorbide mononitrate 40 mg group.
A representative GC-MS trace is shown in Fig. 1. The lower limit of detection and the interday coefficient of variation of the isosorbide mononitrate assay was 5 μg/L and 3.75% (range 2.2–5.18, using a 50-μg/L standard), respectively. Figure 2 shows mean (SD) serum isosorbide mononitrate level plotted against time for each dosage group. Serum levels of isosorbide mononitrate were significantly greater following isosorbide mononitrate 40 mg compared with isosorbide mononitrate 20 mg at all times except time 0 (P < 0.001). In the group given isosorbide mononitrate 20 mg, there was a significant difference between isosorbide mononitrate levels at each of the time points and the others except for time 0 versus 60 minutes (P≤ 0.02). For the isosorbide mononitrate 40 mg, there was a significant difference between isosorbide mononitrate levels at each of the time points, except for 180 minutes versus 360 minutes (P≤ 0.02).
As expected, isosorbide mononitrate, administered in tablets designed for oral administration, is absorbed from the vagina into the bloodstream. To our surprise, however, the serum profile was very different following vaginal administration compared with oral administration. Specifically, the timing and concentration of peak plasma levels were dissimilar. In previous studies where isosorbide mononitrate was given orally, peak levels occurred within 2 hours4. In our study where isosorbide mononitrate was given vaginally, we had not observed a fall in serum levels by the time the last sample was taken, 6 hours after administration. The timing of peak levels was at least 6 hours, and possibly longer after vaginal administration of isosorbide mononitrate. Although we cannot be certain that the true peak serum level was reached, observed peak concentrations following vaginal administration were less than half those following oral or intravenous administration. [Peak mean (SD) serum levels of isosorbide mononitrate were 144 (47) μg/L following vaginal administration, and 400–450 μg/L following a single oral or intravenous administration of 20 mg isosorbide mononitrate4]. It is likely that vaginally administered isosorbide mononitrate is subject to the first uterine pass effect, whereby drugs are preferentially transported from the vagina to the uterus5. This may result in a much greater isosorbide mononitrate concentration in the uterus than in the serum, although we were unable to test this formally. Because a significant proportion of the drug is delivered to the uterus, the amount available for absorption into the bloodstream will be smaller, perhaps accounting for the lower serum levels seen following vaginal, compared with oral or intravenous administration.
We observed a dose-dependent effect of vaginally administered isosorbide mononitrate on serum isosorbide mononitrate concentrations, in keeping with our previous observations that 40 mg of isosorbide mononitrate administered to pregnant women at term has a more marked depressant effect on maternal blood pressure than 20 mg3. In contrast, in a previous study in the first trimester, we failed to show any difference in the incidence of systemic side effects (e.g. headache), blood pressure or clinical ripening effect between 40 and 80 mg isosorbide mononitrate administered vaginally6. Although these studies were undertaken in different trimesters of pregnancy, taken together, these data suggest that plasma levels of isosorbide mononitrate increase in response to increasing vaginal doses of isosorbide mononitrate up to 40 mg. Increasing the vaginal dose of isosorbide mononitrate beyond 40 mg produces little, if any, increase in plasma levels. Further studies involving larger doses of isosorbide mononitrate are required to investigate this hypothesis further. The serum levels generated in the study described in this report were at least equivalent to those in a previous study, where the majority of women reported headache as a side effect7. This is in accord with women's experience of headache following 20 or 40 mg isosorbide mononitrate administered vaginally in our previous study3.
Because surrogate markers of cervical ripening (inhibition of contractile activity8 and production of interleukin-89) can be stimulated in the human cervix by in vitro incubation with a nitric oxide donor, it is likely that the cervical ripening effect of isosorbide mononitrate is mediated in part by direct transport of the drug from the vagina to the cervix. In contrast, the reported side effects of headache, tachycardia and a fall in maternal blood pressure are probably related to circulating levels of isosorbide mononitrate. In other words, the ripening effects of vaginally administered nitric oxide donors are probably due to a local action of the drug, but side effects are due to systemic absorption. If nitric oxide donors are shown to be effective cervical ripening agents in the third trimester, it is likely that more than one dose will be required. We have previously highlighted the importance of targeting the clinical use of nitric oxide donors in a site-specific manner10. The challenge will be to maximise cervical effects while minimising unwanted systemic side effects. One option is to alter the formulation of isosorbide mononitrate for vaginal administration, so that uterine delivery is increased and systemic absorption decreased. Until such a formulation is compounded, the data presented in this paper begin to provide a rational basis on which to determine the appropriate dosing interval for formulations of isosorbide mononitrate available at present.
We have shown that vaginal administration of isosorbide mononitrate is associated with lower plasma levels but a longer period from administration to peak levels, compared with oral or intravenous administration. Recently published RCOG guidelines indicate that prostglandin, administered vaginally for pre-induction cervical ripening, should be repeated after 6 hours if the cervix fails to ripen. Our data highlight the drawbacks of extrapolating from dosage schedule for vaginally administered prostaglandin to isosorbide mononitrate. If the prostaglandin regimen is employed with vaginally administered isosorbide mononitrate, systemic levels will accumulate, with a risk of clinically significant side effects. We suggest that the optimum dosing interval is likely to be significantly longer than 6 hours. The precise timing will be informed by data on the pharmacokinetics of isosorbide mononitrate beyond 6 hours after vaginal administration. These studies are currently under way in our departments.
This work was funded by a grant from Tenovus Scotland (ref. no. S99/10) for which we are extremely grateful.