Longitudinal Ultrasound Evaluation of Dilapan-S During Cervical Ripening

Objective — To assess the diameter change of hygroscopic rod dilation during 12 hours of cervical ripening. Methods — This was an observational, prospective study of term women undergoing labor induction with a bishop score ≤ 6. Women were allocated into two groups (soaked gauze or no gauze) strati ﬁ ed by parity. Using transvaginal ultrasound, maximal rod diameters were obtained in a longitudinal plane. Measurements were taken at four pre-speci ﬁ ed time points (3, 6, 8, and 12 hours). All rods were removed at 12 hours from insertion. Patient satisfaction scores between the groups were assessed. To evaluate if measures were signi ﬁ cantly different among the four time points, a generalized linear model was used. Independent t -tests were used to compare mean rod diameter values and pain measures between the two groups. Fisher Exact tests were used to evaluate categorical satisfaction measures. Results — Forty-four women were recruited with a total of 178 hygroscopic rods placed. Mean rod diameters (mm) were signi ﬁ cantly different among the four time periods (3 hour: 7.9 mm [SD 0.9]; 6 hour: 9.4 mm [SD 0.9]; 8 hour: 10.0 mm [SD 0.9]; 12 hour: 10.9 mm [SD 0.8]; P -value <.001). After stratifying by the use of gauze, there was no difference in rod diameters at 3, 6, 8, and 12 hours respectively. There was no difference in patient satisfaction scores between the two groups. Conclusion — The majority of hygroscopic rod dilation occurs within the ﬁ rst 8 hours of cervical ripening. Placement of saturated gauze does not accelerate rod dilation.


Introduction
I n the United States, 27% of women undergo an induction of labor. 1 Methods of induction vary, with the majority of women requiring cervical ripening. 2 Numerous options are available for cervical ripening including the use of mechanical methods and pharmacologic agents. Each choice has unique advantages and disadvantages that render neither method superior to the other. 3 For decades, the Foley balloon has been the favored mechanical intervention in the United States. 4 While evidence is robust comparing the use of the Foley balloon to other pharmacologic options, [5][6][7] it is limited when comparing one mechanical method to another. A non-inferiority study, the DILAFOL trial, compared Foley balloon placement with a hygroscopic dilator, Dilapan-S. The authors concluded that Dilapan-S was not inferior to the Foley balloon in achieving vaginal delivery in term gravidas with better patient satisfaction in the Dilapan-S group. 4 Dilapan-S is a second-generation synthetic hygroscopic dilator made from Aquacryl, a proprietary hydrogel. 4 Mechanistically, fluid is absorbed from cells in the cervix leading to radial expansion of the dilators resulting in cervical softening and dilation. 4 Dilapan-S has been successfully used for cervical priming in women undergoing second-trimester terminations 8 and received approval from the Food and Drug Administration (FDA) in 2015 for cervical ripening in the third trimester. 9 According to the manufacturer, Dilapan-S rods can remain within the cervix for up to 24 hours with successful cervical ripening occurring within 12-15 hours. 9 While successful cervical ripening occurs within 12 hours, information related to the actual change of Dilapan-S rods during induction of labor is unknown. To fill this gap in knowledge, we sought to measure individual Dilapan-S rod diameters using transvaginal ultrasound at four prespecified time points during 12 hours of cervical ripening. Additionally, manufacturer guidelines comment on the placement of gauze with Dilapan-S. Based on the mechanism of action of Dilapan-S, we further assessed the impact of gauze placement with Dilapan-S versus Dilapan-S alone on measured rod dilation. We hypothesized that Dilapan-S rod diameters would display minimal change after 8 hours and that gauze placement would accelerate rod dilation. Participating subjects were asked to complete satisfaction surveys.

Materials and Methods
This was a single-center, observational prospective study at our affiliated academic institution, Sentara Norfolk General Hospital. Screening for potential study participants was completed by the principal investigator (ES) by reviewing the list of planned scheduled inductions from October 2020 through June 2021. Subjects who met the following criteria were considered eligible to participate in the study: age 18-45 years, scheduled medical or elective induction, gestational age of 37 weeks or more at initiation of induction, Bishop score of 6 or lesson examination performed immediately prior to cervical ripening, singleton fetus in cephalic presentation, and intact membranes. Participants were excluded if they had a prior cesarean delivery, active genital tract infection, active vaginal bleeding, intrauterine fetal demise, known fetal malformations, category 2 fetal tracing on admission, and BMI of 45 or greater. The study was approved by Eastern Virginia Medical School (EVMS)'s Institutional Review Board (IRB#20-04-FB-0097). Medicem (Kamenne Zehrovice, Czech Republic) provided the Dilapan-S rods for study participants, but had no input on study design, execution, data analysis, and interpretation.
Due to a lack of published studies on this topic, a sample size of convenience was utilized. Recruitment for the study was performed at our outpatient office. Written informed consent was obtained from all participating subjects. Seventy-two hours before the scheduled induction, a baseline Bishop score, transvaginal ultrasound to assess cervical length, and transabdominal ultrasound to obtain an estimated fetal weight was obtained by the primary author. Cervical length was assessed according to CLEAR guidelines 10 and estimated fetal weight was based on the Hadlock formula. 11 Upon admission for the scheduled induction, the primary author repeated the Bishop score. On admission, stratified block randomization with block size of four was used to allocate participants on a 1:1 ratio to the two study groups (Dilapan-S alone and Dilapan-S plus gauze). Randomization was stratified by parity. The randomization scheme was generated by EVMS Healthcare Analytics and Delivery Science Institute (HADSI) prior to the start of enrollment using SAS version 9.4 (SAS Institute, Cary, NC). Allocation of the group was not blinded to the study subjects or the healthcare team.
Study participants assigned to each group underwent continuous cardiotocography for 20 minutes prior to placement of the Dilapan-S rods (rod size: 4 Â 65 mm). Placement of the Dilapan-S rods was limited to three physicians with the majority placed by the primary investigator. Placement was completed in a sterile manner under direct visualization and in accordance with the manufacturer's guidelines. 9 The number of rods placed varied per individual and was based on the discretion of the physician. Each participant allocated to the Dilapan-S with saturated gauze group received placement of one, 4 Â 4 inch gauze saturated with 10 mL of normal saline and placed at the level of the external os. Immediately after placement of the Dilapan-S rods, the study participant was asked to complete a patient satisfaction survey 4 (Supplemental Figure 1). Participants were instructed to report any vaginal bleeding, severe pain, or concern for device expulsion. Participants were kept in the hospital during the cervical ripening phase. Fetal monitoring while Dilapan-S was in place followed hospital-based protocols. Analgesia was available upon request. The Dilapan-S rods remained in place in the event of rupture of membranes. The rods were removed if it was deemed the patient was in active labor or at 12 hours from insertion time.
Pre-specified transvaginal ultrasound assessment was performed at hours 3, 6, 8, and 12 post-Dilapan-S placement using a GE Voluson E10 machine (Zipf, Austria) with a transvaginal GE RIC 5-9D transducer. An in vitro experimental study of natural versus synthetic dilators found that Dilapan-S rods display greatest diameter increase within the first 8 hours, with almost maximum change occurring and plateauing after hour 12. 12 This created the foundation for using a 12 hour time duration of Dilapan-S rod placement during this study. The ultrasound methods followed the "As Low As Reasonably Achievable" ALARA principle set forth by the American Institute of Ultrasound in Medicine (AIUM). 13 Preparation and cleaning of the transvaginal probes and safe handling were observed in accordance with department guidelines. Performance of the ultrasound evaluation was limited to the primary investigator and two co-authors of the study.
The Dilapan-S rods were identified on transvaginal ultrasound as anechoic structures in comparison to the echogenic surrounding cervix. On ultrasound, confirmation of correct placement could be visualized with the rods extending beyond the internal os. With the first ultrasound 3 hours post Dilapan-S placement, the spatial orientation of the rods within the cervix was determined. Starting at the midsagittal plane of the cervix and sweeping from left to right, each individual rod was identified in a longitudinal plane and labelled. Once the individual rod was in the longitudinal plane, the length and diameter of the rod were measured from anechoic to anechoic border and recorded (Supplemental Figure 2). Several measurements were obtained along the length of the rod, with the largest measurement recorded. If the rod was difficult to obtain, 3D imaging was used to measure the rod. With each subsequent ultrasound, the same measurements occurred for each individual rod. Repeat cervical length measurements were not obtained outside of the baseline work-up. In women allocated to gauze placement, the original gauze remained in place during each ultrasound exam. After completion of the 12-hour ultrasound, the Dilapan-S rods and gauze, if used, were removed per the manufacturer's instructions. The total number of Dilapan-S rods removed was confirmed with additional medical personnel. A final sterile vaginal exam was performed to determine the post Dilapan-S Bishop score. The participant was then asked to complete a post Dilapan-S patient satisfaction survey 4 (Supplemental Figure 1). Labor management after removal of the Dilapan-S rods was at the discretion of the supervising health care team.
Our primary outcome was the diameter change of Dilapan-S rods (4 Â 65 mm) at four predetermined times (hours 3, 6, 8, and 12). Secondary outcomes were to assess whether diameter change of Dilapan-S rods varied in those with saturated gauze and to assess maternal satisfaction scores between the groups based on a designated questionnaire.
Descriptive statistics were used to analyze demographics and clinical characteristics of participants. To compare categorical demographic and clinical variables between Dilapan-S and Dilapan-S plus gauze groups, Fisher Exact tests or Chi-square tests were used depending on categorical frequencies. To compare continuous demographic variables between Dilapan-S and Dilapan-S plus gauze groups, Independent t-tests or Mann-Whitney U tests were conducted depending on the distribution of the data.
To evaluate if Dilapan-S diameter measures were significantly different among the four time points for Dilapan-S placement, a generalized linear model was used. The overall generalized linear mode was followed with a pairwise post hoc test to assess specific differences among the four time periods. Independent t-tests were used to compare mean Dilapan-S diameter values between the Dilapan-S only and Dilapan-S plus gauze groups at hours 3, 6, 8, and 12. Fisher Exact tests were utilized to compare categorical satisfaction measures among the two therapy groups. Independent t-tests were used to compare pain measures among the two therapy groups. P values of <.05 were considered significant. Data was analyzed using SAS 9.4.

Results
From October 2020 to June 2021, 44 study subjects consented for the study with a total of 178 Dilapan-S rods placed. Most participants were Caucasian (55%), unemployed (50%), and nulliparous (66%). The mean age was 26.5 years (SD 5.67) and the mean baseline cervical length in centimeters was 2.99 (SD 0.84). On average, participants had 4 Dilapan-S rods with the minimum and maximum of 3 and 5, respectively. Demographic characteristics (Table 1) and indications for induction (Supplement Table 1) were not significantly different between the Dilapan-S only and Dilapan-S plus gauze groups, P > .05.
Most participants had a vaginal delivery (75%). The mean time to active labor (defined as time from Dilapan-S removal to 6 cm) was 9.82 hours (SD 6.40), time to delivery (defined as time from initial Dilapan-S placement to delivery) 28.45 hours (SD 8.56), birth weight 3281.82 grams (SD 512.92), and the medians for Apgar scores at minutes 1 and 5 were 8 and 9, respectively (Table 2).
Overall, there was a significant difference among the four time points 3, 6, 8, and 12 hours (F = 341.77, P < .05, Figure 1). All time-point pairwise comparisons were significant, P < .05. Although the average diameters at 3, 6, 8, and 12 hours were larger for the Dilapan-S plus gauze group compared with the average diameters for Dilapan-S only group, the differences were not significant, P > .5 ( Figure 2, Table 3). In Table 3, measurement discrepancies may be noted at different time periods due to lack of complete visualization of the rod at the time of ultrasound. Furthermore, a discrepancy is noted at hour 8 versus hour 12 in the gauze group as one patient went into active labor after 8 hours of placement and had 4 total Dilapan-S in place.
The overall experience of cervical ripening was not significantly different for the Dilapan-S only and the Dilapan-S plus gauze groups, P > .05. Pain measures and experiences related to anxiety, discomfort, daily activities, relaxation, and sleep, were not significantly different between the two groups, P > .05.

Discussion
To our knowledge, this is the first study that has objectively measured the in vivo diameter change of the hygroscopic dilator, Dilapan-S (4 Â 65 mm) during cervical ripening. At each time period there was a statistically significant change in the diameter, confirming the hygroscopic rods continue to dilate over a 12 hour period. Notably, the diameter change is less the longer the rods remain in the cervix. Interestingly, placement of saturated gauze with Dilapan-S did not affect rod diameter. When evaluating the rate of change between the four specified time periods, the greatest increase in rod dilation occurred from placement to the 3-hour time mark. Lastly, when using hour 12 as the determined maximum dilation point, the majority of dilation occurred within the first 8 hours.
Today, various methods of cervical ripening are available. The preferred method is often based on physician practice, hospital protocol, cost, and baseline Bishop score. Mechanical methods are a popular option due to their relatively low cost, minimal storage requirements, and lack of association with fetal heart rate changes. 14 Dilapan-S is one mechanical option that has gained attention owing to its unique properties and proven maternal and fetal safety profile. 15,16 According to the manufacturer's guidelines, 80% of Dilapan-S dilation occurs within the first 4-6 hours, with one rod reaching a maximum diameter of 14.6 mm at 24 hours. 9 Until now, there has been lack of research to confirm this statement. Furthermore, there is lack of published studies confirming how long Dilapan-S should remain within the cervix to reach it's maximal dilation or effect. An international multicenter trial evaluated placement of Dilapan-S at <12 hours and >12 hours and found comparable results regarding change in Bishop score, with improved rates of vaginal delivery when Dilapan-S was removed at hour 12. 17 Among participants undergoing second trimester terminations, Dilapan-S has been proven to be as successful for same day use compared with misoprostol, with one rod documented to reach 12.5 mm of dilation after 4 hours. 18 These articles, along with in vitro data, support a shorter duration of Dilapan-S rod placement to achieve equal and improved clinical outcomes. 12,17,18 Our study confirms in vitro data and supports the notion that Dilapan-S rods achieve the majority .969 Apgar scores 5 (minutes) 9 (2) 9 (1) 9 (2) .610 Note: Data are mean (SD), n (%), or median (range). *P values were obtained from independent t-tests or Mann-Whitney U tests for continuous variables or from Chi-square/Fisher Exact tests for categorical variables. of their dilation within 8 hours of placement during cervical ripening. While statistically significant change was noted from hour 8 to 12, the speed of growth of the hygroscopic rods declines as the hours progress. While we observed a clinically insignificant rate of growth after 8 hours of placement, it is possible that slight change in the dilation, as well as, the duration of Dilapan-S placement could be important. Future studies that compare outcomes between 8 and 12 hours of Dilapan-S placement would be useful.
Our findings support the aforementioned clinical studies showing that a shorter time duration for Dilapan-S placement may translate into an improved change in measured clinical outcomes. The use of gauze did not accelerate the diameter change of Dilapan-S rods. The lack of statistical  significance may be explained by the fact that the study was not powered to detect this difference and limited by the small sample size. Additionally, only one saturated gauze was placed to ensure patient comfort. Our findings may be different with the use of additional soaked gauze. However, this may be at the detriment of patient satisfaction with an unknown clinical benefit. While gauze placement can help to ensure the security of rod placement, we do not believe its use is necessary.
Overall, participant satisfaction scores on the preand post-survey were similar between the study participants. These findings show that the use of gauze does not affect the overall patient experience when using Dilapan-S as a cervical ripening agent. Furthermore, study participants in both groups were willing to use Dilapan-S rods in future pregnancies supporting previously observed outcomes regarding patient satisfaction with Dilapan-S use. 4 Our study is not without several limitations. First, it was a single-center observational study with a relatively small sample size that was not powered to detect differences among the groups. Second, the majority of our participants were nulliparous and response to the hygroscopic rods may vary depending on a variety of clinical factors, particularly parity. Third, this was a convenience sample which can limit the applicability to other obstetric populations if not similar to our study demographic. Finally, although we concluded that near maximal dilation was reached by 8 hours, we cannot conclude that this is equivalent to a 92% clinical effect as a plateau of Dilapan-S dilation was not observed over the 12 hour time period.
A strength of this study is that it is the first to objectively measure hygroscopic rods during cervical ripening. Previous studies evaluating Dilapan-S rods have assessed clinical outcomes as a primary outcome with promising results. [15][16][17] This information can guide clinicians to optimal time placement for Dilapan-S which can result in shorter induction times and decreased hospital stay. Potentially, this can translate into decreased cost and increased overall patient satisfaction.

Conclusion
In conclusion, our study demonstrates that Dilapan-S rods dilate over a 12 hour time period with the largest diameter change in the first 3 hours and the majority of rod dilation occurring within 8 hours of insertion. This observational study sets the stage for future randomized controlled trials to assess induction success with shorter duration of Dilapan-S placement for cervical ripening.