The effect of the onset of labor on the characteristics of the cesarean scar

To assess the effect of cesarean section (CS) timing, elective versus unplanned, on the residual myometrial thickness (RMT) and CS scars.


| INTRODUC TI ON
Implantation of the gestational sac within the niche is a progenitor to placenta accreta spectrum (PAS). 5 and the presence of a niche and RMT thinning are associated with a high risk for developing PAS. 6 In addition to this, the RMT measurement has been assumed to play a role in predicting uterine rupture during the trial of labor after CS. 7 There is emerging evidence that the risk of PAS is higher in subsequent pregnancies after elective pre-labor CS than after unplanned CS after cervical dilatation. 8 A recent study comparing the CS scars between early-labor and pre-labor CS reported a higher incidence of niches after early-labor CS without reporting RMT. 9 Moreover, the clinical implication of reporting RMT as an absolute value is debatable whereas RMT ratio represents a reproducible measurement that could be more useful for comparing patients. 10 Our work was proposed to be the first prospective study that uses the recommendations of Jordans et al. 3 for assessing RMT after CS. We hypothesize that CS timing in relation to the onset of labor affects RMT ratio, and the aim of this study is to assess the effect of CS timing, scheduled versus unplanned in labor, on the characteristics of the CS scar 1 year later.

| MATERIAL S AND ME THODS
This is a prospective observational cohort study investigating the effect of CS timing as a binomial exposure variable, whether elective pre-labor or unplanned intrapartum, on the primary outcome of RMT ratio as a continuous variable. The Ethical Committee of the . The RMT ratio, the presence of a niche and/or fibrosis, as well as the distance from the scar to the internal os (SO) as shown in Figure 1, were recorded. A niche is defined as "an indentation at the site of CS scar with a depth of at least 2 mm" 3 and fibrosis is defined as "a hyperechogenic dent from the serosa into the myometrium". 11 Our null hypothesis assumes that the mean RMT ratios (primary outcome) for the two cohorts lie within half a standard deviation, which was adapted from Roberge et al. 12 The power analysis was performed with 0.05 type I error, 0.1 type II error and 23% standard deviation. The calculated two-sided sample size was 186 individuals with 93 observations for each cohort. All statistical analyses were conducted with Stata ® (ver. 16.1; StataCorp., College Station, TX, USA) and the Wilcoxon rank-sum test, linear regression, and χ 2 test were utilized.

| RE SULTS
Of the recruited eligible patients, 300 non-pregnant women attended a follow-up ultrasound examination within 12 and 24 months postoperatively. We used block randomization to select 93 patients who underwent elective CS and 93 patients who had unplanned intrapartum CS in the analysis. A summary of the characteristics for the study population along with some perioperative circumstances associated with the CS are shown in Table 1.
The collected scar measurements from this study are summarized in Table 2.
Two-sample Wilcoxon rank-sum test for RMT ratio dependence on CS timing resulted in Z value of −0.59 and a P value of 0.553, whereas CS timing showed a significant effect on SO distance (Z −4.94; P < 0.001). The distribution of RMT ratio as well as SO distance among the two cohorts is shown in Figure 2.
The χ 2 test showed no association between CS timing and niches or CS timing and fibrosis (P > 0.99 and P = 0.268, respectively). The relative frequencies of manifesting niches and fibrosis are shown in

| DISCUSS ION
The main outcome of this study showed that RMT ratio was independent of CS timing. RMT is believed to be key for the risk assessment of complications facing women post CS during a subsequent pregnancy, such as PAS and uterine rupture during trial of labor. 13 Several study groups show that the anterior RMT is measurable and reproducible and RMT is measured on the sagittal plane. 3 The  15 Our results failed to demonstrate a difference of RMT ratios between scheduled and unplanned CS.
Some of these contradictory results can be attributed to early scar assessment within as short a time-lapse as 3 months postoperatively. 9 Our cohorts were examined after at least a year and are assumed to have reached their scar healing potential, achieved after 6-9 months post CS. 16 The RMT and the extent of niches seem to be dependent on the surgical technique with better healing and thicker RMT after doublelayer uterine closure. 17 The standard surgical procedure used in our study center is double-layered unlocked, which is associated with a healing ratio of 73% ± 23% according to Roberge et al. 12 As a result, we used 23% as a standard deviation in our power analysis because of the similarity of both closure technique and reporting scar healing as a ratio. 12 The published incidence of niches varied within a broad range from 7% to 65%. 18,19 The variance in niche incidence could be partly attributed to the population demographic difference. Nevertheless, our population exhibits an exceptionally high incidence of both niches and fibrosis, and although we are unable to identify a definite clarification for this high incidence, employing high-frequency matrix transducers by experienced sonographers as well as using the latest niche assessment guidelines could have improved niche imaging and facilitated a higher detection rate of niches. 3 There are several other factors that affect RMT and CS scar healing, including obesity, infection, smoking, diabetes, and multiple CS. 20 We controlled for some factors at the level of study design by excluding patients with repeated CS or diabetes; however, not including demographic information on smoking or obesity can be considered a weakness of our study. The significant difference of peripartum infection rate between the two cohorts can be attributed to the nature of unplanned CS due to obstructed labor or fetal distress. Nevertheless, infection could have skewed the results of the RMT ratio and scar healing in our cohorts towards the null.
As far as we are aware, there has been only one recent published study that investigated the relationship between scar position and intrapartum CS timing. The investigators showed that pre-labor and early-labor CS are associated with scars within the uterine cavity and the height of the scar is inversely correlated with the cervical dilatation. 9 This negative linear relationship is in agreement with the theory of normal cervical dilatation pattern during labor, whereby the cervix is pulled superiorly after effacement. 21 Our data confirm these findings because the scar location was significantly higher within the uterus for elective CS compared with unplanned CS. The inverse linear relationship between SO distance and cervical dilatation at CS in our study reflects the work published by Kamel et al. 9 with a very similar intercept around 10.2 mm, but their slope was more extreme with a decrease of 1.39 mm instead of our 0.45 mm for each centimeter increase in cervical dilatation. The difference of the absolute values can be attributed to the population differences, and the retained similarity of the negative association confirms the validity of this finding.
In conclusion, we demonstrate that the relationship between performing the CS and the onset of labor did not affect the RMT for our study population. The height of the CS scar along the uterine length, referred to as the SO distance, is dependent on the progress of labor at the time of CS, and it shows a negative linear relationship with the measurement of cervical dilatation. Kamel et al. 9 assumed that the location of the scar could explain the predisposition to PAS and that low CS scars closer to the internal os could be considered protective. If we apply this thinking to our results then we could hypothesize that unplanned, in labor CS is associated with lower incidence of PAS, noting that this outcome was not a part of our study.
The strengths of this study are the prospective design, the statistical power analysis, the modern sonographic equipment, randomization, and blinding to exposure during outcome assessment. The association between CS timing and the location of the CS scar is a theoretical facilitator of PAS in a subsequent pregnancy, but proving the causative relationship is difficult because of the relatively low incidence of PAS.
This rarity hinders the clinical implication of cohort studies such as ours for prospective investigation of the causative relationship.

ACK N OWLED G M ENTS
This study was supported by the Dr. Senckenberg Foundation, Frankfurt am Main.

CO N FLI C T S O F I NTE R E S T
The authors have no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
AAN contributed to protocol, data collection, and manuscript writing; LJ contributed to data analysis and manuscript editing; NM con-