Clinical and anamnestic risk factors for the formation of an incompetent scar after one cesarean section

Introduction

The issue of vaginal delivery after a previous cesarean section (CS) is becoming increasingly relevant in the modern world due to the lack of a significant reduction in the share of abdominal deliveries [1–6]. This issue became significant in connection with the so-called “domino principle,” when one previous CS necessarily leads to subsequent operative delivery [7]. The rate of abdominal deliveries increased significantly due to repeat surgical intervention for a single uterine scar, as previously no other delivery option was even considered for this group of patients [3][4]. Vaginal delivery in patients with a scar on the uterus is associated with a high risk of complications such as uterine rupture along the scar, massive bleeding, and organ-removing surgery [5][6][8–13]. For this reason, selecting the ideal candidate for natural childbirth in patients with a single uterine scar remains quite difficult. According to various authors, important factors for a successful attempt at vaginal delivery include a history of natural childbirth, normal body weight of the pregnant woman, and a newborn weight of less than 3500 g [2][14][15]. Studying the anamnestic data of the current pregnancy is relevant when selecting patients for vaginal delivery. However, the indications and contraindications for vaginal delivery vary as this issue is studied [16–20]. One of the complications that can prevent a smooth delivery is when the scar tissue gets thinner during pregnancy and childbirth, causing symptoms of uterine scar failure, which often leads to a bad outcome for the baby [21][22].

The aim of the study was to identify clinical and anamnestic risk factors leading to the formation of an incompetent uterine scar in patients who had undergone a previous CS.

Materials and Methods

A retrospective analysis was performed at a level 3 hospital in Rostov-on-Don and included 102 patient records with a uterine scar after a single CS (form N 096/1u-20) and 102 newborn records (form 097u). The results of pathomorphological studies of placentas (form No. 014-1/u) for 2019–2020 were analyzed.

All patients were divided into three groups. Group I (control group) included 35 (34.3%) women with uterine scars who gave birth vaginally. Group II (main group A) included 35 patients (34.3%) who underwent surgery for pain in the area of the uterine scar before delivery, with intraoperative confirmation of scar failure after one CS. Group III (main group B) included 32 patients (31.4%) with an unsuccessful attempt at vaginal delivery and intraoperatively established scar failure on the uterus after one previous CS operation.

Criteria for inclusion in the study were singleton pregnancy, uterine scar after one previous CS, and scar thickness according to ultrasound data greater than 2.0 mm.

The criteria for exclusion from the groups were antenatal and intrapartum fetal death, preinduction, and induction of labor.

The calculations were performed using MS Excel 2019 spreadsheets (Microsoft, USA) and the Statistica 10 software package (StatSoft Inc., USA). Most of the numerical indicators did not have a normal distribution, so the nonparametric Kruskal-Wallis test was used to compare the three study groups. If there were no significant differences between the three groups according to this criterion (p>0.05), no pairwise two-sided comparison of the mean ranks between the groups was performed. Otherwise, the results of pairwise comparisons of groups p_1,2, p_1,3, and p_2,3 were also performed, as well as the significance of differences in pairwise comparisons of groups using the two-tailed comparison of mean ranks, taking into account the effect of multiple comparisons. Related groups were compared in pairs using Wilcoxon’s test, with the threshold significance of differences, taking into account the Bonferroni correction for multiple comparisons, set at p<0.017. Quantitative data are presented as median, first, and third quartiles Me (Q1; Q3). The distribution of qualitative data was performed using Pearson’s χ² test. The significance threshold was p<0.05. In the pairwise post hoc comparative analysis, Pearson’s χ² test with Yates’ correction was used.

Results

The patients’ age ranged from 30 to 33 years old.

Age differences in the compared groups in the multiple comparative analysis were statistically significant (p=0.005, according to the Kruskal-Wallis test), while statistically significant differences in a pairwise post hoc comparative analysis were observed only between Groups I and III (p_1,3=0.008) and between Groups II and III (p_2,3=0.025). Thus, patients in Group III were significantly older.

The assessment of BMI in the study groups did not reveal any significant differences (p=0.67).

The examination of the parity of patients showed that an equal number of women (94.2%, n=33) in Groups I and II were expecting their second child, one patient in Group I was expecting her fifth child, 71.8% (n=23) in Group III were expecting their second child, four were expecting their fourth child, and five were expecting their third child. The differences in parity between the compared groups are statistically significant (p=0.015, according to Pearson’s χ² test).

An analysis of the anamnestic data revealed (Table 1) that abortions were more common in Groups II and III (28.5% n=10 and 25% n=8) than in Group I of vaginal deliveries (11.4% (n=4)). A pairwise post hoc comparison of the study groups revealed no significant differences (p_1,2=0.136; p_1,3=0.31; p_2,3=0.84, according to Pearson’s χ² test with Yates’ correction). Intrauterine interventions (hysteroscopy, separate diagnostic curettage of the uterine cavity) accounted for 71.4% (n=25) in Group II, 37.5% (n=12) in Group III, and 37.1% (n=13) in Group I. Statistical significance was established for pairwise differences between Groups I and II (p_1,2=0.009, according to Pearson’s χ² test with Yates’ correction) (OR_1,2=0.236; 95% CI = 0.087–0.645, differences are statistically significant) and between Groups II and III (p_1,3 = 0.011, according to Pearson’s χ² test with Yates’ correction). Inflammatory diseases of the pelvic organs, such as acute metroendometritis or salpingo-oophoritis (SOP), were more common in Group II patients at 48.6% (n=17), compared to Group I, where SOP was observed in 20% (n=7), and Group III, where SOP was observed in 15.6% (n=5). In a pairwise post hoc comparison of the groups, the statistical significance between Groups I and II is noteworthy (p_1,2=0.011, according to Pearson’s χ² test with Yates’ correction) (OR_1,2=3.778; 95% CI=1.308–10.913). This circumstance indicates that the inflammatory process determines the initial scar failure.

In Group III, with unsuccessful attempts at vaginal delivery, genital endometriosis was found in 62.5% (n=20) of patients, while in Group I, this factor was found in 5.7% (n=2), and in Group II, in 11.4% (n=4) of patients. A pairwise post hoc comparisons of Groups I and III, II and III were statistically significant (p_1,3<0.001; p_2,3<0.001, according to Pearson’s χ² test with Yates’ correction) (OR_1,3=12.9, 95% CI=3.6–45.7; OR_2,3=27.5; 95% CI=5.6–135.8, differences in odds are statistically significant in both cases). The obtained data indicate the role of genital endometriosis in the formation of an incompetent scar in the group with unsuccessful attempts at vaginal delivery.

Complications after the first CS (lochiometra, acute postpartum endometritis) were more frequently observed in Groups II and III with uterine scar insufficiency (7 cases (20%) in Group II and 3 cases (9.4%) in Group III); and in Group I, in one patient – 2.8%, (p=0.066, according to Pearson’s χ² test).

We studied the indications for the first CS (Table 2) and found that, in Group II, emergency operative delivery was performed in 80% (n=28) of cases, in Group III – in 84% (n=27), and in Group I, this indicator was 40% (n=14). A comparative paired statistical analysis showed that the differences between Groups I and II and between groups I and III were highly significant (p_1,2=0.004; p1.3=0.001, according to Pearson’s χ² test with Yates’ correction) (OR_1,2=6; 95% CI = 2.060–17.479, differences in odds are statistically significant) (OR_1,3 = 8.1; 95% CI = 2.5–26, differences are statistically significant). The main indication for emergency surgery was a disproportion of mixed maternal and fetal origin. This indication for operative delivery was established in half of the patients in Groups II and III (Group II – 48.6% (n=17), Group III – 50% (n=16)), and in Group I – 8.5% (n=3). In pairwise comparisons, Group I differed significantly from both Group II (p_₁,₂ = 0.0001, Pearson’s χ² test with Yates’ correction) and Group III (p₁,₃ = 0.0001, Pearson’s χ² test with Yates’ correction). The OR were 10.0 times higher in Group I compared with Group II (95% CI: 2.6–39.1) and 10.7 times higher compared with Group III (95% CI: 2.7–42.0), with both associations reaching high statistical significance (Table 2).

Scheduled operations were most common in Group I (60%, n = 21), compared with 20% (n = 7) in Group II and 15.6% (n = 5) in Group III, with the overall difference reaching statistical significance (p < 0.001, Pearson’s χ² test). Pairwise post hoc analysis confirmed statistically significant differences between Groups I and II (p₁,₂ = 0.004) and between Groups I and III (p₁,₃ = 0.001) based on Pearson’s χ² test with Yates’ correction. One of the main indications for planned CS was breech presentation, observed in 48.6% (n = 17) of patients in Group I, compared with 14.3% (n = 5) in Group II and 12.5% (n = 4) in Group III. These differences were statistically significant between Groups I and II (p₁,₂ = 0.005) and between Groups I and III (p₁,₃ = 0.004), according to Pearson’s χ² test with Yates’ correction (Table 2).

Analysis of the intergenetic interval across all three groups showed that a duration of less than 2 years and 9 months occurred only among patients with uterine scar insufficiency: 28.5% (n = 10) in Group II and 12.5% (n = 4) in Group III. Pairwise comparison revealed a statistically significant difference between Groups I and II (p₁,₂ = 0.0063).

Analysis of the exacerbation of extragenital diseases during the current pregnancy showed that urinary system disorders (cystitis, pyelonephritis, urolithiasis) were more frequent in groups with uterine scar insufficiency: in Group II – in 48.5% (n=17), in Group III – in 43.8% (n=14), in Group I – only in 14.3% (n=5); the differences are statistically significant, p=0.006, according to Pearson’s χ² test. Statistically significant differences were recorded between Groups I and II, and Groups I and III (p_1,2=0.005; p_1,3=0.017, according to Pearson’s χ² test with Yates’s correction) (OR_1,2= 5.667; 95% CI = 1.784–18, differences are statistically significant) (OR_1,3 = 4.7; 95% CI = 1.439–15.134, differences are statistically significant).

In all groups, most deliveries occurred after 37 weeks of gestation. Premature births were observed in Group II, including one extremely preterm delivery (2.8%) and five late preterm deliveries (14.3%). In Group III, two preterm births (6.25%) occurred at 34 and 36 weeks of gestation. Pairwise comparisons revealed highly significant differences between Groups I and II (p₁,₂ = 0.0013) and between Groups II and III (p₂,₃ = 0.0002).

Neonatal outcomes, assessed using the Apgar score, were most favorable in Group I, where newborns achieved the highest scores at both one and five minutes after birth.

In the complete blood count, leukocytosis exceeding 15 × 10⁹/L was observed in 32.3% of patients in Group III (n = 10), with the difference reaching statistical significance (p < 0.001, Kruskal–Wallis test with multiple comparisons).

A morphological examination of the placenta revealed chorioamnionitis in 78.1% of patients in Group III (n = 25) (p_1,2=0.001, p_1,3=0.001, p_2,3=0.003, according to Pearson’s χ² test with Yates’s correction) (OR_1,2=5.813; 95% CI=1.686–20.045, differences are statistically significant) (OR_1,3=27.679; 95% CI=7.271–105.358, differences are statistically significant). Further morphological examination revealed deciduitis predominantly in groups with uterine scar insufficiency: 37.1% in Group II (n = 13) and 78.1% in Group III (n = 25) (p_1,2=0.008, p_1,3=0.0001, p_2,3=0.002, according to Pearson’s χ² test with Yates’s correction) (Table 3) (OR_1,2=20.091; 95% CI = 2.452–164.641) (OR_1,3 = 121.429; 95% CI = 14.032–1050.842).

Compensatory-adaptive reactions in the placenta were least pronounced in Group III, which comprised patients with unsuccessful attempts at vaginal delivery, and were observed in 37.5% (n = 12) of cases, compared with 74.2% (n = 26) in Group II and 62.9% (n = 22) in Group I. A statistically significant difference was identified between Groups II and III (p₂,₃ = 0.006, Pearson’s χ² test with Yates’ correction) (Table 3). In Group III, the lower frequency of compensatory-adaptive reactions may explain the lower Apgar scores in neonates born to mothers in this group. Notably, in 9.6% of cases (n = 3), operative delivery in Group III was indicated by intrapartum fetal distress, and intraoperative findings confirmed uterine scar dehiscence in all of these patients.

Villous maturation disorders were most frequently observed in Group III (96.8%, n = 31), whereas the rates in Groups I and II were similar – 51.4% (n = 19) and 52.4% (n = 18), respectively. Post hoc pairwise comparisons demonstrated highly significant differences between Groups I and III (p₁,₃ = 0.0005) and between Groups II and III (p₂,₃ = 0.0002), according to Pearson’s χ² test with Yates’ correction (Table 3) (OR_1,3=26.105; 95% CI=3.198–213, differences are statistically significant).

Placental vascular changes were observed in 96.8% of patients in Group III (n = 31), 71.2% in Group II (n = 26), and 48.5% in Group I (n = 17). Detailed comparative analysis showed highly significant differences between Groups I and III (p₁,₃ = 0.01) and between Groups II and III (p₂,₃ = 0.025), according to Pearson’s χ² test with Yates’ correction (Table 3) (OR_1,2=3.059; 95% CI=1.117–8.373, differences are statistically significant) (OR_1,3=32.824, 95% CI 4.024–267.718 differences are statistically significant).

Involutive–destructive changes in the placenta were more common in the control group (74%, n = 26) and in Group III (96%, n = 30) compared with Group II (54%, n = 20). Post hoc pairwise comparisons showed a highly significant difference only between Groups II and III (p₂,₃ = 0.002), according to Pearson’s χ² test with Yates’ correction, indicating more pronounced placental alterations in Group III (Table 3).

Discussion

The present study found no association between the development of uterine scar dehiscence and either patient age or body mass index (BMI).

Intrauterine interventions, which were more frequently observed in Group II, may have directly impaired complete myometrial repair. These procedures likely resulted in inflammatory changes in the endometrium and/or myometrium, subsequently leading to placental alterations, as confirmed by histopathological examination. According to Shchukina et al. (2018) [22], intrauterine interventions are also associated with impaired healing processes. Inflammatory diseases of the pelvic organs were notably more frequent in Group II, suggesting a potential link with disrupted myometrial repair. Similar associations between inflammatory conditions and impaired reparative processes were reported in several other studies [23][24].

According to our data, a history of genital endometriosis may be a significant factor contributing to uterine scar incompetence. Similar findings were reported by Okulova [23] and by Tsikhai et al. (2016), who documented the presence of endometrial heterotopia within the tissue of incompetent uterine scars [24].

The most common indication for the first CS in cases that subsequently developed uterine scar incompetence was cephalopelvic disproportion. A key factor potentially impairing the healing process may be incorrect identification of the border between the lower uterine segment and the cervix, when the incision is placed too low and cervical glands producing mucus are included in the suture line. The secretion from these glands may interfere with tissue repair. This assumption is supported by reports from several authors indicating that cesarean delivery performed during the active phase of labor – particularly when the incision is placed 2 cm below the uterovesical fold – as well as difficult fetal head extraction (often encountered in cephalopelvic disproportion), may increase the risk of impaired myometrial healing [25][26].

Another important factor in the development of uterine scar incompetence may be a short intergenetic interval (< 2 years 9 months), as this duration is considered by the WHO to be necessary for complete tissue recovery.

In our study, the occurrence of intrapartum fetal distress in Group III was likely associated not only with the risk of uterine scar incompetence but also with reduced compensatory-adaptive responses in the placenta. According to several studies, intrapartum fetal distress can indeed develop in the context of uterine scar incompetence, including cases accompanied by abnormalities of labor. This association was also highlighted in the work of Geetha [27].

Conclusions

As a result of our retrospective analysis, the key clinical and anamnestic factors contributing to the development of uterine scar incompetence were identified.

Emergency cesarean delivery due to cephalopelvic disproportion may increase the likelihood of incorrect identification of the lower uterine segment, potentially resulting in the inclusion of cervical glands within the incision during closure. Tissue healing processes require a defined period to complete the necessary stages of repair, including neovascularization; for myometrial scar tissue, this period is estimated at 2 years and 9 months. A subsequent pregnancy occurring before this interval is associated with a higher risk of developing uterine scar incompetence. The postpartum course is also of considerable significance, as groups with uterine scar incompetence demonstrated a higher incidence of complications such as lochiometra and acute postpartum endometritis.

In patients with pre-existing uterine scar incompetence, labor was more frequently preterm. Neonatal outcomes, however, were more favorable in the group that underwent vaginal delivery.

Our study demonstrated that uterine scar incompetence may be accompanied by acute intrapartum fetal distress, which, according to placental morphological examination, is associated with a reduction in compensatory-adaptive reactions. Groups with an incompetent scar were characterized by inflammatory changes in the placenta, impaired chorionic villus maturation, and involutive-destructive alterations.