Abstract
Objective
In animal models, internal iliac artery occlusion caused gestational hypertension; however, whether this phenomena occurs in humans is still unknown.
Materials and methods
This retrospective cohort study used data from the Birth Certificate Application of Taiwan and linked to the National Health Insurance Research Database and Taiwan Maternal and Child Health Database from 2008 to 2017. Women who underwent internal iliac artery occlusion before pregnant were identified according to diagnosis and procedure codes. The occlusion group included 328 births in 286 women with a history of internal iliac artery occlusion, and the non-occlusion control group included 2,024,882 births in 1,391,288 women.
Results
There were no significant differences in gestational hypertension-associated diseases including preeclampsia, eclampsia and HELLP syndrome between the occlusion and non-occlusion groups (4.3 % vs 3.4 %, p = 0.4). The adjusted odds ratios (ORs) of placental previa, placenta accreta spectrum and stillbirth were 1.69 (95 % confidence interval [CI] = 1.12–2.56), 3.99 (95 % CI = 2.52–6.31), 2.57 (95 % CI = 1.13–5.83), respectively, with the non-occlusion group as reference. The adjusted ORs of preterm delivery in the occlusion group were 1.48 (95 % CI = 1.08–2.04) and 2.79 (95 % CI = 1.62–4.82) for a gestational age below 37 weeks and 32 weeks, respectively.
Conclusion
Women who underwent internal iliac artery occlusion did not have a higher risk of gestational hypertension and related disease. Their offspring also had similar risks of small for gestational age, poor Apgar score, birth defects and neonatal mortality within 28 days. However, their risks of placental previa, placenta accreta spectrum, stillbirth and preterm delivery were increased.
Introduction
Internal iliac artery (or hypogastric artery) occlusion can preserve the uterus and reduce mortality from postpartum hemorrhage [ ]. Internal iliac artery occlusion can also reduce the amount of blood lost during myomectomy and reduce the risk of leiomyoma recurrence [ ]. However, uterine blood flow may decrease after internal iliac artery occlusion, and this decrease may affect pregnancy outcomes. In 1940, Ogden et al. reported that clamping the abdominal aorta of pregnant rats below the renal arteries resulted in gestational hypertension [ ]. In 1985, Cavanagh et al. performed similar procedures on pregnant primates, which resulted in preeclampsia [ ]. These animal models established the hypothesis that uteroplacental ischemia can cause gestational hypertension and preeclampsia [ ].
Even though the animal models implied poor pregnancy outcomes, the results in humans have been conflicting. Several cohort studies have indicated a low incidence of gestational hypertension among pregnant women with a history of uterine artery occlusion or embolization, irrespective of the etiology, as well as those who underwent bleeding control during a myomectomy [ ] or postpartum hemorrhage [ , ]. One small clinical trial conducted in eight Ontario hospitals reported that 4 of 24 pregnant women with previous uterine artery embolization for symptomatic uterine myomas developed gestational hypertension [ ]. In addition, 4 of 18 live births were small for gestational age (less than 5th percentile) [ ]. These studies were limited by a small sample size and lack of a control group, however they still provided some information on humans. Some pregnant women may arrange prenatal care or delivery at medical facilities other than where they received artery occlusion, and this can make follow-up and evaluation of perinatal effects difficult. Thus, we design a nationwide, population, retrospective cohort study using the Taiwan Maternal and Child Health Database (TMCHD) to study the perinatal outcomes of women after internal iliac artery occlusion.
Methods
Data sources
The Taiwan National Health Insurance (NHI) system is a single-payer program that was established in 1995 and covers over 99.6 % of the population in Taiwan [ ]. The National Health Insurance Research Database (NHIRD) is derived from the NHI program, and includes data on patient demographics, diagnoses, surgery or interventions during hospital admission, and outpatient claims.
The TMCHD is linked to medical claims of offspring and their parents from the Taiwan Birth Registration Database, Birth Certificate Application, National Register of Death, and NHIRD. The TMCHD contains 99.78 % of all parent-child triads (defined as one offspring and both parents) in Taiwan [ ]. These national registry databases have been evaluated and shown to be valid and complete [ ].
Study design
To investigate the perinatal outcomes, we used a retrospective cohort design. We identified 2,034,925 births in 1,399,684 women with pregnancies ending after more than 20 weeks’ gestation from the Birth Certificate Application from January 1, 2008 to December 31, 2017. Of these women, 6391 were excluded as their data were not linked to the NHIRD. Thus, the data of 2,027,981 births in 1,393,293 women were collected. Internal iliac artery occlusion was defined as women who had undergone uterine artery ligation (NHI procedure code 69040B), hypogastric artery ligation (NHI procedure code 69012B), and trans-arterial embolization (NHI procedure code 33075B plus a diagnosis related to a hemorrhage site at the pelvis, Supplementary 1 ). We excluded women who received iliac artery occlusion and gave birth during the same admission. After excluding 1054 births before the women received internal iliac artery occlusion and 1739 births in women who received iliac artery occlusion and gave birth during the same admission, population 1 included 328 births in 286 women who underwent internal iliac artery occlusion and 2,024,882 births in 1,391,288 women who did not undergo occlusion as comparison group (non-occlusion group). In order to study neonatal mortality and birth defect rates, we excluded stillbirths (n = 22,673) and those in whom birth records were not linked to the TMCHD (n = 142,180). We defined this cohort as population 2, which included 1,860,357 live births in 1,319,202 women. Population 1 was further stratified by surgical intervention (uterine artery ligation and hypogastric artery ligation) and trans-arterial embolization to analyze differences in perinatal outcomes. We also propose that distinct causes of internal iliac artery occlusion may exert varying effects on perinatal outcomes. Population 1 was categorized based on obstetric hemorrhage and non-obstetric hemorrhage, and subsequently analyzed separately (see Fig. 1 ).
The primary outcomes were to evaluate whether internal iliac artery occlusion was related to maternal gestational hypertension and related diseases (including preeclampsia, preeclampsia, and HELLP syndrome, defined as ICD-9-CM: 642; ICD-10-CM: O11.1–11.3 or O13-16) and stillbirth (ICD-9-CM: V27 excluding V27.0, V27.2, and V27.5; ICD-10-CM: Z37 excluding Z37.0, Z37.2, and Z37.5). Data were extracted if there was at least one admission or one outpatient visit. The secondary outcomes were to evaluate whether there were relationships between a history of internal iliac artery ligation with gestational diabetes, placental abruption, placenta previa, placenta accreta spectrum, preterm delivery, small for gestational age, Apgar score, neonatal mortality within 28 days, and birth defects (the diagnosis and extractive methods are listed in Supplementary 2 ). The Institutional Review Board of Chang Gung Medical Foundation approved the study (IRB Permit No 202001737B0C501) and waived the requirement for informed consent for deidentified data.
Statistical analysis
According to the Health and Welfare Data Science Center’s policy, data for items with five or fewer cases are not reported in the NHIRD or TMCHD. Pearson’s chi-squared test or Fisher’s exact test with Monte Carlo estimates was used to evaluate each categorical item. We used a multivariable logistic regression model to adjust for the covariates of gestational hypertension and perinatal outcomes, including maternal age [ ] and parity at pregnancy [ ], income level, delivery method, chronic kidney disease [ ], antiphospholipid syndrome [ ], systemic lupus erythematosus [ ], chronic hypertension [ ] and diabetes mellitus [ ] before pregnancy (the diagnosis codes are listed in Supplementary 3 ). Because women with multiple deliveries were included in the study, we used a generalized estimating equation with an exchangeable correlation structure to assess the effects of model parameters. SAS software, version 9.4, was used to conduct all statistical analyses (SAS Institute, Cary, NC). P < 0.05 was set as the statistical significance level for 2-tailed tests.
Results
The occlusion group included 328 births in 286 women, and the control (non-occlusion group) included 2,024,882 births in 1,391,288 women. The major method of internal iliac artery occlusion was by trans-arterial embolization (86 %, 282 of 328). Obstetric hemorrhage was the major etiology for internal iliac artery occlusion (90.5 %, 297 of 328). The occlusion group was older, more multiparous, and had a tendency to deliver by cesarean section compared with the non-occlusion group. There were no significant differences in financial status and maternal comorbidities including pre-existing diabetes mellitus, systemic lupus erythematosus, anti-phospholipid syndrome and chronic kidney disease between the two groups. However, a significantly higher percentage of women in the occlusion group had maternal chronic hypertension (2.1 % vs 0.7 %, p < 0.05) ( Table 1 ).
| Characteristic | Internal iliac artery occlusion (N = 328) | No occlusion of the internal iliac artery (N = 2,024,882) | P value | ||
|---|---|---|---|---|---|
| n | % | n | % | ||
| Age at pregnancy (mean ± SD) | 32.5 ± 4.6 | 30.4 ± 4.8 | <0.001 | ||
| Age at pregnancy (years) | <0.001 | ||||
| <25 | 20 | 6.1 | 235,159 | 11.6 | |
| 25–29 | 58 | 17.7 | 593,080 | 29.3 | |
| 30–34 | 131 | 39.3 | 800,296 | 39.5 | |
| 35–39 | 106 | 32.3 | 345,858 | 17.1 | |
| ≧40 | 13 | 4.0 | 50,489 | 2.5 | |
| Income (NTD per month) | 0.500 | ||||
| 0–15,840 | 42 | 12.8 | 232,638 | 11.5 | |
| 15,841–25,000 | 126 | 38.4 | 838,059 | 41.4 | |
| >25,000 | 160 | 48.8 | 954,185 | 47.1 | |
| Delivery method | <0.001 | ||||
| Cesarean section | 174 | 53.0 | 725,696 | 35.8 | |
| Vaginal delivery | 154 | 47.0 | 1,299,186 | 64.2 | |
| Parity | <0.001 | ||||
| Primiparous | 31 | 9.5 | 1,113,279 | 55.0 | |
| Multiparous | 297 | 90.5 | 911,603 | 45.0 | |
| Maternal comorbidity | |||||
| Chronic hypertension before pregnancy | 7 | 2.1 | 14,072 | 0.7 | 0.002 |
| Diabetes mellitus before pregnancy | N/A a | – | 10,823 | 0.5 | 0.101 b |
| Systemic lupus erythematosus | N/A a | – | 4428 | 0.2 | 1.000 b |
| Anti-phospholipid syndrome | N/A a | – | 3749 | 0.2 | 0.456 b |
| Chronic kidney disease | N/A a | – | 601 | 0.0 | 0.093 b |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
