Outcomes of Early-onset Preeclampsia With Severe Features at the University Hospital in Southern Thailand: a 15-year Experience

Purpose: To evaluate outcomes and factors associated with adverse outcomes among patients with early-onset preeclampsia with severe features at Songklanagarind Hospital. Methods: A retrospective study of 326 singleton women with early-onset preeclampsia with severe features treated at Songklanagarind Hospital between 2004-2019 was conducted. Baseline characteristics, management and outcomes were reviewed. Multivariate logistic regression was used to evaluate predictors of adverse outcomes. Statistical signicance was set at p < 0.05. Results: There were no maternal mortalities, with 3.1% stillbirths and 6.7% neonatal deaths. High maternal serum creatinine (OR 3.26, 95% CI 1.27-8.36, p = 0.01) was signicantly associated with adverse maternal outcomes. Early gestational age at delivery [< 28 weeks (OR 16.63, 95% CI 6.95-39.80, p <0.01), 28-32 weeks (OR 3.24, 95% CI 1.54-6.85, p <0.01)], maternal diabetes mellitus (OR 5.62, 95% CI 1.43-22.06, p = 0.01), high maternal serum creatinine (OR 2.66, 95%CI 1.20-5.93, p = 0.02) and elevated serum aminotransferases (OR 2.26, 95% CI 1.19-4.29, p = 0.01) were associated with serious adverse perinatal outcomes. Conclusions: Early-onset preeclampsia with severe features had favorable outcomes. Maternal diabetes mellitus, high serum creatinine, elevated serum aminotransferases and early gestational age at delivery were factors associated with poor outcomes.


Introduction
Preeclampsia is a pregnancy-speci c multiple-organ disorder associated with high rates of maternal and neonatal morbidity and mortality. Recent studies reported incidence ranges from 2.7-8.2% globally and 2.2% in Thailand [1,2]. Appropriate diagnosis and timely management are crucial as more than half of these morbidities and mortalities are preventable [3].
The diagnosis of preeclampsia is made by recognizing new onset hypertension plus proteinuria and/or other signs and symptoms of organ dysfunction during pregnancy [4]. Preeclampsia is classi ed into early-and late-onset, de ned as those requiring delivery before or after 34 weeks of gestation [5]. The pathophysiology of early-onset preeclampsia is abnormal placental development leading to de cient uterine spiral artery remodeling, while late-onset preeclampsia mainly arises from degenerative placenta accompanied by maternal predisposition to cardiovascular or metabolic disease. Both types of preeclampsia lead to dysfunctional maternal peripheral endothelial cells and systemic in ammatory responses, which lead to organ dysfunction and death in some women [6][7][8][9]. The fetuses are also affected from utero-placental insu ciencies. They have higher risk of growth restriction, hypoxia and death [8,9].
Since the cause of preeclampsia is principally from the placenta, termination of pregnancy is the de nitive treatment for the affected women. However, when preeclampsia is diagnosed early in gestation, physicians have to take into account the high neonatal mortality and morbidity related to preterm birth.
[10] Attempts to prolong pregnancy at risk of preeclampsia using expectant management practices may improve neonatal outcomes. On the other hand, pregnancy prolongation may increase morbidity and mortality in some women [4,11].
Songklanagarind Hospital is the university hospital and tertiary care center in Southern Thailand, and thus receives a large number of referral preeclampsia women from all of the southern provinces. In 2018, the incidence of preeclampsia with severe features in our hospital was 4.2% of the total births (140/3299), of which about 11% (15/140) were early-onset. Early-onset preeclampsia with severe features is a condition which needs suitable guidelines, meticulous evaluation, and good physician judgment. To date, there have been no studies evaluating maternal and neonatal outcomes among cases of early-onset preeclampsia with severe features in Southern Thailand. Therefore, we aimed to evaluate maternal and neonatal outcomes among women with early-onset preeclampsia with severe features and also possible predictive factors for poor maternal and neonatal outcomes.

Material And Methods
A retrospective analytical study was conducted after approval by the Institutional Review Board of the Women were identi ed as having adverse outcomes if they had at least one of the following: 1) death; 2) abruptio placentae; 3) pulmonary edema; 4) posterior reversible encephalopathy syndrome (PRES); and 5) postpartum hemorrhage. Serious adverse perinatal outcome was de ned as at least one of the following; 1) stillbirth; 2) neonatal death; and 3) low 5-minutes Apgar score (< 7).
Data were recorded using EpiData Version 3.1. Statistical analysis was done using Stata 14.2.
Mean (SD), median (IQR) or number (%) were used as appropriate regarding the type of variables and data distribution. Multivariate logistic regression was used to estimate odds ratios for the association between variables and adverse maternal and perinatal outcomes. Statistical signi cance was set at p < 0.05.

Results
There were a total of 41,884 singleton deliveries during the 15-year period (from November 1, 2004 to October 31, 2019) at Songklanagarind Hospital, of whom 2087 were diagnosed with preeclampsia and 392 women were diagnosed with early-onset preeclampsia with severe features. Sixty-six women were excluded (23 due to severe medical problems and 43 due to serious antepartum complications). Finally, 326 early-onset preeclampsia with severe features women were analyzed. (Fig. 1) The demographic data are shown in Table 1. Most of the diagnoses were based on elevated blood pressure de ned as systolic blood pressure ≥ 160 mm Hg and/or diastolic blood pressure ≥ 110 mm Hg (314/326 women, 96.3%) and signi cant proteinuria de ned as urine protein ≥ 300 mg/24 hours or spot urine protein to creatinine ratio ≥ 0.3 or urine protein dipstick ≥ 2+ (301/326 women, 92.3%), with 178/326 women (54.6%) reporting the symptoms of headache, blurred vision or epigastrium pain. The pre-delivery management followed the ACOG recommendations [4,14] in nearly all cases. Only 5/326 (1.5%) women did not receive magnesium sulfate for maternal neurostabilization. Gestational age at start (weeks) 14.5 (12,17) Duration of ASA usage (weeks) 14 (11,16) Most deliveries were induced (301/326, 92%), with uncontrolled blood pressure being the most frequent indication. Cesarean section was performed in the majority of cases. Hydralazine was the most frequently used anti-hypertensive drug followed by nifedipine and labetalol. Nicardipine was used only in women whose blood pressure failed to be controlled by the above-mentioned drugs. Expectant management was offered in less than half of the women, with time from diagnosis to delivery ranging from 2 hours to 29 days. (Table 2) The maternal and neonatal outcomes are shown in Table 3. There was no maternal death. High percentages of uncontrolled blood pressure and persistent symptoms were observed. HELLP syndrome was noted in approximately one-third of cases. There were no eclamptic women during expectancy. Rates of other complications including pulmonary edema, abruptio placentae, postpartum hemorrhage and posterior reversible encephalopathy syndrome (PRES) were low. Three out of ve women with abruptio placentae were detected during expectant management (within 48 hours after diagnosis), with the rst sign being non-reassuring fetal heart rate; and two out of three had di culties in blood pressure control. The other two women with abruptio placentae were detected at the time of diagnosis and received immediate termination of pregnancy. Regarding nine cases of postpartum hemorrhage, all had successful conservative management. Regarding perinatal outcomes, about 90% of cases were alive with high rate of neonatal intensive care unit admission. Other complications were related to preterm delivery.
Univariate analysis showed that high maternal serum creatinine (> 1.1 mg/dL) was the only factor associated with adverse maternal outcomes (OR 3.26, 95% CI 1.27-8.36, p = 0.01). Multivariate analysis showed that early delivery at gestational age (GA) < 32 weeks, maternal diabetes mellitus, high maternal serum creatinine and elevated serum aminotransferases (AST or ALT > 70 IU/L) were risk factors for serious adverse perinatal outcomes adjusted by gestational age at diagnosis and maternal low platelet count. (Table 4) Table 4 Multivariate analysis of factors associated with serious adverse perinatal outcomes a

Discussion
In this study, women with early-onset preeclampsia with severe features had favorable outcomes. High maternal serum creatinine was associated with adverse maternal outcomes, while delivery before 32 weeks of gestation, maternal diabetes mellitus, high serum creatinine and elevated serum aminotransferases increased the risk of adverse perinatal outcomes.
Similar to other studies, there was very low rate of maternal death among women with early-onset preeclampsia with severe features [13][14][15][16][17][18]. The incidence of HELLP syndrome was similar to that in a previous report in Thailand, about 28% [13], but was more than those reported in other countries, which ranged from 12-15% [14][15][16][17], which might be explained by the differences in gene expression in placenta, concentrations of anti-angiogenic factors or degrees of in ammatory responses in different ethnicities [19]. The rates of abruptio placentae and eclampsia reported in our study were low. Since two out of three women with abruptio placentae during expectant management had uncontrolled blood pressure, and all had non-reassuring fetal heart monitoring, we suggest that physicians should emphasize on blood pressure control and close observation of fetal health status. Poorly controlled blood pressure needs expedited delivery. In regard to perinatal outcomes, the perinatal mortalities were not different among various parts of the world [13][14][15][16][17].
We found that high maternal serum creatinine was the only factor associated with adverse maternal outcomes. It is understandable as kidney is one of the most frequently affected organs from in ammatory response and/or vasoconstriction in preeclampsia [6,20]. Once kidney dysfunction is evident, it indicates that other organs might be deteriorated. It is generally accepted that preterm delivery and maternal diabetes mellitus are related to adverse neonatal outcomes [10,21]. They were also associated with poor perinatal outcomes in our study. Apart from gestational age at delivery and maternal diabetes mellitus, we found that high maternal serum creatinine and elevated serum aminotransferases, which re ect the severity of preeclampsia, were also associated with serious adverse neonatal outcomes. Ganzevoort et al. reported that early gestational age at admission was the only signi cant factor in uencing adverse neonatal outcomes. However, they did not evaluate the associations between maternal diabetes mellitus or maternal serum creatinine level and adverse outcomes [22].
The strengths of this study were; 1) large sample size; 2) complete and reliable database; 3) diagnoses and managements adherent to ACOG guidelines; and 4) determination of factors associated with adverse maternal and perinatal outcomes.
The limitation of our study was that it was a retrospective study. There were a few incomplete data such as laboratory results from referring hospitals, and patient histories and physical examination ndings depended on the level of completion of medical records at the time of diagnosis. As aspirin prescription was just recently added to the preeclampsia prevention guideline in 2016 [23], there was only a small percentage (6.7%) of aspirin prescription in our data. Because of this small percentage, the effects of aspirin prescription on the severity of preeclampsia might not be apparent.
According to the favorable outcomes of the management of early-onset preeclampsia with severe features in our study, we encourage the use of the ACOG practice guideline. In addition, knowing factors associated with adverse maternal and perinatal outcomes can guide physicians in patient counselling and providing optimal neonatal care planning.

Conclusion
Women with early-onset preeclampsia with severe features had favorable outcomes. High maternal serum creatinine was associated with adverse maternal outcomes, while early gestational age at delivery, maternal diabetes, high serum creatinine and elevated serum aminotransferases increased the risk of adverse perinatal outcomes. The study ow chart