Acute kidney injury associated with preeclampsia or hemolysis, elevated liver enzymes, and low platelets syndrome

Abstract

Objective:

To determine the prevalence of acute kidney injury (AKI), placental abruption and postpartum hemorrhage in patients with preeclampsia or HELLP syndrome.

Study design:

A retrospective study of patients with preeclampsia or HELLP syndrome treated at the University of Mississippi Medical Center from January 2000 through December 2010.

Main outcome measures:

Relationships among the obstetric complications of placental abruption, postpartum hemorrhage, and AKI (serum creatinine > 107 μmol/L) of women with preeclampsia or HELLP syndrome. Additional analysis was undertaken to explore if there was a correlation between postpartum hemorrhage/placental abruption and the severity of HELLP syndrome according to the Mississippi classification system.

Results:

Data from 1276 women over 11 years were included in the analysis. 67 of 466 patients (14.4%) with HELLP syndrome and 38 of 810 preeclampsia patients (4.7%) met criteria for AKI. Women with either placental abruption or postpartum hemorrhage had statistically significant increased odds of also having AKI (p < 0.01). Women with HELLP and AKI were also more likely to experience either placental abruption or postpartum hemorrhage. Women with Class 1 HELLP with placental abruption or postpartum hemorrhage were also more likely to have AKI than women with preeclampsia.

Conclusion:

HELLP syndrome, AKI and placental abruption or postpartum hemorrhage appear to be interrelated. AKI occurs more frequently in women with HELLP syndrome with or without associated postpartum hemorrhage and placental abruption.

1. Introduction

The occurrence of acute kidney injury (AKI) during pregnancy has been increasing in both frequency and severity in developed countries [1,2]. Moreover, the prevalence of AKI has amplified nearly 10-fold in industrialized countries, despite radical improvements in reproductive healthcare [3]. AKI during maternal gestation typically occurs in women with other pregnancy related conditions, such as preeclampsia or HELLP (hemolysis elevated liver enzymes low platelet count) syndrome, eclampsia and/or postpartum hemorrhage [4-6].

Regardless of the underlying cause, AKI is a potentially life-threatening disorder that can have long-lasting consequences for both mother and infant [7]. In the current study, our primary objective is to determine the incidence of AKI in patients with preeclampsia or HELLP syndrome. Due to the likelihood of placental abruption and postpartum hemorrhage among women with AKI [8-10], we also sought to determine if AKI increased the occurrence of these comorbidities.

2. Materials & methods

In this retrospective chart review the medical records of all pregnant patients coded with a discharge diagnosis of preeclampsia or HELLP Syndrome (January 1, 2000 through December 31, 2010) at the University of Mississippi Medical Center were reviewed under IRB approval. Detailed patient characteristic data has previously been published [8,11-13].

Preeclampsia was defined based on systolic blood pressure ≥ 140 mmHg and diastolic ≥90 mmHg after 20 weeks of gestation and the presence of proteinuria, while HELLP syndrome was diagnosed based on the Mississippi classification (Table 1) [8,14]. AKI was defined as a peak serum creatinine ≥107 μmol/L in the absence of other renal disease. The data abstracted included maternal and gestational age at delivery, parity, peak serum creatinine, placental abruption, postpartum hemorrhage, acute renal failure, and maternal death. The recorded occurrence of placental abruption and estimated blood loss (EBL) at delivery or incidence of obstetric postpartum hemorrhage were noted. Postpartum hemorrhage was defined as EBL ≥ 500 mL in women undergoing vaginal delivery and ≥1000 mL in women with caesarean deliveries. If a patient had documented, pre-existing renal disease, she was excluded from the current analysis.

Table 1.

HELLP Syndrome classification system.

HELLP Class	Mississippi classification
1	Platelet Count ≤ 50,000/μL AST or ALT ≥ 70 IU/L Total LDH ≥ 600 IU/L
2	50,000/μL ≤ Platelet count ≤ 100,000/μL AST or ALT ≥ 70 IU/L Total LDH ≥ 600 IU/L
3	100,000/μL ≤ Platelet count ≤ 150,000/μL AST or ALT ≥ 40 IU/L Total LDH ≥ 600 IU/L

2.1. Statistical analysis

Data were analyzed using R statistical software (version 3.3.1). Descriptive summary statistics were computed and summarized in terms of means and proportions. Mean level of continuous variables (mean maternal age and mean gestational age at delivery) were compared among diagnosis type categories using t-test. A chi-square test or Fisher’s exact test (depending on participant counts within each category) was employed to explore relationships between categorical variables. Finally, adjusted odds ratios and 95% confidence intervals were obtained using multivariate logistic regression. Statistical significance was considered at P-value < 0.05.

3. Results

A total of 1,295 women were identified with preeclampsia or HELLP syndrome during the study period. The findings from 1,276 patients were included in the study; 15 women were excluded due to incomplete records and 4 were determined to have evidence of pre-existing renal disease (Fig. 1). 466 women had HELLP syndrome and 810 women met criteria for preeclampsia; women with a diagnosis of preeclampsia did not have evidence of HELLP syndrome. Overall, 105 (8.2%) women developed AKI, 67 (14.4%) of whom had a diagnosis of HELLP syndrome, which was significantly greater than the 38 (4.7%) women with a diagnosis of preeclampsia (p < 0.0001). There were three maternal deaths in the total group of women studied. One of the three had Class 3 HELLP + AKI + Placental abruption; one had preeclampsia + AKI and the other had preeclampsia only.

Fig. 1.

Flow diagram of women diagnosed with PreE (preeclampsia) or HELLP syndrome at the University of Mississippi Medical Center between January 1, 2000 through December 31, 2010.

There were no statistically significant differences in maternal age (P = 0.61), maternal race (P = 0.09), gestational age at delivery (P = 0.26) or null parity (P = 1) between women with HELLP syndrome and women with preeclampsia, who also developed AKI. Similar results are reported among women who experienced postpartum hemorrhage and/or placental abruption regardless of HELLP, preeclampsia or AKI diagnosis (Table 2). The single exception to this were women with HELLP syndrome, absent of any evidence of AKI, postpartum hemorrhage or placental abruption; these patients delivered significantly earlier compared to their preeclamptic counterparts (31.3 ± 4.7 vs. 32.5 ± 4.1 weeks; P < 0.001).

Table 2.

Characteristics in Patients with AKI vs. No AKI.

		AKI			Non-AKI
		PE	HELLP	P value	PE	HELLP	P value
Maternal Age (yrs)	Hemorrhage	24.8 ± 6.6	24.8 ± 6	0.99	25.4 ± 6.2	25.3 ± 6.4	0.98
	Placental Abruption	25.6 ± 5.6	24.9 ± 6.7	0.83	24.8 ± 5.8	24.2 ± 4.7	0.70
	Hemorrhage+Placental Abruption	25 ±: 4.2	24.83 ±: 6.4	0.97	25.1 ± 5.7	24.6 ± 2.6	0.84
	No Comorbidity	27.9 ± 7.2	24.8 ± 6.3	0.06	24.7 ± 6.3	24.0 ± 5.8	0.09
Gestational Age (wks)	Hemorrhage	34.3 ± 5.5	31.6 ± 3.1	0.22	31.1 ± 5.1	32.0 ± 3.7	0.59
	Placental Abruption	30.4 ± 2.5	29.6 ± 3.3	0.59	30.6 ± 3.9	29.2 ± 3.3	0.21
	Hemorrhage+Placental Abruption	30.0 ± 2.2	30.0 ± 1.4	0.97	30.4 ± 4.7	31.5 ± 3.0	0.59
	No Comorbidity	31.5 ± 5.0	30.7 ± 5.2	0.52	32.5 ± 4.1	31.3 ± 4.7	< 0.001
Nulliparous (%)	Hemorrhage	25	27	1	47	38	0.73
	Placental Abruption	20	25	1	47	40	0.77
	Hemorrhage+Placental Abruption	0	17	1	58	25	0.19
	No Comorbidity	34	50	0.22	43	48	0.10

Hemorrhage is postpartum hemorrhage.

There was a significant difference when women were diagnosed with AKI (antepartum vs postpartum; P = 0.01). Women with preeclampsia were more likely to be diagnosed with AKI in the antepartum period (n = 29; 75%) compared to women with HELLP syndrome who had equal chances of being diagnosed with AKI in the antepartum (n = 29; 46%) or postpartum (n = 34; 54%) period. The timing of AKI diagnosis could not be extracted from the medical records of 4 women with HELLP syndrome and 2 women with preeclampsia.

3.1. Placental abruption and postpartum hemorrhage are more prevalent in women with HELLP + AKI

Seventy-one women developed placental abruption (5.5%), 21 also had a diagnosis of AKI (Fig. 2A). A sub-analysis of women with placental abruption found that women with HELLP syndrome with AKI (n = 16) were significantly more likely to have placental abruption compared to women with preeclampsia with AKI (n = 5; P = 0.005). There was not a significant difference in the timing of the diagnosis of AKI among women who also had placental abruption (P = 0.52). A total of 54 (4.2%) women were diagnosed with postpartum hemorrhage, 19 of which were also diagnosed with AKI (Fig. 2B). After subanalysis, women with HELLP syndrome with AKI (n = 15) were significantly more likely to have postpartum hemorrhage compared to women with preeclampsia with AKI (n = 4; P = 0.02). Among this same group of women, the timing of AKI diagnosis was not significantly different (P = 0.55). Twenty-eight (2.1%) of the women in this study developed both placental abruption and postpartum hemorrhage and 8 of these women also had a diagnosis of AKI (Fig. 2C). A subanalysis of this group of women did not find a statistically significant difference among women with AKI (P = 0.09). There was also not a significant difference in the timing of the diagnosis of AKI among women who had both placental abruption and postpartum hemorrhage (P = 0.7). Furthermore, when the relationship between placental abruption and AKI was assessed, women in our study group who had placental abruption had a six-fold increased risk of also having AKI (O.R. = 6.57; 95% CI 3.18–12.88; Fig. 2D). Similar results were true for women with postpartum hemorrhage, as they had an eleven-fold increased risk of developing AKI (O.R. = 11.11; 95% CI 4.82 – 24.93; Fig. 2D). However, women who experienced both postpartum hemorrhage and placental abruption paradoxically had significantly decreased odds of also having AKI (O.R. = 0.08; 95% CI 0.02 – 0.31; Fig. 2D).

Fig. 2.

The percent of placental abruption, postpartum hemorrhage or the combination of both was calculated among women with preeclampsia and HELLP with and without AKI. There was a significant association between the diagnosis of HELLP with AKI vs. preeclampsia with AKI and the development of placental abruption (P = 0.005; A) and in the development of postpartum hemorrhage (P = 0.02; B). There was not a statistically significant relationship between AKI and the development of placental abruption and postpartum hemorrhage (P = 0.09; C). The numbers above the bars indicate the numbers of women in each category. Adjusted odd-ratios with 95% confidence interval (CI) from multiple logistic regression were determined. Computed CI’s do not intersect with 1, indicating odd-ratios of placental abruption, postpartum hemorrhage and placental abruption combined with postpartum hemorrhage that are significantly different from 1 (D).

3.2. Women with Class 1 HELLP syndrome are more likely to develop AKI

Previously we reported that the Mississippi classification system of HELLP syndrome, based primarily on severity of thrombocytopenia, is important for the management of women with HELLP [8]. The odds of developing AKI in women who developed Class 2 HELLP were 73% lower (O.R. = 0.27; 95% CI 0.14 – 0.52) compared to women with the most severe disease as Class 1 HELLP. Similarly, women with Class 3 HELLP had odds of AKI that were 68% lower (O.R. = 0.32; 95% CI 0.17 – 0.60) compared to women with Class 1 HELLP (Fig. 3A). We performed a sub-analysis among women with HELLP + AKI and did not find a statistically significant association between the classes of HELLP syndrome with regard to the incidence of postpartum hemorrhage (P = 0.22) or placental abruption (P = 0.11) or the combination of placental abruption and postpartum hemorrhage (P = 0.29; Fig. 3B).

Fig. 3.

Women with HELLP syndrome were divided into Class 1, 2 and 3 based on the Mississippi classification system [8] and subdivided based on AKI status (A). There was a significant difference in the development of AKI among the different classes of HELLP syndrome (P = 0.04). When the occurrence of placental abruption, postpartum (PP) hemorrhage or the combination of both placental abruption and PP hemorrhage was compared to women with HELLP syndrome in the same class minus AKI, there was not a significant difference among the different classes of HELLP syndrome. The numbers above the bars indicate the numbers of women in each category.

4. Discussion

In this study of women with preeclampsia and HELLP syndrome, the most remarkable finding is the apparent association of HELLP syndrome with AKI and a very high prevalence of placental abruption or postpartum hemorrhage in women with HELLP syndrome + AKI. Important consequences of placental abruption to the mother include postpartum hemorrhage, possible blood transfusion, disseminated intravascular coagulopathy and a host of other complications related to sudden hypovolemia and significant stress on the mother’s coagulation system [15].

It has been suggested that postpartum hemorrhage may also be a contributing factor to AKI in pregnant patients with or without an apparent associated placental abruption [15-17]; however, in our study we did not find a significant association between the development of AKI and postpartum hemorrhage with or without placental abruption. Our findings of an increased incidence of placental abruption in women with HELLP syndrome are in agreement with other studies. For instance, Bentata et al performed a mini meta-analysis of pregnancy-related AKI in developing countries and reported that the incidence of HELLP syndrome and placental abruption were increased in women with AKI [10]. Recently a group of investigators reported that women with HELLP syndrome have an increased risk of placental abruption, but they were unable to determine if there was a relationship between placental abruption to any other risk factor [18].

In the current study we found that among women with Class 1 HELLP who experienced postpartum hemorrhage, 75% of them also had AKI, compared to 50% of the Class 2 and 28.6% of Class 3 HELLP patients (Fig. 3B). 80% of the women with Class 1 HELLP, 60% of women with Class 2 HELLP, and 25% of women with Class 3 HELLP who experienced placental abruption also had AKI. These results suggest that women with Class 1 HELLP are at an increased risk of developing placental abruption or postpartum hemorrhage compared to women with less severe classes of HELLP syndrome. There appears to be a relationship between the defective placental/obstetrical hemorrhagic tendencies of patients with HELLP syndrome and the occurrence of AKI, which greatly exceeds what is observed in patients with preeclampsia.

The HELLP syndrome-related propensity to injure the kidney, prematurely separate the placental attachments and initiate excessive intrapartum and postpartum hemorrhage may develop due to non-immunological, non-inflammatory causative factors. In recent years, the pathogenesis of HELLP syndrome has been widely attributed to an exaggerated inflammatory, immunologically driven injury process to the vascular system that theoretically should be responsive to corticosteroids [19,20]. A similar observation of 5–24% renal morbidity in patients across the spectrum of HELLP severity, inclusive of Class 1, 2, and 3, compared to only 2% of 688 patients with severe preeclampsia absent of HELLP syndrome has also been reported [8].

While most studies report that pregnancy-related AKI is associated with an increase in maternal death, only 1.9% of mothers in this study died. A relationship between the maternal kidney, the placenta, and hypertensive pregnancy, with or without HELLP syndrome, is not a new concept. Women with a history of AKI prior to pregnancy are well known to have an increased risk of developing pregnancy complications [21]. Women with preeclampsia were more likely to be diagnosed with AKI in the antepartum period compared to women with HELLP syndrome who were equally as likely to develop AKI in the antepartum as they were in the postpartum. As the underlying mechanisms surrounding AKI and its role in hypertensive pregnancies is still unclear it is uncertain if there is a relationship between placental factors and renal injury which would increase the risk for AKI diagnosis in the antepartum period.

As an observational study, our findings suffer from the limitations inherent in all retrospective investigations. We are confident that all cases were identified and included in the analysis, minimizing the risk of selection bias. The study population is fairly large. Criteria for selection are subject to the criticism of using a relatively high serum creatinine threshold for inclusion, since it is well known that biomarkers for AKI can be positive before elevations in serum creatinine are evident; the diagnosis of AKI limited to women who only had serum creatinine levels greater than 107 μmol/L. Unfortunately as this study is a secondary analysis of a preeclampsia and HELLP syndrome database we were unable to include women who may have had preeclampsia during this time frame based on the current preeclampsia guidelines [22]. As recent studies utilizing the new diagnostic guidelines for preeclampsia have reported significantly higher rates of AKI than what is reported in the current study (15.3–16.9%) it is probable that more women with preeclampsia would indeed have a diagnosis of AKI [23,24]. We acknowledge that there is likely more AKI in HELLP syndrome patients than our criteria and methodology, absent of biomarker information, could provide. Unfortunately, urine samples from the study population were neither procured nor saved for later analysis, thus obviating analysis using newly available biomarkers to detect early evidence of renal compromise.

5. Conclusions

Results from the current study suggest that perhaps the pathogenesis of HELLP syndrome involves even greater potential for AKI than preeclampsia, resulting in major long-term implications for affected patients that may develop end-stage renal disease later in life. Further prospective studies are needed to clarify the pathogenesis and chronology of renal and placental injury related to HELLP syndrome. Investigation of the interrelationship of acute kidney injury, which occurs within the pregnancy spectrum of preeclampsia, HELLP syndrome, and the thrombotic microangiopathies, including atypical hemolytic uremic syndrome, could lead to new diagnostic tests and treatments that would benefit patients suffering the morbidity and mortality of these challenging disorders, as well as provide healthcare providers with another route to healthcare intervention [25].