Abstract
Differentiating between acute fatty liver of pregnancy (AFLP) and haemolysis, elevated liver enzyme and low platelet (HELLP) syndrome may be difficult. The treatment for both conditions is expedited delivery, with the infrequent exception of expectant management with close monitoring and consideration of dexamethasone with HELLP in the setting of severe prematurity. Correct diagnosis may have implications for maternal morbidity and mortality, risk of recurrence, risk to the child, and the risk of long-term maternal disease. The Swansea criteria lack specificity, a significant proportion of women with HELLP fulfilling criteria. Low antithrombin III, elevated circulating soluble Fms-like tyrosine kinase-1, hypocholesterolaemia, hypofibrinogenaemia, and reduced portal vein blood flow velocity have been proposed as differentiating AFLP from HELLP. A case of presumed AFLP associated with hypocholesterolaemia is presented, and the literature regarding markers in the differentiation of AFLP from HELLP is discussed. Hypocholesterolaemia may be useful in differentiating between AFLP and HELLP syndrome, particularly in resource-limited settings.
Keywords
Introduction
Acute fatty liver of pregnancy (AFLP) and haemolysis, elevated liver enzyme and low platelet (HELLP) syndrome may share common clinical features and laboratory investigations, complicating differentiation of the two disorders. The treatment for both conditions is expedited delivery, with the infrequent exception of expectant management with close monitoring and consideration of dexamethasone with HELLP in the setting of severe prematurity. Correct diagnosis may have implications for index pregnancy maternal morbidity and mortality, risk of recurrence in subsequent pregnancy, future risk to the neonate/child especially with fasting or with intercurrent illness, and the risk of long-term maternal disease.
The diagnosis of AFLP is based upon the Swansea criteria, while classification of HELLP is based upon the Tennessee or Mississippi criteria.1,2 The Swansea criteria were derived empirically, and cut-offs for criteria do not take into account physiological changes in laboratory values during pregnancy.3,4 Three recent studies found that 37%, 61%, and 94%, respectively, of women with HELLP syndrome fulfilled at least six of the Swansea criteria (overall 59%).5–7 The combination of hyperbilirubinaemia, elevated hepatic aminotransferases, prolonged prothrombin time, and antithrombin III (ATIII) level less than 65%, in the absence of alternative causes of acute hepatic failure has been proposed as an alternative criteria for the diagnosis of AFLP in pregnancy.4,8 Other features proposed to be useful in the early diagnosis of AFLP and differentiation from HELLP syndrome include hypocholesterolaemia, levels of soluble fms-like tyrosine kinase-1 (sFlt-1), placental growth factor and fibrinogen, and portal vein blood flow velocity.9–13
A case of presumed AFLP associated with severe hypocholesterolaemia is presented, and the literature regarding the potential value of serum cholesterol and sFLT-1 in differentiating between AFLP and HELLP syndrome is discussed.
Case
A 31-year-old woman was found to have a blood pressure of 180/110 mmHg at a routine antenatal visit at 37 weeks’ gestation in her first pregnancy. The woman had noted significant thirst and polyuria for the previous 2 weeks but was otherwise asymptomatic. Her medical history was significant for type 2 diabetes mellitus (T2DM) diagnosed at age 23 years, managed with metformin extended release 2 g / day, and gastro-oesophageal reflux treated with rabeprazole. The woman's pre-pregnancy body mass index was 20.8 kg/m2. HbA1c was 37 mmol/mol/ 5.5% (< 48/ < 6.5%) and liver enzyme levels were normal at 6 weeks' gestation. The woman's mother had developed T2DM in her fifth decade having previously had gestational diabetes mellitus. Testing for genetic mutations associated with maturity-onset diabetes of the young had not been performed.
Physical examination revealed a woman with pulse 90/min, blood pressure 145/105 mmHg, dual heart sounds, clear chest, no abdominal tenderness, and brisk reflexes but no clonus. Pathology testing revealed hyperbilirubinaemia, elevated hepatic aminotransferases, acute kidney injury, a prothrombin time of 14.3 s, hypocholesterolaemia, and hyperuricaemia (Table 1). Plasma glucose, albumin, and serum sodium were initially normal. Urine protein: creatinine ratio was 0.4 mg/mmol (normal < 0.3), although urine microscopy demonstrated more than 500 * 106 leucocytes / L (<10) with epithelial cells and moderate bacteria suggestive of contamination. There was no evidence for haemolysis on examination of a blood smear, serum haptoglobin was 0.28 g/L (third trimester reference interval 0.27–0.36), and reticulocyte count was 4% (non-pregnant reference interval 0.2–2.0). The liver was normal in size, echogenicity, and appearance on ultrasound.
Laboratory results.
AST: aspartate aminotransferase; ALT: alanine aminotransferase; LDH: lactate dehydrogenase; PT: prothrombin time; INR: international normalised ratio; APTT: activated partial thromboplastin time; Cr: creatinine; Hb: haemoglobin; WBC: white cell count; Plts: platelet count; AT III: antithrombin III; urine PCR: protein:Cr ratio.
The initial diagnosis was of preeclampsia with severe features of abnormal liver enzymes and acute kidney injury. The woman also fulfilled six of the Swansea criteria for the diagnosis of AFLP.
The woman was induced with oxytocin and delivered a live male birthweight 3000 g by vacuum vaginal delivery with epidural anaesthesia. Postpartum a rapid improvement was seen in hepatic aminotransferases, there was a mild temporary rise in prothrombin time and further fall in cholesterol, and a brief fall occurred in fibrinogen and platelet count without fulfilling pregnancy-specific criteria for disseminated intravascular coagulation (DIC). An ATIII level was 26%. The rise in prothrombin time and fall in fibrinogen without evidence for DIC, and very low ATIII were consistent with the woman having AFLP as well as preeclampsia. Testing for mutations in LCHAD and MODY genes is pending. Prior to pregnancy, and 3 months postpartum while taking metformin 2 g/day, the woman's fasting total cholesterol levels were 4.6 mmol/L and 5.0 mmol/L, respectively.
Discussion
Differentiation between AFLP and HELLP syndrome is important regarding maternal morbidity and mortality in the index pregnancy, and the risks of recurrence in subsequent pregnancies, to the child, and long-term maternal health.
Recent case series of AFLP describe maternal and perinatal fetal/neonatal mortality rates of 0–17% and 10%-27%, respectively.14–17 Maternal and perinatal fetal/neonatal mortality rates with HELLP in recent studies have been reported to be 0.2–1.2% and 7–21%, respectively.18,19 A recent comparative study found that AFLP was associated with higher rates of maternal morbidity than with HELLP syndrome, including acute kidney injury (62% vs 15%, p < 0.001), sepsis (48 vs 10%, p < 0.001), DIC (67% vs 9%, p < 0.001), and hypoglycaemia (29% vs 2.5%, p = 0.001). 19
The risk of recurrence of HELLP syndrome in subsequent pregnancies has been reported as 2–24%, while 18–28% of women developed preeclampsia, and 24–52% developed hypertensive disorders of pregnancy overall.20–24 A review of 10 small case series of women with an index pregnancy complicated by AFLP found 1 case of ALP in 35 subsequent pregnancies (2.9%). 25
Studies examining long-term maternal morbidity following a pregnancy complicated by HELLP found new-onset hypertension in 33%, requirement for chronic dialysis in 2.4%, and psychiatric disorders in 26–32%. 24 Women with a history of preeclampsia are at increased long-term risk of chronic kidney disease, end-stage kidney disease, chronic hypertension, cardiovascular disease, diabetes mellitus, cerebrovascular disease, mental health disorders, and hospitalisation for liver disease.26–30 A survey in Brazil found, however, that 98% of women who had experienced preeclampsia were unaware of the possible long-term repercussions. 31 Other than a single study of 25 women finding no ongoing issues with hepatic or kidney disease 6 months postpartum following AFLP, the authors did not identify any studies examining long-term risks following a diagnosis of AFLP. 32
AFLP may be associated with maternal heterozygosity and neonatal homozygosity or compound heterozygosity for genetic variants associated with long-chain hydroxyacyl-CoA dehydrogenase deficiency and mitochondrial trifunctional protein deficiency. The most common pathogenic variant is in the c.G1528C in exon 15 of the α-subunit which alters amino acid 474 from glutamic acid to glutamine (E474Q) of the HADHA gene, which causes isolated LCHAD deficiency. There is significant geographical variation in the frequency of HADHA gene variants, with c.G1528 > C carriers in 1/169 in Poland, 1/680 in the Netherlands, and 0/1200 screened individuals from China. 33 Neonates with severe phenotype may present within days of birth with hypoglycaemia, hepatomegaly, and cardiomyopathy. 34 The intermediate phenotype is characterised by hypoketotic hypoglycaemia precipitated by infection or fasting in infancy. Late-onset phenotype is characterised by myopathy, neuropathy, and retinopathy. Parents of children from pregnancies complicated by maternal AFLP, and health professionals involved in their care, need to be aware of the avoidance of the risk of acute complications particularly during intercurrent illness associated for vomiting or fasting.
Total cholesterol levels rise physiologically from second trimester due to increased hepatic production, with values between 35 and 37 weeks’ gestation ranging between 4.5 and 9.2 mmol/L in healthy singleton pregnancy. 35 Hypocholesterolaemia has been defined as a serum total cholesterol less than the 5th percentile of the general population, adjusted for age, sex and race. 36 Authors have described cut-off levels for hypocholesterolaemia in non-pregnant adults lower than 3.1–4.2 mmol/L.37,38
Six case series have assessed serum total cholesterol at presentation in a total of 156 women with AFLP, finding a mean level of 2.6 mmol/L (Table 2).
Mean serum total cholesterol levels at presentation and nadir in case series of AFLP.
n: number of individuals; NS: not stated; AFLP: acute fatty liver of pregnancy.
A meta-analysis of seven studies examining lipid profiles in preeclampsia demonstrated that total and LDL cholesterol levels were significantly higher, while HDL levels were lower than in a control group. 44
Byrne et al. described significantly lower values for serum total cholesterol in women AFLP compared with women with HELLP (p < 0.001), although some overlap existed between the two groups. 9
Severe depression of ATIII was first described in four women with AFLP by Liebman et al. in 1983. 45 Sixty-one cases of AFLP where ATIII levels were described found the mean level to be 16.1% (range 0–69%). 41 A review of case reports of AFLP in Japan found that 47 of 48 cases (98%) had an ATIII level less than 65%. 17 Mean levels of ATIII in preeclampsia and HELLP range between 60 and 85%.8,46–50
Levels of sFlt-1 and placental growth factor are significantly higher in women with AFLP than HELLP. Two studies including 18 women with AFLP and 60 women with HELLP found that a sFlt-1 cut-off of 31 100 pg/mL allowed differentiation between the two disorders with 100% sensitivity and specificity.10,11 A linear correlation was found between the cumulative numbers of Swansea criteria and sFlt-1 levels. 10
Serum cholesterol levels are low in both acute and chronic liver failure due to multiple aetiologies in the general population.51,52 In non-pregnant individuals with acute hepatitis E low levels of cholesterol are predictive of development of acute liver failure. 53 Similarly, low levels of ATIII occur in both acute and chronic liver failure.54–57 Thus, low levels of cholesterol and ATIII are unlikely to be useful in differentiating between AFLP and other causes of acute liver failure in pregnancy.
A study in the United States found that non-pregnancy-specific diagnoses accounted for 35 of 70 cases of acute liver failure in pregnancy. 7 Thirty-two of these 35 women, and all of the other 35 women with pregnancy-specific aetiologies (HELLP and AFLP) fulfilled Swansea criteria. In areas where hepatitis E virus is endemic, viral hepatitis may be the cause of acute liver failure in 45–75% of women with acute liver failure in pregnancy.58,59 Studies investigating whether sFLT1 levels differentiate between AFLP and other causes of acute liver failure in pregnancy would be valuable.
The diagnoses of preeclampsia and AFLP are not mutually exclusive, several case reports describing the co-existence of the two conditions.60–64
Conclusion
Together with low ATIII and elevated sFlt-1 values, hypocholesterolaemia may be a useful marker to rapidly distinguish AFLP from HELLP in the acute setting, particularly in resource-limited areas, which may not have access to ATIII or sFlt-1 assays.
Footnotes
Ethics approval
Ethical approval was provided by Mater Health Human Research and Ethics Committee QACR/MML/120521 (V2)
Informed consent
Written informed consent was provided by the patient for publication of this article.
Authors’ Contributions
NG, JP, and AM cared for the patient, performed a literature review, and wrote the manuscript.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Guarantor
NG
