Possible additional criteria for the diagnosis of preeclampsia with severe features

Abstract

Preeclampsia is a disorder affecting multiple organ systems. Preeclampsia with severe features may prompt consideration of delivery. The diagnostic criteria for preeclampsia with severe features, while focusing upon maternal cardiopulmonary, neurological, hepatic, renal and haematological systems, vary considerably in international practice guidelines. In the absence of alternative causes, severe hyponatraemia, pleural effusions and ascites, and abrupt severe maternal bradycardia are proposed as possible additional criteria for the diagnosis of preeclampsia.

Clinical practice guidelines demonstrate considerable variation in diagnostic criteria for preeclampsia with severe features (PET-SF) (Table 1). While these focus upon changes in the cardiopulmonary, hepatic, renal, neurological and haematological systems, as a multisystem disorder severe preeclampsia may cause other pathology, which may additionally represent PET-SF and prompt consideration of expedited delivery.

Table 1.

Diagnostic criteria for preeclampsia with severe features (PET-SF) – frequency of inclusion in 11 international practice guidelines.^28–31

Feature of PET-SF	Frequency in clinical practice guidelines
Severe hypertension	7
Acute kidney injury	7
Platelet count <50×109/l	7
Headache/visual disturbance	6
Right upper quadrant pain	6
Elevated AST, ALT or LDH	5
Heavy proteinuria	4
Pulmonary oedema	4
Eclampsia	4
Hepatic haematoma/rupture	4
New indication for dialysis	3
Gestational age <34 weeks	3
O2 sats <90%	2
Chest pain/dyspnoea	2
Myocardial infarction	2
Requirement for inotropes	2
Fetal growth restriction	2
Non-reassuring fetal heart rate	2
Oligohydramnios	2
Chest pain/dyspnoea	2
Nausea/vomiting	2
Oliguria	1
Clonus	1

AST: Elevated aspartate aminotransferase, ALT: alanine aminotranferase, LDH: lactate dehydrogenase.

Hyponatraemia, defined as a serum sodium <130 mmol/l, has been described in 9% to 9.7% of cases of preeclampsia.^1,2 A physiological fall in serum sodium of approximately 5 mmol/l occurs during pregnancy, and it is important that measures are made using a direct ion-specific electrode as false elevation in levels may occur when significant hypoproteinaemia is present if an indirect method is employed. One study found that pregnancies with preeclampsia and a serum sodium <128 mmol/l were complicated by acute kidney injury (AKI) in 50%, fetal growth restriction (FGR) in 46.7%, haemolysis, elevated liver enzyme, low platelets (HELLP) syndrome in 16.7%, the use of magnesium sulphate infusion in 53.3%, and maternal admission to an intensive care unit in 35.4%. ¹ The duration between the first detection of serum sodium <128 mmol/L and delivery was <24 h in 71% of women. The lowest serum sodium occurred prior to delivery in 78% of women, excluding excessive peripartum fluids or oxytocin therapy being a factor in these patients. Urine sodium was <20 mmol/L in 6 of the 11 patients measured, thus reduced placental vasopressinase production was less likely to be the major factor in causation of hyponatraemia. Serum sodium <128 mmol/l was associated with a significantly increased risk of AKI, FGR and maternal intensive care admission compared with preeclampsia with serum sodium of 128 to 129 mmol/l.

Razavi et al. ² described severe features occurring in 61.3% of cases of preeclampsia where serum sodium was <130 mmol/l compared with severe features in 16.7% of cases where serum sodium was normal. Remer and Sompolinsky ³ found a strong correlation between hyponatraemia in preeclampsia and AKI, HELLP syndrome, severe hypertension, early delivery, low birth weight and neonatal intensive care admission. Hyponatraemia typically resolves rapidly postpartum without the need for hypertonic saline.

It is important that alternative causes are excluded prior to attributing hyponatraemia to preeclampsia. An audit of 194 women with serum sodium <130 mmol/l after 20 weeks of gestation found the major causes to be preeclampsia (66.5%), infection (13.4%) and hyperglycaemia (3.1%). Other causes of hyponatraemia in pregnancy include acute fatty liver of pregnancy, postpartum haemorrhage, AKI, vomiting, water intoxication during labour, prolonged labour, caesarean section, cortisol deficiency (Sheehan's syndrome, hypophysitis, Addison's disease) and medications (oxytocin, non-steroidal anti-inflammatory medications, proton pump inhibitors).

Ascites and/or pleural effusion have been detected clinically and on ultrasound in 10% to 21% and 44% to 54%, respectively, of women with PET-SF.^4–7 Woods et al. ⁵ found that maternal ascites were associated with a sixfold increased risk of congestive cardiac failure and a ninefold increased risk of adult respiratory distress syndrome in women with HELLP syndrome. Guan et al. ⁸ found that 52.6% of women with PET-SF and ascites developed cardiopulmonary dysfunction 5.1 ± 2.6 days following the diagnosis of ascites. A prospective cohort study of Rwandan women admitted with PET-SF found the presence of ascites on ultrasound was associated with earlier delivery, higher stillbirth rates, lower birth weight, higher rates of neonatal intensive care admission, and prolonged maternal hospital stay. ⁴ A matched cohort study found the presence of ascites with PET-SF was associated with an increased risk of a composite adverse maternal outcome of eclampsia, pulmonary oedema, AKI or disseminated intravascular coagulation (adjusted OR 16.4 (95% CI 2.9–93.3)). ⁶ A retrospective study of the expectant management of early-onset PET-SF found that the serum albumin and presence of ascites/pleural effusions were factors affecting prolongation of pregnancy. ⁹ Several authors have concluded the development of ascites/pleural effusions in women with PET-SF may be an indication for immediate delivery.^9,10

The major factors in the development of ascites/pleural effusions in preeclampsia are thought to be increased capillary permeability and reduced capillary oncotic pressure due to hypoalbuminaemia. Microvascular permeability is significantly increased in preeclampsia related to elevated circulating levels of tumour necrosis factor-alpha. ¹¹ Vazquez-Rodriguez and Veloz-Martinez ⁷ reported a negative correlation of −0.43 between ascites and a measure of capillary permeability (Briones index) in 222 women with PET-SF and ascites. ¹²

Le et al. ⁹ reported mean plasma albumin levels of 24.08 ± 5.76 g/l (third-trimester reference interval 23–42) in preeclamptic women with hydrothorax and ascites compared with 28.19 ± 5.58 g/l in women with preeclampsia without effusions (p < 0.05). ⁹ Vazquez-Rodriguez and Veloz-Martinez ⁷ reported a weak negative correlation of −0.03 between measured plasma colloid osmotic pressure and the development of ascites. ¹²

Echocardiographic features in PET-SF include left ventricular diastolic dysfunction, increased left ventricular wall thickness, left atrial enlargement, elevated right ventricular systolic pressure and abnormal right ventricular strain. ¹³ PET-SF is associated with elevated levels of brain natriuretic peptide, with a strong correlation with echocardiographic measures of left ventricular systolic and diastolic dysfunction. ¹⁴ Calculated capillary hypostatic pressure, however, is significantly lower in PET-SF than in mild preeclampsia or healthy pregnancy, and where performed in six cases of PET-SF complicated by ascites/pleural effusions, echocardiography revealed normal left and right ventricular function. ¹⁵

Other causes of ascites and/or pleural effusions in pregnancy which need to be considered include infection, neoplasia, thromboembolic disease, cardiac failure, cirrhosis, non-cirrhotic portal hypertension, pancreatitis, mirror syndrome and hypoalbuminaemia.

Four case reports describe the abrupt onset of sinus bradycardia (pulse rate 35–50 bpm) in women with PET-SF, followed within hours by the development of HELLP syndrome or eclampsia.^16–19 None of the patients received β-blocker or alpha-methyldopa therapy, and betamethasone had been administered in only one case, this more than 48 h prior to the onset of bradycardia. A study comparing third-trimester heart rate in 15 women with preeclampsia with 44 normotensive pregnant women found a significantly lower mean heart rate in the women with preeclampsia (71 ± 14 vs. 85 ± 10 beats per minute; p < 0.005). ²⁰ Sinus bradycardia with PET-SF must be differentiated from bradycardia due to magnesium sulphate, heart block following a dural puncture, or antihypertensive therapy with negative chronotropic activity. Bradycardia has also been infrequently described in non-pregnant individuals following high dose glucocorticoid therapy, with a single case reporting the onset of bradycardia 26 h after administration of betamethasone to a pregnant woman. ²¹

The phenomenon of a pulse rate lower than expected for a patient's disease process has been postulated to relate to changes in the autonomic nervous or immune systems. Abnormalities of autonomic function have been demonstrated in 93.6% of women with preeclampsia. ²² A systematic review found, however, that hypertensive disorders of pregnancy were associated with sympathetic overdrive, and parasympathetic withdrawal, which might be expected to result in increased heart rate. ²³

Relative bradycardia, a pulse rate lower than expected for a given body temperature, may occur in several conditions associated with elevated levels of proinflammatory cytokines including acute viral hepatitis, coronavirus, dengue fever, intracellular Gram-negative pathogens, lymphoma, other causes of neoplastic fever and acute pancreatitis. Proinflammatory cytokine levels are significantly elevated in preeclampsia. ²⁴ The administration of interleukin-6 (IL-6)-induced significant bradycardia in animal studies. ²⁵

A possible pathophysiological mechanism linking the findings of hyponatraemia, ascites, pleural effusions and bradycardia in PET-SF is the elevation of proinflammatory cytokines in preeclampsia. ²⁴ IL-6 increases vagal tone and decreases heart rate variability, plays an important role in the non-osmotic regulation of anti-diuretic hormone, and promotes a sustained loss of endothelial barrier function, predisposing to capillary leak.

Preeclampsia is a disorder which may have protean manifestations, and PET-SF may occur in the absence of hypertension or proteinuria. While the ratio of soluble fms-like tyrosine kinase-1 to placental growth factor has high accuracy for differentiating preeclampsia from non-preeclampsia patients, its power for differentiating severe forms of the disease is considerably lower, and delivery decisions should not be made using the ratio test alone.^26,27

It is important for health professionals involved in the care of women with preeclampsia to be aware that severe hyponatraemia, ascites and pleural effusions, and abrupt severe bradycardia may signal severe disease and prompt consideration of delivery, depending upon gestation.