Gordon's syndrome in pregnancy

Abstract

Gordon's syndrome, also known as pseudohypoaldosteronism type II and familial hyperkalaemic hypertension, is a rare inherited condition characterised by familial hyperkalaemia, normal anion gap hyperchloraemic metabolic acidosis, low renin with normal glomerular filtration rate and hypertension. The outcome of 11 pregnancies in 3 women with Gordon's syndrome is presented and combined with 13 pregnancies in 7 women previously described. Pregnancy in women with Gordon's syndrome appears to be associated with a significant risk of adverse pregnancy outcomes, particularly where there is maternal hypertension preconception. No pregnancy registry exists for Gordon's syndrome. The available data is limited to case reports and small case series and may be affected by bias. A pregnancy registry would be valuable to assist in preconception counselling and management during pregnancy. The goal of this study was to summarise the available cases describing pregnancy outcomes with maternal Gordon's syndrome.

Introduction

Hyperkalaemia is uncommon in pregnancy. Pregnancy-specific causes include magnesium sulphate infusion, labetalol therapy and rebound hyperkalaemia following beta-agonist infusion for tocolysis. Familial hyperkalaemia has been described with Gordon's syndrome and pseudohyperkalaemia.

The prevalence of Gordon's syndrome is not known, though the condition may be underdiagnosed due to manifestations being mild, incomplete, intermittent or poorly ascertained. ¹ The characteristic features of hyperkalaemia and hypertension are variably present. Serum potassium levels fluctuate in Gordon's syndrome. Among 15 affected individuals never treated with thiazide diuretics, serum potassium was within the normal range in 8 of 36 samples (22%) tested. ² Additionally, the absence of hypertension does not exclude the diagnosis of Gordon's syndrome. The mean age of detection of hypertension in a large kindred of individuals with with-no-lysine kinase (WNK) gene mutations was 26 ± 7.5 years in males and 36.7 ± 12.5 years in females. ² A systematic review of 27 patients with Gordon's syndrome caused by kelch-like family member 3 (KLHL3) mutation diagnosed at a mean age of 28.2 years found that 74% of patients presented with hypertension, 77% with hyperkalaemia and 59% with metabolic acidosis. ³

Other variable features in Gordon's syndrome include hypercalciuria, short stature, serum aldosterone lower than expected for the degree of hyperkalaemia, dental anomalies, osteopenia, muscle weakness and rarely intellectual impairment (Figure 1).

Figure 1.

Features of Gordon's syndrome.

Gordon's syndrome is caused by the gain of function mutations in the genes regulating the thiazide-sensitive sodium-chloride transporter channel (NCC) located on the distal convoluted tubules of renal nephrons. ⁴ Excessive activation of the NCC results in increased sodium and chloride reabsorption in the distal nephron causing volume expansion, low renin and hypertension. ⁵ The reduction in luminal electronegativity lowers the forces for potassium and hydrogen ion secretion resulting in hyperkalaemia and metabolic acidosis.

Four genes have been implicated: WNK1 on chromosome 12, WNK4 on chromosome 17, cullin 3 (CUL3) on chromosome 2 and KLHL3 on chromosome 5. ⁶ Intronic deletion mutations of WNK1 and missense mutations in WNK4 result in a gain of function (increased WNK1 expression) that increases the NCC activity in the distal nephron. CUL3 and KLHL3 encode proteins involved in the degradation of the WNK1 and WNK4 proteins. The frequency of gene mutations in 86 families with Gordon's syndrome has been reported to be KLHL3 (48%), CUL3 (29%), WNK1 (9%), WNK4 (9%) and no pathogenic gene mutation was found in 5%. ⁷

Gordon syndrome is usually inherited as an autosomal dominant condition with high penetrance although autosomal recessive inheritance has been described with KLHL3. Many pathogenic de novo mutations in CUL3 have been reported. ⁸

The clinical and biochemical manifestations of Gordon's syndrome vary depending on the type of gene mutation. Individuals with a heterozygous CUL3 pathogenic variant tend to have more severe hyperkalaemia and metabolic acidosis, earlier and more severe hypertension, and a greater likelihood of growth impairment compared to those with KLHL3, WNK1, or WNK4 mutations. ⁹

Individuals with biallelic KLHL3 variants tend to have more severe clinical manifestations than those with heterozygous pathogenic variants in KLHL3. In general, clinical manifestations of Gordon's syndrome appear to be milder in individuals with a heterozygous WNK1 or WNK4 pathogenic variant compared to those with a heterozygous CUL3 or a heterozygous or biallelic KLHL3 pathogenic variant(s). ⁹

Patients with WNK4 mutations tend to have hypercalciuria whereas those with WNK1 mutations have normocalciuria. ²

Case series

Patient 1

A 31-year-old woman in her fourth pregnancy was referred to a medical antenatal clinic because of a history of unexplained hyperkalaemia with normal renal function. Her first pregnancy resulted in spontaneous labour at 32 weeks' gestation due to cervical dysfunction. Fetal Dopplers on ultrasound were normal. Her second child displayed ectrodactyly, missing two fingers on one hand. Her third child was noted to have short long bones on fetal ultrasound, this pregnancy also being complicated by postpartum preeclampsia. In her fourth pregnancy, the woman's serum potassium was 5 mmol/l (normal 3.2–4.3 mmol/l) with a hyperchloraemic normal anion gap metabolic acidosis (HCl-NAGMA), and serum creatinine was 51 μmol/l (30–77). Mass renin was 5 mU/l (pregnancy range 15 ± 2) and plasma aldosterone was 644 pmol/l (pregnancy range 1590 ± 150 pmol/l) at 24 weeks gestation. Further investigation revealed that her mother, maternal grandfather, and maternal great-uncle had hyperkalaemia without hypertension. The woman's serum potassium remained stable and blood pressure was normal during pregnancy. Fetal ultrasound at 28 weeks' gestation demonstrated femur length on the 1st centile with estimated fetal weight on the 23rd centile, abdominal circumference on the 28th centile and head circumference on the 59th centile by Hadlock IV formula parameters, and normal fetal Doppler measurements. At 32 weeks' gestation the woman presented with placental abruption and delivered an 1800 g male by emergency caesarean section (birthweight 50th percentile). ¹⁰ The newborn serum potassium ranged between 5.2 and 6.0 mmol/l (normal neonatal serum potassium 3.5–6.0 mmol/l) with HCl-NAGMA. The baby was not noted to have any genetic syndrome. The mother and baby were referred for testing for WNK1, WNK4, KLHL3 and CUL3 gene mutations, the results of which are pending.

Patient 2

A 29-year-old woman in her first pregnancy was referred to the obstetric medical clinic at 28 weeks' gestation because of hyperkalaemia. There was no other significant past history. Her father and sister were known to have hyperkalaemia and her father also had hypertension. The patient's serum potassium ranged between 4.6 and 6 mmol/l during her pregnancy with an HCl-NAGMA and normal renal function. At 30 weeks gestation mass renin was 5 mU/l and plasma aldosterone was 1420 pmol/l. Twenty-four hours of urine calcium was 13 mmol/day (non-pregnant reference range 1.2–7.5 mmol/day). Pseudohyperkalaemia was excluded. The pregnancy was uncomplicated, proceeding to vaginal delivery of a 3713 g female at term (birthweight 85th centile). Neonatal serum potassium ranged between 6.8 and 8 mmol with an HCl-NAGMA, mass renin of 17 mU/l (normal neonatal range 64 ± 16 mU/l), plasma aldosterone of 1383 pmol/l (normal 2260 ± 202 pmol/l) and hypercalciuria consistent with Gordon's syndrome. Maternal testing did not detect mutations in the WNK1 or WNK4 genes. Testing for CUL3 and KLHL3 gene mutations is pending.

Patient 2's second pregnancy was unremarkable other than the fetal ultrasound showing asymmetrical leg length. Neonatal serum potassium on day 5 of life was 5.9 mmol/l, with normal serum bicarbonate, chloride and renin.

Genetic testing on patient 2 and her 2 children is awaited.

Patient 3

Patient 3 is the 33-year-old sister of patient 2. Investigations when not pregnant revealed serum potassium ranging between 5.7 and 6.1 mmol/l, mass renin 3.7 mU/l (normal 3–40 mU/l), plasma aldosterone <69 pmol/l (normal 30–800 pmol/l), and an HCl-NAGMA. Her blood pressure was normal. Patient 3 has had 5 pregnancies. Her 4th pregnancy was complicated by delivery at 32 weeks gestation due to preterm premature rupture of membranes. The other 4 pregnancies were uncomplicated. Her third child had biochemical features consistent with Gordon's syndrome with a neonatal serum potassium of 6.7 mmol/l, mass renin of 5 mU/l and HCl-NAGMA.

A significant limitation of this small case series presented is the absence of genetic testing in the children of patient 2, and the families of patients 1 and 3.

Discussion

Previously 13 pregnancies in 7 women with Gordon's syndrome have been described in the literature^11–17 (Table 1). These were combined with the 11 pregnancies in the 3 women in this series and the outcomes reviewed.

Table 1.

Pregnancy outcomes with maternal Gordon's syndrome.

Patient pregnancy	Maternal age (years)	Maternal hypertension	Diuretic pre/during pregnancy	Pregnancy complications	Gestation birth	Birth weight (g)	Infant affected	Gene mutation
1.1	24	No	No/No	APH, Short cervix	32	1694	NS	NS
1.2	27	No	No/No	Ectrodactyly	40	3160	Yes	NS
1.3	29	No	No/No	PET; short long bones, bilateral hydronephrosis, echogenic bowel on fetal US	39	2805	NS	NS
1.4	31	No	No/No	Placental abruption	31.5	1800	Yes	NS
2.1	29	No	No/No	Nil	39	3997	Yes	NS
2.2	32	No	No/No	Asymmetry leg length	39	4281	No	NS
3.1	20	No	No/No	Nil	NS	NS	No	NS
3.2	22	No	No/No	Nil	39	NS	No	NS
3.3	25	No	No/No	Nil	NS	3790	Yes	NS
3.4	29	No	No/No	PPROM	32	NS	No	NS
3.5	30	No	No/No	Nil	NS	NS	No	NS
4.1 11	19	Yes	Yes/Yes	PET	32	1100	Yes	NS
5.1 12	33	Yes	Yes/ Yes	PET, FGR	35.6	1430	Yes	NS
5.2 12	34	Yes	Yes/Yes	FGR	37.6	1775	Yes	NS
6.1 13	28	Yes	Yes/No	Hypertension at term	40	3000	NS	WNK4
6.2 13	29	Yes	Yes/No	Nil	38	2720	NS	WNK4
7.1 13	36	Yes	Yes/ No	Hypertension at term	40	NS	NS	WNK4
7.2 13	40	Yes	Yes/No	Hypertension at term	40	NS	NS	WNK4
8.1 14	28	Yes	Yes/Yes	PET	34	1890	Yes	NS
8.2 14	NS	Yes	Yes/Yes	PET, FGR	Preterm	1100	Yes	NS
9.1 15	30	No	No/No	FGR	37.2	2244	No	NS
10.1 16	28	No	Yes/No	Nil	Term	NS	No	KLHL3
11.1 17	33	Yes	No/No	Eclampsia, fetal death	NS	NS	NS	NS
11.2 17	35	Yes	Yes/Yes	Nil	Term	NS	NS	NS

APH: antepartum haemorrhage; NS: not stated; PET: preeclampsia; FGR: fetal growth restriction; WNK4: with-no-lysine kinase 4; KLHL3: kelch-like family member 3; US: ultrasound; PPROM: preterm premature rupture of membranes.

Overall, 11 pregnancies occurred in 5 women who had hypertension preconception. Pregnancy complications were eclampsia and fetal death in one pregnancy, preeclampsia (4 cases), fetal growth restriction (FGR) (3), gestational hypertension at term (3), and delivery prior to 37 weeks gestation in 4 pregnancies.

Thirteen pregnancies occurred in 5 women with normal blood pressure preconception. These pregnancies were complicated by preeclampsia (1 case), FGR (1), placental abruption (1), delivery prior to 37 weeks (3), fetal ectrodactyly (1), fetal echogenic bowel and short long bones (1).

Eleven pregnancies were to women taking thiazide diuretics preconception. Thiazides were continued during 6 pregnancies. These 6 pregnancies were complicated by preeclampsia (4 cases) and FGR (4). No congenital anomalies occurred.

The results of genetic testing were available for only 4 women. One woman had WNK4 and another KLHL3 gene mutations. Two sisters tested negative for WNK1 and WNK4; KLHL3 and CUL3 testing were not performed. Pregnancies in these women were uncomplicated other than one case of preterm prelabour rupture of membranes and three cases of hypertension at term.

In patients 1, 2 and 3 serum potassium was stable during pregnancy when compared with non-pregnant values. Awad et al. ¹⁶ reported the resolution of hyperkalaemia during pregnancy in a woman with a KLHL3 mutation despite cessation of preconception thiazide diuretics. Mayan et al. ¹³ described two women with WNK4 mutations with hypertension and hyperkalaemia in whom blood pressure remained normal but hyperkalaemia returned during pregnancy when thiazides were ceased.

The risk of hypertensive complications of pregnancy and FGR appears to be higher in women where Gordon's syndrome has been complicated by the development of hypertension prior to conception than in normotensive women. The mechanism of other pregnancy and fetal complications is unclear. FGR, preterm labour and dysmorphic features are associated with distal renal tubular acidosis (RTA) due to toluene use, but not with RTA secondary to connective tissue disease. ¹⁸ Rowe et al. ¹⁹ described 3 pregnancies in 2 women with distal RTA who developed diastolic hypertension in the third trimester, attributed to the physiological hypervolaemia of pregnancy superimposed upon the renal condition. Maternal metabolic acidosis may impair fetal bone growth and development. ²⁰ Single nucleotide polymorphisms in WNK1 have been shown to be associated with an increased risk for preeclampsia. ²¹ Levels of CUL3 and its neddylation-modified factor are decreased in spiral arteries and placentas of women with preeclampsia. ²² Impaired activity of CUL3 in vascular smooth muscle causes abnormal vasoconstriction.

Brachydactyly or ectrodactyly has not been previously described with Gordon's syndrome. Hypertension with brachydactyly syndrome is a rare autosomal dominant disorder associated with severe salt-independent, age-dependent hypertension, type E brachydactyly, short stature and hypokalaemia due to mutation in phosphodiesterase 3A. ²³ The disorder is localised to 12p12.2-p11.2.

Salt restriction and/or thiazide diuretic therapy may result in the complete resolution of hypertension, hyperkalaemia, acid–base disturbance, and low renin in Gordon's syndrome.^24,25 Low dose thiazide diuretics inhibit the activity of NCC in the distal nephrons. Theoretical concerns regarding the possibility of maternal hypovolaemia have been raised regarding the use of diuretics in pregnancy. A Cochrane review examining the possible role of diuretics in the prevention of preeclampsia found no increased risk of stillbirth, neonatal death, preterm birth, FGR, low Agar scores or neonatal thrombocytopenia compared with placebo. ²⁶ There is a lack of data regarding the possible teratogenic effects of thiazide diuretics. A retrospective study found an increased risk of various malformations in the offspring of women who purchased prescription diuretics during early pregnancy, however, this may have been confounded by high rates of diabetes mellitus, as well as the increased risk of malformations in women with obesity, hypertension and undiagnosed diabetes mellitus. ²⁷ Nevertheless it would seem prudent to avoid the use of thiazide diuretics during the first trimester.

The biochemical diagnosis of Gordon's syndrome may be affected by physiological changes in pregnancy. Compared with preconception values serum potassium falls by 0.2 to 0.3 mmol/l, plasma aldosterone and renin levels rise 4- to 8-fold, urine calcium increases by up to 300%, and serum bicarbonate fall due to metabolic compensation for the physiological respiratory alkalosis of pregnancy. In the two previous case reports of Gordon's syndrome where plasma renin activity was measured during pregnancy, the results were below the lower end of the normal range for non-pregnant individuals.^13,16 Similarly, in patients 1 and 2, mass renin was at the lower end of the normal range for non-pregnant individuals.

Conclusion

Gordon's syndrome appears to be associated with a high risk of pregnancy complications, especially in women with preconception hypertension. This conclusion, however, is based upon case studies and small case series, and do not take into account other comorbidities for adverse pregnancy outcome such as age or obesity, the presence or absence of preconception hypertension, and the possibility of missed diagnosis of the disorder. An international pregnancy registry would be valuable to better assess pregnancy outcomes with Gordon's syndrome. Future research directions may include assessing whether known maternal mutations or fetal mutations predict adverse pregnancy outcomes. It is unclear as to whether therapy with thiazide diuretics improves maternal and fetal outcomes in pregnancy complicated by Gordon's syndrome.