Mid-gestational maternal cardiovascular profile in preterm and term pre-eclampsia: a prospective study

Authors

  • K Melchiorre,

    1. Fetal Maternal Medicine Unit, Department of Obstetrics and Gynaecology, St George's Hospital, University of London, London, UK
    2. Department of Obstetrics and Gynaecology, University of Chieti, Chieti, Italy
    Search for more papers by this author
  • G Sutherland,

    1. Department of Cardiology and Cardiothoracic Surgery, St George's Hospital, University of London, London, UK
    Search for more papers by this author
  • R Sharma,

    1. Department of Cardiology and Cardiothoracic Surgery, St George's Hospital, University of London, London, UK
    Search for more papers by this author
  • M Nanni,

    1. Fetal Maternal Medicine Unit, Department of Obstetrics and Gynaecology, St George's Hospital, University of London, London, UK
    Search for more papers by this author
  • B Thilaganathan

    Corresponding author
    • Fetal Maternal Medicine Unit, Department of Obstetrics and Gynaecology, St George's Hospital, University of London, London, UK
    Search for more papers by this author

Correspondence: Basky Thilaganathan, Fetal–Maternal Medicine Unit, St George's University of London, London SW17 0RE, UK. Email basky@pobox.com

Abstract

Objective

Pre-eclampsia (PE) is associated with maternal cardiac remodelling and biventricular diastolic dysfunction. Preterm PE alone can also be associated with severe left ventricular hypertrophy and biventricular systolic dysfunction. The aim of this study was to assess whether the maternal cardiovascular profile at mid-gestation in nulliparous normotensive women differs in women destined to develop preterm PE versus those who will develop PE at term.

Design

Prospective study.

Setting

Tertiary referral university centre.

Population

A total of 269 women, including 152 at increased risk of developing PE as determined by mid-gestational uterine artery Doppler assessment.

Methods

Women underwent blood pressure profiling, echocardio-graphy, cardiac tissue Doppler and strain rate analysis at 20–23 weeks of gestation.

Main outcome measures

Mid-gestational cardiovascular profile in women with normal pregnancy and those that subsequently developed preterm or term PE.

Results

Pre-eclampsia subsequently developed in 46 women, including 18 with preterm PE. Women who subsequently developed PE, irrespective of gestation, had evidence of left ventricular concentric remodelling (33%) which was not found in the control women (P < 0.0001). Only women who developed preterm PE exhibited a high resistance-low volume haemodynamic state at mid-gestation. The latter group also had evidence of left ventricular diastolic or systolic dysfunction (33%) and segmental impaired myocardial relaxation (72%).

Conclusions

Asymptomatic cardiac diastolic dysfunction is evident at mid-gestation in women who subsequently develop preterm PE but not in those who develop term PE. These cardiac findings are useful in understanding the pathophysiology of PE and corroborate the concept that PE is not a single disorder, but a cluster of symptoms that have several different aetiologies.

Introduction

Pre-eclampsia (PE) and fetal growth restriction are known to be associated with altered cardiac geometry, impaired myocardial relaxation and biventricular diastolic dysfunction.[1-11] Preterm but not term PE, is also associated with severe left ventricular hypertrophy and biventricular systolic dysfunction in a significant proportion of women.[1-10] Postpartum follow up of women whose pregnancies were complicated by PE has shown persistence of altered cardiac geometry and left ventricular diastolic dysfunction in those who had preterm PE, but not for women who developed PE at term.[1-3, 5, 7, 10] The long-term cardiovascular morbidity of women who developed PE in pregnancy is well characterised and is worse for those who developed preterm PE.[12, 13] These findings suggest that maternal cardiovascular susceptibility may play an important role in the pathogenesis of preterm more than term PE.

The aim of this study was to assess whether the maternal cardiovascular profile at mid-gestation in nulliparous normotensive women differs in women destined to develop preterm PE versus PE at term.

Methods

This was a prospective observational case–control study carried out over a 3-year period from January 2008. The study included only nulliparous women with singleton pregnancy and at increased risk of developing PE. Risk for developing PE was determined by uterine artery Doppler assessment at the routine ultrasound assessment at 20–23 weeks of gestation. Women with a mean uterine artery pulsatility index (PI) greater than the 95th centile (screen-positive women) were asked to take part in the study.[14] Nulliparous women with normal uterine artery Doppler PI were recruited consecutively during the same period and matched for gestational age at assessment. Women with medical comorbidities, smokers, on medication or with fetal abnormalities were excluded from recruitment to the study. At the time of this study, women were not routinely prescribed aspirin in pregnancy. All women provided informed consent and the study was approved by the local institutional review committee.

Patients were categorised accordingly to mid-gestation uterine artery Doppler screening test and pregnancy outcome in one of the following mutually exclusive groups: screen negative (normal uterine artery PI) women with uncomplicated pregnancy delivering at term (controls), screen positive (high uterine artery PI) women with uneventful pregnancy, screen positive women who required delivery before 37 weeks because of the severity of PE (preterm PE) and screen positive women who developed PE at or after 37 weeks (term PE). All other categories were excluded from the analysis. PE was classified according to the International Society for the Study of Hypertension in Pregnancy guidelines.[15] In brief, women recruited into the study were normotensive and nonproteinuric at mid-gestation with complete recovery within 12 weeks postpartum. Women with postpartum persistent hypertension or proteinuria (essential hypertension or nephropathy) were excluded from the analysis.

The study assessments included weight, height, blood pressure profile, 12-lead electrocardiogram and echocardiography. Maternal blood pressure was measured manually from the brachial artery using a mercury sphygmomanometer according to the guidelines of the National High Blood Pressure Education Programme Working Group on High Blood Pressure in Pregnancy.[16] All women were studied by standard two-dimensional and Doppler transthoracic echocardiography, tissue Doppler, strain and strain rate analysis as described in detail previously[1-3, 17, 18] and outlined in brief below. Echocardiography and offline tissue Doppler and strain rate analyses were performed within 1 week of the mid-gestational uterine Doppler assessment by one investigator blinded for pregnancy outcome (KM).

Left-sided cardiovascular system assessment

Left ventricular (LV) global diastolic dysfunction was defined as the inability of the heart to fill to a normal volume without an increase in chamber filling pressures. The LV diastolic function and left heart chamber filling pressures were assessed and graded using standard diagnostic algorithms with the recommended adjustments reflecting the concomitant systolic function and maternal age[19] and further adjustments reflecting the pregnant state[2] (see Supporting Information, Figure S1). Normal LV geometry was defined as normal LV mass index (LVMI < 95 g/m2) and normal relative wall thickness (RWT < 0.42).[20] Altered cardiac geometry was defined according to the following three mutually exclusive categories[20]: concentric remodelling—normal LVMI with increased RWT (>0.42); eccentric hypertrophy—increased LVMI (>95 g/m2) with normal RWT; concentric hypertrophy—increased LVMI (>95g/m2) and RWT (>0.42). LV radial systolic dysfunction was defined as ejection fraction less than 55%.[20] Left ventricular longitudinal systolic dysfunction was defined as average peak systolic velocity at the level of the left and septal sites of mitral valve annulus index two standard deviations (SDs) below the expected mean for age; LV global systo-diastolic dysfunction was defined as LV global diastolic dysfunction in the presence of ejection fraction less than 55%.[20] The severity of LV systolic dysfunction and remodelling/hypertrophy was graded according to the American Society and European Association of Echocardiography guidelines[20] (see also Supporting Information, Table S1).

Strain and strain rate assessment

Myocardial function was assessed using tissue colour Doppler-derived strain and strain rate technique. Peak systolic strain rate was considered abnormal if it was two SDs below the expected mean for age. If at least one myocardial segment was affected, it was termed impaired myocardial contractility. Regional diastolic dysfunction was defined as early to late strain rate ratio below or equal to one. If at least one myocardial segment was affected, it was termed impaired myocardial relaxation.

Correction of indices

All conventional echocardiographic indices were adjusted for body surface area and all tissue Doppler velocity indices were adjusted to the end-diastolic LV long axis length.

Statistical analysis

Data were analysed using SPSS 15 software (SPSS, Chicago, IL, USA). Variables were compared using Mann–Whitney U test or chi-square test, as appropriate. Paired group comparisons were only undertaken if Kruskal–Wallis testing indicated significant differences. A value of P < 0.05 was considered statistically significant and all tests were two-sided. Repeatability and reproducibility of conventional echocardiographic and tissue Doppler and strain rate indices was previously reported as very good.[2]

Sample size

Sample size was calculated for the primary outcome, cardiac diastolic function, using a power formula with an alpha level of 0.05 and a power of 85%. A sample size of 19 in each group was calculated to observe a >25% difference in tissue Doppler indices between cohorts.[21] The positive predictive value of uterine artery Doppler was used to calculate the number of high-risk women required to obtain an adequate sample size in each PE cohort.[22, 23]

Results

A total of 269 women were recruited at mid-gestation, of whom 152 with uterine artery Doppler PI above the 95th centile (high-risk women; see Supporting Information, Figure S2). Thirty-six (23.7%) of these women also had bilateral notching compared with five (4.2%; P = 0.0001) in the control group. Three women (1%) were excluded from the analysis because a full echocardiographic assessment could not be performed. Pre-eclampsia subsequently developed in 46 women from this cohort (18 preterm, 28 at term). There were no significant differences in the demographic characteristics of the controls and women who subsequently developed PE at term (Table 1). Women who developed preterm PE were significantly heavier than those with normal pregnancy outcome (control and high-risk uneventful groups) and more likely to be of Afro-Caribbean ethnicity compared with the other groups (Table 1).

Table 1. Demographic and pregnancy characteristics in the cohorts of women with uneventful pregnancy (controls) and pregnancy complicated by preterm and term PE
ParameterLow-risk women with uneventful outcome (n = 105)High-risk women with uneventful outcome (n = 63)High-risk women who developed term PE (n = 28)High-risk women who developed preterm PE (n = 18)
  1. BMI, body mass index; BSA, body surface area.

  2. P values for gestational age at delivery and birth weight centiles are not reported as they were determined to be different by the study design.

  3. Values are given as median (interquartile range) or number of women (percentage).

  4. a

    P < 0.05 versus ‘low-risk women with uneventful outcome’.

  5. b

    P < 0.05 versus ‘high-risk women with uneventful outcome’.

  6. c

    P < 0.05 versus ‘high-risk women who developed term PE’.

Maternal age (years)31.5 (26–34)31.8 (26–35)32 (30–37)30 (24–34)c
Caucasian73 (69.5)47 (74.6)22 (78.6)7 (38.9)a,b,c
Afro-Caribbean18 (17.1)10 (15.8)3 (10.7)8 (44.4)a,b,c
Asian14 (13.3)6 (9.5)3 (10.7)3 (17)
Prepregnancy BMI (kg/m2)22.5 (20.8–25.8)22.6 (20.8–25.0)23.2 (21.6–27.3)25.8 (22–31.5)a,b
BMI (kg/m2) at assessment23.9 (21.7–27.2)23.9 (21.5–26.9)25.3 (22.3–28.4)26 (23–32)
BSA (m2) at assessment1.75 (1.66–1.83)1.72 (1.60–1.79)1.71 (1.53–1.72)1.79 (1.65–1.87)
Gestational age at delivery (weeks)40.3 (39.2–41.1)40.1 (38.5–40.7)39.3 (38.6–40)31.7 (28.7–34.4)
Birthweight (centile)36 (20–64)23.8 (16–54)17 (12–32)4.7 (2–7)

Haemodynamic parameters

Women who subsequently developed term and preterm PE had significantly higher mean arterial pressure and total vascular resistance index (TVRI) at mid-gestation than those with normal pregnancy outcomes (Table 2 and Fig. 1). Women who subsequently developed preterm PE but not those who developed term PE had lower stroke volume index and cardiac index (CI) than controls (Table 2).

Table 2. Mid-gestation haemodynamic and LV geometric and systolic functional parameters in women with uneventful pregnancy (controls) versus women who subsequently developed preterm and term pre-eclampsia (PE)
 Low-risk women with uneventful outcome (n = 105)High-risk women with uneventful outcome (n = 63)High-risk women who developed term PE (n = 28)High-risk women who developed preterm PE (n = 18)
  1. Values are given as median (interquartile range) or number of women (percentage).

  2. HR, heart rate; MAP, mean arterial pressure; SVI, stroke volume index; CI, cardiac index; CO, cardiac output; TVRI, total vascular resistance index; RWT, relative wall thickness; LVM, left ventricular mass; LVMI, left ventricular mass index: LVM normalised for body surface area (BSA); Sm, peak systolic myocardial velocity at the level of the basal segment of the interventricular septum and anterolateral wall.

  3. a

    P < 0.05 versus ‘low-risk women with uneventful outcome’.

  4. b

    P < 0.05 versus ‘high-risk women with uneventful outcome’.

  5. c

    P < 0.05 versus ‘high-risk women who developed term PE’.

  6. d

    P < 0.05 versus ‘high-risk women who developed preterm PE’.

Haemodynamic
HR (beats/min)81 (75–87)79 (71–85)c85 (72–89)81 (74–92)
MAP (mmHg)80 (73–83)78 (73–82)83 (80–93)a, b90 (81–101)a,b,c
SVI (ml/m2)40 (34–47)39 (33–45)35 (32–44)38 (27–39)a
SV (ml)72 (60–86)67 (56–80)63 (46–76)a67 (57–70)a
CI (l/min/m2)3.2 (2.6–3.9)3.0 (2.6–3.5)3.2 (2.8–3.5)2.6 (2.3–2.8)a,b,c
CO (l/min)5.6 (4.7–6.8)5.1 (4.5–6.1)5.2 (4.4–6.4)4.7 (4.3–5.7)a
TVRI ((dynes/s/cm5) × m2)1993 (1696–2334)2045 (1744–2378)2138 (1995–2469)a3090 (2351–3376)a,b,c
TVR (dynes/s/cm5)1067 (883–1354)1248 (1016–1411)1440 (1048–1690)a1570 (1280–1652)a
LV geometric indices
RWT0.32 (0.28–0.36)0.32 (0.30–0.36)0.37 (0.31–0.46)a,b0.44 (0.35–0.49)a,b,c
LVMI (g/m2)62 (53–71)64 (56–71)65 (57–71)62 (57–71)
Wall stress index (dyne/cm2/m2)39 (29–49)37 (28–44)39 (27–49)36 (32–52)
Radial systolic function
Ejection fraction (%)62 (55–67)68 (62–70)a,c,d60 (51–69)64 (48–67)
Longitudinal systolic function
Sm basal septal7.1 (6.3–7.5)7.1 (6.3–7.4)7.0 (6.1–7.9)6.4 (5.5–7)
Sm basal lateral7.2 (6.0–9.0)8.0 (7.0–9.0)7.0 (6.0–8.0)5.9 (5.5–8.6)a,b,c
Figure 1.

Mid-gestational TVRI and RWT in the study population. Category: 1 = normal uterine artery Doppler with normal outcome group (green dots); 2 = high uterine artery Doppler with normal outcome group (blue dots); 3 = high uterine artery Doppler developing term pre-eclampsia group (purple dots); 4 = high uterine artery Doppler developing preterm pre-eclampsia group (red dots).

Left ventricular geometry

Women who subsequently developed term and preterm PE had higher LV RWT and unchanged LV mass and wall stress indices at mid-gestation compared with women with normal pregnancy outcomes (Table 2). The prevalence of altered geometry was significantly higher in both PE groups at mid-gestation compared with women with normal pregnancy outcomes (preterm PE: 55.6%, term PE: 32.1%, controls: 3.8%; high-risk uneventful group: 3.1%; = 0.0001; Table 3 and Fig. 1). LV concentric remodelling or hypertrophy was asymmetric with basal septal bulging seen in only three women who subsequently developed preterm PE (17%) but not in any of the other women (P = 0.01).

Table 3. Mid-gestation left-sided cardiovascular system in women with uneventful pregnancy (controls) versus women who subsequently developed preterm and term PE
 Low-risk women with uneventful outcome (n = 105)High-risk women with uneventful outcome (n = 63)High-risk women who developed term PE (n = 28)High-risk women who developed preterm PE (n = 18)
  1. Values are given as number of women (percentage) with abnormal findings.

  2. a

    P < 0.05 versus ‘low-risk women with uneventful outcome’.

  3. b

    P < 0.05 versus ‘high-risk women with uneventful outcome’.

  4. c

    P < 0.05 versus ‘high-risk women who developed term PE’.

Altered geometry
Concentric remodelling008 (28.6)a,b7 (38.9)a,b
Eccentric hypertrophy4 (3.8)2 (3.1)1 (3.4)1 (5.6)
Concentric hypertrophy0002 (11.1)
Total altered geometry4 (3.8)2 (3.1)9 (32.1)a,b10 (55.6)a,b
Myocardial function
Impaired relaxation22 (21)10 (15.8)6 (21.4)13 (72)a,b,c
Impaired contractility5 (4.7)3 (4.7)3 (10.7)4 (22.2)
Chamber function
Diastolic dysfunction2 (1.9)2 (3.1)3 (10.7)6 (33.3)a,b
Longitudinal systolic dysfunction0003 (16.7)a,b

LV diastolic function

Six of 18 women (33%) who went on to develop preterm PE demonstrated LV global diastolic dysfunction at mid-gestation compared with two in the control and two in the high-risk uneventful groups (1.9% and 3.1%, respectively; P = 0.0001; Tables 3 and 4). Thirteen of 18 women (72%) who subsequently developed preterm PE exhibited impaired myocardial relaxation compared with 21% of controls and 20% of high-risk uneventful women (P = 0.0001; Table 3, Fig. 2 and see Supporting Information, Table S2). The prevalence of LV global diastolic dysfunction and impaired myocardial relaxation in the group of women who later developed term PE was not significantly different compared with the control and high-risk uneventful groups (10.7% versus 1.9% and 3.1%; 21.4% versus 21.0% and 20%, respectively; P > 0.05; Tables 3–5 and see Supporting Information, Table S2).

Table 4. Mid-gestation diastolic indices of controls and women who developed preterm and term PE
 Low-risk women with uneventful outcome (n = 105)High-risk women with uneventful outcome (n = 63)High-risk women who developed term PE (n = 28)High-risk women who developed preterm PE (n = 18)
  1. Values are given as either median (interquartile range) or number of women (percentage).

  2. E, peak early diastole transmitral wave velocity; A, peak late diastole transmitral wave velocity; DT, deceleration time of E wave; IVRT, isovolumetric relaxation time; S, peak systolic pulmonary venous flow velocity; D, peak anterograde early diastolic pulmonary venous flow velocity; AR, peak retrograde late diastolic pulmonary venous flow velocity; ARdur, AR duration; (ARdur–Adur), the time difference between pulmonary AR-wave duration and mitral A-wave duration; E1, peak early diastolic velocity at mitral valve annulus; Em, peak early diastolic myocardial velocity at the level of the basal myocardium; Am, peak late diastolic myocardial velocity at the level of the basal myocardium; average E/E1 ratio, E to average of lateral and septal E1 velocities.

  3. a

    P < 0.05 versus ‘low risk women with uneventful outcome’.

  4. b

    P < 0.05 versus ‘high risk women with uneventful outcome’.

  5. c

    P < 0.05 versus ‘high risk women who developed term PE’.

Left atrium
Left atrial volume index (ml/m2)35 (28–40)34 (26–39)33 (22–40)35 (23–42)
Mitral inflow
E/A ratio1.8 (1.6–2.2)1.6 (1.3–1.9)1.6 (1.3–2.2)1.4 (0.9–1.6)a
DT (ms)116 (110–170)125 (123–177)140 (134–188)a171 (153–227)a,c,b
IVRT (ms)74 (62–77)76 (74–80)88 (79–95)a83 (71–89)
Pulmonary venous flow
S/D ratio1 (0.8–1.2)1 (0.7–1.2)0.87 (0.73–1.1)1.1 (0.85–1.3)a,c
AR cm/s0.23 (0.21–0.26)0.25 (0.21–0.27)0.26 (0.23–0.36)0.27 (0.25–0.35)a,c
AR dur–A dur−15 (–19–12)−16 (–19–13)−3 (–8–26)13 (–8–41)a,c,b
Tissue Doppler indices
Septal E1 (cm/s)12 (11–15)12 (11–16)11 (10–12)10 (6–14)a,c,b
Lateral E1 (cm/s)18 (16–19)18 (16–20)14 (8–17)12 (11–17)a,c,b
Septal Em/Am2.7 (1.4–3.1)2.0 (1.7–2.8)2.1 (1.6–2.4)1.2 (0.9–2.2)a,c,b
Lateral Em/Am2.9 (2.1–3.8)3.2 (2.2–3.7)3.1 (1.8–3.9)2.2 (1.3–3.9)
Left chamber filling pressure
Average E/E14.8 (4.2–5.7)4.6 (3.9–5.8)6.2 (5.3–6.7)6.5 (5.3–7)
Table 5. Summary of mid-gestational cardiac changes in women whose pregnancies were complicated by either term or preterm PE
Cardiac parameterTerm pre-eclampsiaPreterm pre-eclampsia
  1. MAP, mean arterial pressure; TVRI, total vascular resistance index; SVI, stroke volume index; CI, cardiac index; RWT, relative wall thickness.

Haemodynamic⇑MAP ⇑TVRI⇑MAP, ⇑TVRI,⇓SVI, ⇓CI
Remodelling⇑RWT, ⇑altered geometry⇑RWT, ⇑altered geometry, ⇑hypertrophy
Myocardial function⇑Impaired relaxation
Chamber (ventricular) function⇑Diastolic and systolic dysfunction
Figure 2.

Mid-gestational early to late diastolic myocardial velocity ratio at the level of the septal site of the mitral annulus (Em/Am: y-axis) and early to late diastolic strain rate ratio at the level of the basal septum (SRE/SRA: x-axis) in the study population. Category: 1 = normal uterine artery Doppler with normal outcome group (blue dots); 2 = high uterine artery Doppler with normal outcome group (green dots); 3 = high uterine artery Doppler developing term pre-eclampsia group (purple dots); 4 = high uterine artery Doppler developing preterm pre-eclampsia group (red dots).

LV systolic function

Only women who subsequently developed preterm PE exhibited LV longitudinal systolic dysfunction at the level of the lateral ventricle free wall (n = 3, 17%) compared with all other groups (P = 0.006; Tables 2 and 3 and see Supporting Information, Table S1). Both LV radial systolic function and myocardial contractility were preserved in all cases and controls and was even enhanced in the high-risk uneventful groups, as demonstrated by the increased ejection fraction in this group compared with control women (Table 2).

Discussion

Women requiring iatrogenic delivery before 37 weeks of gestation for worsening PE exhibit more severe cardiovascular impairment at mid-gestation compared with those who are delivered at term with PE or with an uneventful pregnancy.

Haemodynamic parameters

The women who went on to develop preterm PE demonstrated significantly higher TVRIs and lower CIs at mid-gestation than both control women and women who subsequently developed term PE. This finding is consistent with that of Valensise et al.[9] assessing maternal cardiac function at mid-gestation in asymptomatic women who later developed early-onset PE. High TVRI/low CI has also been demonstrated in the clinical phase of preterm PE by Visser and Wallenburg[24] who performed a well-designed, prospective study looking at haemodynamic indices assessed by invasive monitoring. These authors demonstrated that untreated women with PE presented with high TVRI/low CI whereas women with PE who were treated with antihypertensives had a similar CI to normotensive controls.[24] The authors concluded that the reported conflicting haemodynamic profiles in PE are artefacts caused by clinical management.[24] This haemodynamic model has been confirmed by several studies on the cardiovascular state in women with untreated preterm PE.[3, 6, 8] This high-impedance/low-volume haemodynamic state seen in women destined to develop preterm PE suggests that there is an increased LV afterload and contracted circulating volume even at mid-gestation.

In contrast, women who subsequently developed PE at term presented with mid-gestational high TVRI/unchanged CI haemodynamic state compared with control women. This finding is consistent with our previous investigations[3] and those of Hibbard et al.[25] on women with acute PE. This finding is in contrast with those of Bosio et al.,[26] Easterling et al.[27] and Valensise et al.,[9] who found a low TVR/high cardiac output state in the preclinical phase of term PE. It should be noted that in the studies that demonstrated discordant findings[9, 26, 27] cardiac indices were not corrected for maternal body surface area even though women with PE presented with significant anthropometric differences compared with controls.[28] More relevant to the discrepant cardiovascular profiles is the disease heterogeneity of PE, as demonstrated by Mei et al.,[29] who found seven different haemodynamic models in acute PE.

LV geometry

There is a significantly higher prevalence of LV remodelling/hypertrophy at mid-gestation in both preterm PE and term PE women. This finding is likely to represent a compensatory response to the increased afterload that is evident from the higher mid-gestational mean arterial pressures seen in women with PE. Left ventricular remodelling is required to minimise wall stress in the presence of increased afterload as a recognised mechanism for preserving the balance between myocardial oxygen demand and supply. In women destined to develop PE it seems to be an effective response, as wall stress indices remain unaltered between cohorts. Our results are in agreement with those of previous authors who have similarly demonstrated compensatory altered LV geometry in the preclinical phase of both preterm and term PE.[9] In particular, Valensise et al.[30] demonstrated altered mid-gestational cardiac geometry in women who subsequently developed fetal growth restriction and gestational hypertension. In another study, the authors demonstrated that third-trimester concentric remodelling in women affected by early mild gestational hypertension was an independent predictor for the development of PE.[10] They also showed that LVMI and RWT were significantly higher at 24 weeks of gestation in women who subsequently developed PE compared with control women.[9]

Left-sided cardiac chamber function

The combined use of tissue Doppler and conventional echocardiography taking into account maternal haemodynamic and heart geometry to assess cardiac function demonstrated a very high prevalence (33%) of LV diastolic dysfunction at mid-gestation in women who subsequently developed preterm PE compared with the women with term PE or uneventful pregnancy. There was no significant difference in mid-gestational LV diastolic function between women with term PE and control women. This is a novel finding as no previous study has systematically assessed diastolic function in the preclinical stages of PE. Diastolic dysfunction is related to increased afterload and LV stiffness, as demonstrated by significantly higher mean arterial pressure, TVRI, RWT and LV concentric hypertrophy noted in women destined to develop preterm PE. Preterm PE, but not term PE, also exhibited longitudinal systolic dysfunction at the level of the lateral LV free wall with preserved radial function. This pattern of systolic impairment affecting only the longitudinal function is similar to that seen in early essential hypertension in nonpregnant women and is indicative of afterload-induced impairment of subendocardial myocardial fibres.[18] Our findings of significant LV diastolic dysfunction in preterm PE suggest that there is exhaustion of cardiac reserve by mid-gestation in these women. This may favour the subsequent deterioration of cardiac function seen with the diagnosis of preterm PE and the higher long-term cardiac morbidity/mortality seen in preterm versus term PE and controls. The only other study that assessed cardiac function in the preclinical phase of the disease in early and late PE was performed by Valensise et al.[9]

However, the latter study was limited by the interpretation of diastolic function indices in isolation, which is inadequate for detecting the grade 2 of diastolic dysfunction (pseudonormal pattern) were mitral inflow indices are normal despite the presence of advanced dysfunction. Furthermore, the authors did not account for the age-dependency of diastolic indices nor did they use recognised clinical algorithms for the diagnosis and grading of diastolic dysfunction that take into account the actual maternal haemodynamic and heart remodelling and are hence more reliable than the use of cardiac indices in isolation. These limitations are entirely understandable given the less extensive understanding of diastolic function and inexperience with the use of tissue Doppler imaging at the time that the study of Valensise et al.[9] was performed.

Myocardial function

There is a paucity of published data on intrinsic myocardial function using myocardial strain and strain rate techniques in pregnancy despite their extensive validation and correlation with invasive indices of intrinsic cardiac function.[31-35] Myocardial strain is a measure of the magnitude of tissue deformation and strain rate is the rate at which said deformation occurs.[31] Myocardial deformation is an active energy-dependent process both in systole and diastole and therefore highly sensitive to increased wall stress, afterload and ischaemia.[31] Deformation indices have already been used to assess myocardial function in a wide range of cardiac conditions and found to give fundamental information on myocardial properties beyond that afforded by conventional echocardiography and tissue Doppler velocity indices.[31-35] The relevance of strain/strain rate assessment is that it is able to evaluate impaired myocardial contractility/relaxation before the development of overt chamber systo-diastolic dysfunction.[31-35] Our study findings demonstrate impairment of myocardial deformation in the absence of overt chamber dysfunction in a significant proportion of women with preterm PE at 22 weeks of gestation. Impairment of myocardial diastolic function is most evident at the basal segment of the inferoseptal wall in agreement with early hypertension in nonpregnant women.[18] In contrast, intrinsic myocardial contractility was preserved in all cases and controls.

Prehypertension state

Women who subsequently developed preterm PE had higher mean arterial pressure than women who developed term PE and control women at mid-gestation. The standard definition of hypertension as blood pressure more than or equal to 140/90 mmHg is based on the observation that the risk of cardiovascular morbidity increases significantly above this level. However, recent data have shown that an increased risk of cardiovascular morbidity is present in persons with blood pressure levels as low as 130/80 mmHg and that this risk increases progressively with rising blood pressure.[36-38] Data on prehypertension in young women are poor, and little is known on the risk factors for prehypertension. However, it may be speculated that a pre-existing prehypertensive state played a role in the development of cardiovascular compromise in the preclinical phase of preterm PE.

High-risk women developing PE versus high-risk women with uneventful pregnancy outcome

The group of uterine artery Doppler screen positive (high PI) women with normal pregnancy outcome did not exhibit any functional cardiac abnormalities compared with screen negative women with uneventful pregnancy. Instead, there was evidence of enhanced radial systolic function, as demonstrated by increased ejection fraction. This finding is in agreement with that from Valensise's group, although the methodology they used was different.[39] This supports the concept that PE is a complex disorder related to the ability of the maternal cardiovascular system to adapt to placental dysfunction. It seems that the women with abnormal uterine artery Doppler have different cardiovascular profiles at mid-gestation depending on whether they later develop preterm PE or not.

Conclusions

Preterm and term PE exhibit different cardiovascular profiles at mid-gestation of pregnancy, before the onset of overt disease. Asymptomatic cardiac diastolic dysfunction and impaired myocardial relaxation at mid-gestation is only seen in women who subsequently develop preterm PE but not term PE. It is now evident that women who developed preterm PE in pregnancy have a much higher incidence of developing symptomatic heart failure many years after delivery.[12, 13] Although it is not possible to distinguish pre-existing cardiac dysfunction from that acquired as a result of pregnancy, these cardiac findings may be useful in understanding the cardiovascular pathophysiology of PE.

Disclosure of interests

None.

Contribution to authorship

BT conceived and designed the research, performed statistical analysis, interpreted the data, drafted the manuscript and made critical revisions of the manuscript for important intellectual content. KM designed the research and performed echocardiography, tissue Doppler, strain and strain rate analyses, performed statistical analysis, interpreted the data, drafted the manuscript and made critical revisions of the manuscript for important intellectual content. GS interpreted the data and made critical revisions of the manuscript for important intellectual content. MN collected data. RS made critical revisions of the manuscript for important intellectual content. All authors contributed to and approved the final version.

Details of ethics approval

This study was approved by the Wandsworth Local Research Ethics Committee (Ref No.: 01.78.5).

Funding

KM was funded for her PhD thesis by the University of Chieti, Italy.

Acknowledgements

We are grateful to the midwives and sonographers of the Fetal Maternal Medicine Unit, St George's Hospital for their invaluable help with recruitment of participants.

Ancillary