Maternal total vascular resistance and concentric geometry: a key to identify uncomplicated gestational hypertension

Authors


Prof H Valensise, Università di Roma “Tor Vergata”, Dipartimento di Ginecologia ed Ostetricia, Isola Tiberina 39, 00100, Roma, Italy. Email valensise@med.uniroma2.it

Abstract

Objective  To evaluate the prognostic impact of elevated total vascular resistance (TVR) on the outcome of pregnancy in early mild gestational hypertension (EMGH).

Design  Prospective observational study.

Setting  Data collected from women with EMGH referred to the obstetrics outpatient clinic of Tor Vergata University from June 2003 to June 2005.

Population  A total of 268 women with EMGH (systolic and diastolic blood pressure [BP] 140–150 mmHg and 90–99 mmHg, respectively, without significant proteinuria).

Methods  Women had a maternal echocardiographic examination and BP examination within 24 hours of diagnosis. From this, the TVR was calculated and the geometric pattern of the left ventricle assessed.

Main outcome measures  Fetal/maternal adverse outcomes (pre-eclampsia, preterm delivery, placental abruption, other maternal medical problems, fetal distress, neonatal low birthweight, admittance to neonatal intensive care unit and perinatal death).

Results  Ninety-two out of the 268 pregnancies showed adverse outcomes (34.3%). The best independent predictor for the composite of maternal and fetal complications was TVR (OR 64.4, 95% CI 25.9–160.1). The cutoff value was 1340 dyn seconds/cm5 with a sensitivity and a specificity of 90 and 91%, respectively. Concentric geometry of the left ventricle was also an independent predictor (OR 4.72, 95% CI 1.85–12.04).

Conclusions  Echocardiography could help in identifying women with EMGH who subsequently develop maternal and fetal complications, allowing a classification in high-risk (TVR > 1340 dyn seconds/cm5, concentric geometry of the left ventricle) and low-risk women (TVR < 1340 dyn seconds/cm5, nonconcentric geometry of the left ventricle) for adverse outcomes of pregnancy.

Introduction

Recent reports assess a progression to maternal/fetal complications in about 45% of women with early mild gestational hypertension1 (EMGH). Nevertheless, the presence of elevated systolic blood pressure (SBP) and diastolic blood pressure (DBP) values in gestational hypertension is often associated with normal outcomes. The classifications that we use in clinical practice, such as that from Davey and MacGillivray,2 are important for research and epidemiological purposes but do not identify which patients will develop complications. Moreover, other reports have demonstrated that normotensive women with abnormal placentation may develop complications (i.e. fetal growth restriction [FGR]).3,4 These results demonstrate the poor predictive value of blood pressure (BP) alone in the prediction of fetal complications.

In the search for factors that can predict adverse outcomes (abnormal uterine Doppler, serum levels of inhibin, etc), attention has been focused on uterine cardiovascular parameters.5–8 However, many reports have shown that the cardiovascular adaptation to pregnancy is altered during complicated pregnancies9–21 and that altered geometry of the left ventricle in gestational hypertension is associated with adverse pregnancy outcomes.15 Women with gestational hypertension who subsequently develop complications also showed some functional differences compared with uncomplicated women with gestational hypertension, in particular, higher total vascular resistance (TVR).15 The aim of the present study was to assess the ‘risk’ of developing maternal and/or fetal complications (e.g. pre-eclampsia, FGR) in EMGH on the basis of maternal TVR and cardiac structural and functional parameters.

Methods

Patient selection

Two hundred and seventy-nine consecutive pregnant women with EMGH referred to the outpatient clinic of Tor Vergata University between June 2003 and June 2005 were recruited between 28 and 31 weeks of gestation. Two hundred and sixty-eight women agreed to participate in the study. Since TVR can change during pregnancy, recruitment was confined to a limited range of time (28–31 weeks of gestation) in which the influence of gestational age on TVR might be the lowest, obtaining homogeneous TVR values. A control group of 41 normotensive pregnant women matched for age and gestational age was also enrolled during the same period of time and gave informed consent. Women with EMGH had the following inclusion criteria:

  • 1The finding of EMGH: SBP values of 140–150 mmHg and DBP values of 90–99 mmHg in two different measurements 4–6 hours apart, according to the definition of Davey and MacGillivray.2
  • 2Normal BP values before 20 weeks of gestation.
  • 3Absence of proteinuria on urinary dipstick at the time of enrolment.
  • 4Absence of thrombocytopenia, elevated liver enzymes, haemolysis, elevated liver enzymes and low platelet count (HELLP) syndrome or coagulation abnormalities.
  • 5Absence of any antihypertensive treatment.
  • 6Estimated fetal weight of >10th percentile.
  • 7Normal fetal umbilical artery (UA) and middle cerebral artery (MCA) Doppler and amniotic fluid index.

Exclusion criteria included the following:

  • 1Undetermined gestational age.
  • 2Tobacco use.
  • 3Twin pregnancies.
  • 4Maternal heart disease.
  • 5Pre-existing maternal chronic medical problems.
  • 6Chromosomal and/or suspected echographic fetal abnormalities.

Women with EMGH were not given pharmacological treatments, unless they developed moderate to severe gestational hypertension later in pregnancy, as suggested by international guidelines.22 Pharmacological treatment included calcium antagonists, α-methyldopa and clonidine.

The BP was measured using a mercury sphygmomanometer in the seated position with the arm at the level of the heart and the feet supported or on the ground. Korotkoff phase V was used for the determination of DBP.23

The approval of the University Ethics Committee was obtained, and written informed consent was collected from all participants.

Fetal and uterine artery ultrasound examination

For all ultrasound examinations, a 3.5-MHz sector ultrasound transducer (Esaote Technos, Genova, Italy) was used with the high-pass filter at 100 Hz. All participants underwent uterine artery colour Doppler examination with the calculation of the resistance index (RI) as previously described.11–15 A mean RI of both uterine arteries >0.6 was considered abnormal.

Fetal Doppler measurements were obtained from the UA and MCA with the calculation of pulsatility indices (PI) as previously described.3,4,15 A UA PI of >95th centile or an MCA PI of <5th centile for gestational age was considered abnormal.

Fetal biometry and estimated fetal weight were assessed according to the local reference values.24

Echocardiographic evaluation

A two-dimensional (2D) echocardiogram with M-mode and Doppler was performed within 24 hours of the diagnosis of EMGH, and the TVR and systolic and morphologic parameters of the left ventricle calculated.

The M-mode, 2D and Doppler echocardiographic evaluations were performed with the woman in a lateral position with a 1.8- to 3.6-MHz transducer in harmonic imaging interfaced with a commercially available echocardiographic machine (Siemens Acuson Sequoia C256, Mountain View, CA, USA). The left ventricular end diastolic diameter (LVDd) and left ventricular end systolic diameters (LVDs), interventricular septum diastolic thickness (IVSd) and posterior wall diastolic thickness (PWd) were detected in the parasternal long axis view during M-mode tracing, according to the recommendation of the American Society of Echocardiography.25 Left ventricular mass (LVM) in grams was calculated by the Devereux formula.26

Left ventricular geometric pattern

The LVM index (LVMi) was then calculated as follows: LVMi = LVM/m2.7, where ‘m’ is the height (in metres) of the women.27 The relative wall thickness (RWT) was calculated as the ratio (IVSd + PWd)/LVDd.11,28,29

Systolic function

The stroke volume (SV) was calculated as the product of the aortic valve area and the aortic flow velocity time integral as previously described.11 The cardiac output (CO) was calculated as the product of the SV and the heart rate (HR) derived from electrocardiographic monitoring.

Total vascular resistance

At the end of the maternal echocardiographic examination, the BP was measured from the brachial artery with a manual cuff. The TVR was calculated in dyn seconds/cm5 according to the following formula:

image

where MBP is mean blood pressure calculated as DBP + (SBP−DBP)/3.

Outcome

The pregnancy was followed until term by an investigator, blinded as to the results of maternal echocardiography, to determine its outcome. Women who subsequently developed moderate to severe hypertension were pharmacologically treated. Poor BP control despite aggressive pharmacological treatment was considered an indication for delivery, although the exact timing of this was dependent on the individual clinician.

Pregnancies were classified as complicated when any of the following adverse maternal and fetoneonatal outcomes occurred:

  • 1The occurrence of proteinuria of >300 mg/24 hours (pre-eclampsia).
  • 2The development of moderate to severe hypertension resulting in an induced preterm delivery before 34 weeks for maternal conditions (women with continued instability of their BP despite aggressive pharmacological treatment) and/or deterioration of intrauterine fetal conditions (i.e. Doppler and/or cardiotocographic signs of fetal distress).
  • 3Placental abruption.
  • 4Other maternal complications: thrombocytopenia, elevated liver enzymes, HELLP syndrome or coagulation abnormalities.,
  • 5FGR with a birthweight <10th centile for gestational age of the reference population.24
  • 6Admittance to neonatal intensive care unit.
  • 7Perinatal death.

Postpartum control

Women with EMGH and controls had a clinical and echocardiographic examination 6–8 weeks after delivery to assess maternal TVR and left ventricular morphologic parameters.

Statistical analysis

Women were classified as either with uncomplicated or complicated EMGH. Values are expressed as mean ± SD.

Comparisons between groups were performed using the one-way analysis of variance (ANOVA) with Student–Newman Keuls correction for multiple comparisons, whereas the intragroup comparison between pregnancy and postpartum data was performed using the ANOVA for repeated measurements. The Mann–Whitney U test was used for non-normally distributed data.

Receiver–operator characteristic (ROC) curves and χ2 tests for RWT and TVR were constructed to find the best predictor parameters and the optimal cutoff values to identify subsequently complicated EMGH pregnancies (as defined above). The test accuracy was used to identify the best cutoff value with accuracy calculated according to the following formula: (true positive + true negative)/(true positive + true negative + false positive + false negative).

Univariate and multivariate binary logistic regression analyses were used to find predictors of complicated pregnancy during EMGH. The analysis included:

  • 1History of pre-eclampsia, gestational hypertension and gestational diabetes; family history of pre-eclampsia, diabetes, premature vascular disease, autoimmune or renal disorders.
  • 2Primiparity.
  • 3An age of >35 years.
  • 4Concentric geometry (RWT ≥ 0.45).
  • 5TVR (TVR ≥ 1340 dyn seconds/cm5).

To test intra-observer and inter-observer variability, two independent observers measured data on videotape recordings from 20 randomly selected women. The same data were than re-measured on tape after 1 month by one of the two observers.

Results

Data were collected for all 268 women with EMGH and for 41 controls. No participant was lost to follow up. Out of 268 EMGH pregnancies reviewed after delivery, 176 women (65.7%) with gestational hypertension had an uncomplicated pregnancy (uncomplicated EMGH group), while 92 (34.3%) women developed maternal and/or fetal complications (complicated EMGH group).

Table 1 shows the main complications developed after the enrolment in the study: 39 pregnancies had isolated maternal complications, 26 pregnancies had isolated fetal–neonatal complications and 27 had both maternal and fetal complications.

Table 1.  Main maternal and fetal/neonatal complications subsequently developed in women with gestational hypertension
Complications occurred in the study groupn, 92/268
Maternal complications39
Appearance of proteinuria >300 mg/24 hours (pre-eclampsia)17
Evolution towards moderate to severe gestational hypertension with induced preterm delivery <34 weeks13
Placental abruption3
HELLP syndrome (two women), coagulation abnormalities (two women), elevated liver enzymes (one woman), thrombocytopenia (one woman)6
Fetal/neonatal complications26
FGR19
Admittance to neonatal intensive care unit5
Perinatal death2
Maternal and fetal/neonatal complications27
Pre-eclampsia and FGR18
Evolution towards moderate to severe gestational hypertension and FGR with induced preterm delivery <34 weeks8
HELLP syndrome and neonatal death1
Total complications92

Table 2 shows the baseline characteristics of the women at the moment of the diagnosis, retrospectively classified as uncomplicated and complicated EMGH. Women in the complicated EMGH group gave birth significantly earlier and to neonates of significantly lower birthweight compared with the uncomplicated EMGH group. Moreover, primiparas more often developed complications as shown from the significantly higher percentage of nulliparous women in the complicated EMGH group.

Table 2.  Baseline features of the study groups at the moment of the diagnosis of gestational hypertension
ParameterControls, n= 41Uncomplicated gestational hypertension, n= 176Complicated gestational hypertension, n= 92P value
  1. 1, controls; 2, uncomplicated gestational hypertension; 3, complicated gestational hypertension; NS, non significant. Values expressed as mean ± SD.

Age (years)32 ± 332 ± 433 ± 4NS
Primiparity, n (%)11 (26.8)51 (29.0)40 (43.5)0.01, 3 vs 2; 3 vs 1
Height (m)1.63 ± 0.051.63 ± 0.061.64 ± 0.05NS
Pre-pregnancy body mass index (kg/m2)23.5 ± 3.023.9 ± 2.024.1 ± 2.1NS
Gestational age at delivery (week)39.1 ± 1.038.3 ± 1.232.3 ± 4.00.001, 3 vs 2; 3 vs 1
Weight centile49 ± 1553 ± 2518 ± 150.01, 3 vs 2; 3 vs 1

Tables 3a and 3b show the haemodynamic and left ventricular morphological features of controls, women with uncomplicated and complicated EMGH at enrolment (28–31 weeks) and postpartum.

Table 3.  M-mode-derived and 2D-derived parameters at 28–31 weeks of gestation and 6–8 weeks postpartum
ParameterControls, n= 41Uncomplicated gestational hypertension, n= 176Complicated gestational hypertension, n= 92P value
  1. 1, controls; 2, uncomplicated gestational hypertension; 3, complicated gestational hypertension; NS, non significant. Values expressed as mean ± SD.

Panel a: at 28–31 weeks of gestation
HR87 ± 690 ± 1280 ± 13<0.001, 1 vs 3; 2 vs 3
SBP (mmHg)111 ± 7144 ± 10145 ± 10<0.001, 1 vs 2; 1 vs 3
DB (mmHg)62 ± 783 ± 1285 ± 9<0.001, 1 vs 2; 1 vs 3
Mean arterial pressure (mmHg)78 ± 6103 ± 9105 ± 7<0.001, 1 vs 2; 1 vs 3
TVR (dyn seconds/cm5)949 ± 1501138 ± 1831754 ± 425<0.001, 1 vs 2; 1 vs 3; 2 vs 3
LVM (g)148 ± 20176 ± 25169 ± 32<0.001, 1 vs 2; 1 vs 3; 2 vs 3
LVMi (g/m2.7)40 ± 647 ± 845 ± 9<0.001, 1 vs 2; 1 vs 3
LVDd (cm)4.77 ± 0.264.88 ± 0.274.58 ± 0.26<0.001, 1 vs 2; 1 vs 3; 2 vs 3
IVSd (cm)0.92 ± 0.081.02 ± 0.081.05 ± 0.12<0.001, 1 vs 2; 1 vs 3; 2 vs 3
PWd (cm)0.89 ± 0.070.98 ± 0.081.03 ± 0.12<0.001, 1 vs 2; 1 vs 3; 2 vs 3
RWT0.38 ± 0.040.41 ± 0.030.46 ± 0.05<0.001, 1 vs 2; 1 vs 3; 2 vs 3
SV (ml)77 ± 1082 ± 1162 ± 11<0.001, 1 vs 2; 1 vs 3; 2 vs 3
CO (l)6.75 ± 0.967.4 ± 1.25.0 ± 1.0<0.001, 1 vs 2; 1 vs 3; 2 vs 3
Panel b: at 6–8 weeks postpartum
HR76 ± 871 ± 1172 ± 80.010, 1 vs 2; 1 vs 3
SBP (mmHg)116 ± 8122 ± 9128 ± 8<0.001, 1 vs 2; 1 vs 3; 2 vs 3
DBP (mmHg)66 ± 871 ± 1078 ± 9<0.001, 1 vs 2; 1 vs 3; 2 vs 3
Mean arterial pressure (mmHg)82 ± 789 ± 995 ± 8<0.001, 1 vs 2; 1 vs 3; 2 vs 3
TVR (dyn seconds/cm5)1288 ± 1561405 ± 3111574 ± 323<0.001, 1 vs 2; 1 vs 3; 2 vs 3
LVM (g)111 ± 14122 ± 18132 ± 20<0.001, 1 vs 2; 1 vs 3; 2 vs 3
LVMi (g/m2.7)32 ± 533 ± 536 ± 6<0.001, 1 vs 3; 2 vs 3
LVDd (cm)4.53 ± 0.194.73 ± 0.184.61 ± 0.24<0.001, 1 vs 2; 1 vs 3; 2 vs 3
IVSd (cm)0.79 ± 0.080.80 ± 0.090.89 ± 0.10<0.001, 1 vs 3; 2 vs 3
PWd (cm)0.77 ± 0.060.79 ± 0.080.86 ± 0.09<0.001, 1 vs 3; 2 vs 3
RWT0.34 ± 0.030.34 ± 0.040.36 ± 0.05<0.001, 1 vs 3; 2 vs 3
SV (ml)69 ± 1073 ± 1070 ± 90.001, 1 vs 2; 2 vs 3
CO (l)5.20 ± 0.575.25 ± 1.005.01 ± 0.85NS

Intragroup comparisons in controls showed significantly (P < 0.001) higher HR, LVM and LVMi, LVDd, IVSd and PW thickness, RWT, SV and CO whereas TVR was lower during pregnancy compared with the postpartum values.

Intragroup comparisons in the uncomplicated EMGH group showed a significantly (P < 0.001) higher HR, SBP, DBP, MBP, LVM and LVMi, LVDd, IVSd and PW thickness, RWT, SV and CO in pregnancy compared with the postpartum values, while the TVR was significantly lower.

Intragroup comparisons in the complicated EMGH group showed a significantly (P < 0.001) higher SBP, DBP, MBP, LVM and LVMi, IVSd and PW thickness, RWT and TVR in pregnancy compared with the postpartum values, while the SV was lower. The CO was unchanged.

No differences were present when comparing primiparas versus multiparas except for age (31 ± 3 versus 36 ± 4 years; P < 0.001) and the prevalence of complications (44.0 versus 29.4%, P < 0.05).

Figure 1 shows the ROC curve for TVR in the overall population. The cutoff value for TVR was 1340 dyn seconds/cm5 and showed a 90% sensitivity, 91% specificity, a positive predictive value (PPV) of 85% and a negative predictive value (NPV) of 95%. A separate ROC curve was also built for primiparas and multiparas (not shown). The accuracy reached 91% with a cutoff value for TVR in primiparas at 1340 and at 1400 dyn·seconds/cm5, whereas in multiparas the accuracy was 91% at both 1340 and 1350 dyn seconds/cm5. Table 4 shows the sensitivity, specificity, PPV and NPV for primiparas and multiparas at different cutoff values. On the basis of the test accuracy, 1340 dyn seconds/cm5 was considered the best cutoff for both primiparous and multiparous women.

Figure 1.

ROC curve for the prediction of complications (composite of maternal and fetoneonatal outcomes) using TVR.

Table 4.  Cutoff values for TVR in multiparous and primiparous women with sensitivity, specificity, PPV, NPV and accuracy
TVR cutoffMultiparous womenPrimiparous women
Sensitivity (%)Specificity (%)PPV (%)NPV (%)Accuracy (%)Sensitivity (%)Specificity (%)PPV (%)NPV (%)Accuracy (%)
8501005301003310064510047
90010083110035100124710051
950100143310039100164810053
1000100253610047100204910055
1050100374010055100295310060
1100100464310062100355510064
1150100544710067100415710067
120098625299729849609670
125098776499839869719781
130094867497899884839890
134087938394919588869691
135087938394919388869490
140075958790899092909291
145067979088887094908084
150062989186876398967782
155058989185866098967681

Figure 2 shows the ROC curve for the RWT. It shows the best cutoff value to be 0.45, with a 60% sensitivity, 93% specificity, 81% PPV and 82% NPV. A separate ROC curve was built for primiparas and multiparas (not shown). A cutoff value of 0.44 for RWT in primiparas showed a 63% sensitivity, 84% specificity, 76% PPV and 74% NPV, whereas with a cutoff value of 0.45 there was a 48% sensitivity, 92% specificity, 83% PPV and 69% NPV. A cutoff value of 0.44 for RWT in multiparas showed a 67% sensitivity, 90% specificity, 73% PPV and 87% NPV, while a cutoff value of 0.45 had a 58% sensitivity, 95% specificity, 83% PPV and 84% NPV. A concentric geometry was therefore diagnosed when RWT ≥ 0.45.

Figure 2.

ROC curve for the prediction of complications (composite of maternal and fetoneonatal outcomes) using RWT.

At enrolment, the uterine artery Doppler was abnormal in 42 women out of the 92 subsequently retrospectively classified as with complicated EMGH (45.7%) and in 65 of 176 women successively classified as with uncomplicated EMGH (36.5%). A concentric geometry was present in 27 of the uncomplicated (15.3%) and 58 of the complicated EMGH group (63.0%). In the uncomplicated EMGH group, TVR was equal or higher than 1340 dyn seconds/cm5 in 15 out of 176 (8.5%), whereas in the complicated EMGH group, 83 (90.2%) women had TVR values at or above the cutoff. Thirty-nine (22.2%) of the uncomplicated and 26 (28.3%) of the complicated EMGH pregnancies were aged >35 years. Fifty-one women (29.0%) of those with uncomplicated EMGH were primiparous, whereas 40 (43.5%) of the complicated EMGH group were primiparas.

The univariate and multivariate analysis is reported in Table 5: the TVR and concentric geometry appear to be the best independent predictors of complicated EMGH. Altered Doppler velocimetry of the uterine artery was not an independent predictor of complicated EMGH.

Table 5.  Univariate and multivariate binary logistic regression analysis for the prediction of complicated EMGH
Regression analysisPredictorOdds ratio95% CIP value
Univariate analysisFamily history2.591.54–4.35<0.001
Primiparity1.891.11–3.190.018
Age (>35 years)1.380.78–2.460.269
High uterine RI2.321.38–3.900.002
Concentric geometry (RWT ≥ 0.45)9.415.22–16.97<0.001
TVR (≥1340 dyn seconds/cm5)98.9941.56–235.75<0.001
Multivariate analysisFamily history1.370.50–3.760.536
Primiparity1.170.33–4.130.806
Age (>35 years)1.010.25–4.060.989
High uterine RI2.050.73–5.780.173
Concentric geometry (RWT ≥ 0.45)4.721.85–12.040.001
TVR (≥1340 dyn seconds/cm5)64.4125.90–160.13<0.001

Inter-observer and intra-observer variability

Intra-observer and inter-observer variability in terms of coefficient of variation (CV) and regression coefficient are reported. For the interventricular septum thickness, the CVs were 7.3% (r= 0.98) and 8.0% (r= 0.97) for intra- and inter-observer error, respectively. For the PW thickness, they were 8.1% (r= 0.98) and 8.2% (r= 0.97); for the LVDd, they were 5.0% (r= 0.98) and 7.9% (r= 0.96); for the LVDs, they were 7.1% (r= 0.96) and 8.8% (r= 0.95); for the LVM, they were 8.9% (r= 0.95) and 9.0% (r= 0.94) and for the TVR they were 7.5% (r= 0.97) and 8.1% (r= 0.96).

Discussion

There is limited information available about factors that predict the appearance of clinical complications in women with EMGH. Our aim was to search for predictive parameters of subsequent maternal and/or fetal complications during EMGH among maternal TVR, cardiac structural and functional parameters. The novel finding of our study is the independent prognostic prediction of adverse outcomes by both TVR and concentric geometry of the left ventricle.

The need to find accurate prognostic indicators for adverse maternal and fetal outcomes comes from the poor predictive power of BP values.30 Other authors have reported that in mild gestational hypertension remote from term, 46% of women eventually develop pre-eclampsia, with progression to severe disease in 9.6%.1 The development of proteinuria was associated with an earlier gestational age at delivery, lower birthweight and an increased incidence of small-for-gestational-age newborns, but this was not related to the BP level.1 Our results confirm these previous observations of a low yield of BP values as predictors of complications during gestational hypertension. In fact, in our study, there was no difference in BP levels at enrolment between women with uncomplicated and complicated EMGH.

Hypertensive disease of pregnancy is characterised by defective blood volume expansion.31,32 BP values and CO, both included in the TVR formula, could reflect this impaired condition better than the BP values alone. In fact, high values of TVR identify those women who show both increased BP and low CO.

Several reports show that maternal haemodynamics in pregnancies complicated by FGR (with or without maternal hypertension) and established pre-eclampsia are characterised by a defective plasma volume expansion, low CO and high TVR.3,4,9,10,14–21,33 On the other hand, there is still debate about the haemodynamic model of isolated gestational hypertension and the state of hypertension that precedes pre-eclampsia.11–13,15,33,34 In particular, the studies from Bosio et al.33 and Easterling et al.34 have shown that the preclinical phase of pre-eclampsia is characterised by a low TVR and high CO. In the series of Easterling et al.,34 women maintained this haemodynamic characteristic when pre-eclampsia developed, whereas the study from Bosio et al.33 showed a reduction in CO and an increase of peripheral resistance in established pre-eclampsia. Other studies have shown high TVR both in gestational hypertension and in the preclinical phase of pre-eclampsia.11,15 The differences between these two ‘models’ may be explained mainly by the fact that many of the women who developed pre-eclampsia in the studies from Easterling et al.34 and Bosio et al.33 had a high body mass index and obesity might in itself cause a haemodynamic state with high CO and low TVR. In the present study, women with uncomplicated EMGH showed lower TVR levels than subjects with complicated EMGH, although this was higher than controls. In our series, more than half of the complicated EMGH group had FGR (45 out of 92), which is usually characterised by a hypovolemic state3,4,20 with a low CO and high TVR.3,4,18 Nevertheless, in our study, women with EMGH who subsequently developed an isolated maternal complication (pre-eclampsia, induced preterm delivery, coagulation abnormalities) also showed a high TVR at enrolment. The fact that our participants were not obese may well explain the differences compared with series of Bosio et al.33 Moreover, in the 20 women who developed pre-eclampsia described by Bosio et al.,33 proteinuria occurred very late in the third trimester, whereas most of our participants had the complication before 34 weeks of gestation, denoting a more severe state of the disease and suggesting that the echocardiographic evaluation was closer to the development of complications than in the other study.

The high NPV of TVR for maternal and/or fetal adverse outcomes with the identification of a cutoff value could help us to identify women with EMGH with a low risk of developing complications. As might be expected, the rate of complications in EMGH primiparas is higher than in multiparas and this might explain the higher PPV and NPV of TVR in these women. All women selected in this study (both complicated and uncomplicated) had similar BP values at entry, confirming that BP values alone are not an effective predictor of adverse outcomes.

Concentric geometry has also emerged as an independent predictor for complicated EMGH, thus confirming previous findings.15 This structural response might be strictly linked to the higher TVR: the increased RWT might reduce the myocardial wall stress due to the higher BP regimen and TVR present in this group of women.

SV and CO appear to be lower in the complicated EMGH group. This haemodynamic feature, already reported in gestational and essential hypertension,15,35 probably represents an inadequate cardiovascular adaptation in response to an insufficient plasma volume expansion, determining a pressure overload and volume underload.

Another interesting result related to the difference between antepartum and postpartum data is the different haemodynamic features of the three groups: controls, uncomplicated EMGH and complicated EMGH. Simmons et al.14 reported that women with pre-eclampsia tend to decrease their TVR, while normotensive women show an increase from antenatal to postpartum. Our report is consistent with those previous findings, showing that the postpartum TVR in uncomplicated EMGH behaves as in normotensive women, with an increase seen at 6–8 weeks postpartum. Moreover, the finding of an elevated postpartum TVR and BP values in all women with EMGH compared with the controls might indicate a possible pre-gestational predisposition to pregnancy complications, as previously hypothesised.15 This hypothesis could be confirmed by the continuing postpartum follow up.

Our study confirms the important role of the altered uterine resistances and abnormal uterine artery Doppler waveforms in predicting complications during established gestational hypertension as previously reported,15,36,37 although it is not an independent predictor when analysed together with the cardiac parameters.

Hence, on the basis of echocardiographic results, we propose a possible clinical classification of gestational hypertension based on the presence or absence of cardiovascular features as shown in Table 6.

Table 6.  Proposed classification of gestational hypertension
 Low-risk hypertensive groupHigh-risk hypertensive group
TVR<1340 dyn seconds/cm5≥1340 dyn seconds/cm5
Concentric geometry (RWT)Absent (<0.45)Present (≥0.45)

Conclusion

Our study shows that before complications appear, the maternal haemodynamics of women with EMGH destined to have complications differ dramatically from those destined to have uncomplicated gestational hypertension. In particular, uncomplicated EMGH is characterised by low TVR and normal CO. A classification of gestational hypertension based on the presence or absence of cardiovascular features (elevated TVR, concentric geometry) might be helpful in identifying women for pharmacological treatment in order to reduce subsequent maternal and/or fetal complications. In fact, the identification of high-risk woman might allow the use of antihypertensive treatment with a rationale based on the need of reduction of the TVR and not only on the basis of BP values. The linkage between BP and cardiac function opens a new scenario in which the idea of the normalisation of the hypertensive patients should consider both parameters. The choice of the antihypertensive drugs could then be individualised and their efficacy followed during treatment.

Ancillary