Peripartum cardiomyopathy (PC) is a rare disease, recognized as early as the 18th century,1 and its diagnostic criteria were established in 1937.2,3 It is characterized by heart failure during the last month of pregnancy through the fifth month postpartum, without heart disease before the last gestational month, and no determinable cause.3

Because of its rarity, geographical differences and heterogeneous presentation, diagnosis may be difficult.3 Traditionally, it has been related to old maternal age, black race, greater parity and multiple gestation, but the underlying cause remains elusive.4 The reported prevalence of this disorder ranges from one in 100 to one in 15000 pregnancies.2,5,6 In a recent report, the incidence in 241497 deliveries was one in 4025, and was highest among African-Americans.7

In contrast to idiopathic dilated cardiomyopathy, left ventricular (LV) dilation and systolic dysfunction return to normal in more than 50% of patients within six months,2 although in a recent study in 100 African women with PC, the authors reported that ejection fraction (EF) returned to normal in only 23% of the patients.8 Higher EF and smaller LV diameter at the time of diagnosis have been shown to be associated with recovery and with a better prognosis,4,8–10 although there is some controversy.11 Although the presence of fibrosis, detected by biopsy12 and by cardiac magnetic resonance imaging (MRI),13,14 has been described in PC, its role in recovery of function is not known.

Dobutamine is a synthetic sympathomimetic amine that directly stimulates beta-1 receptors in the myocardium to increase myocardial contractility. Dobutamine stress echocardiography has been shown to be safe and accurate in detecting coronary artery disease15 and evaluating myocardial viability in patients with LV dysfunction.16 More recently, it has been used to analyze contractile reserve in patients with PC in order to predict recovery of function.17

The objectives of the present study were: (1) to assess whether low-dose dobutamine stress echocardiography, performed at an early stage of PC, can predict recovery of LV function in these patients; and (2) to assess whether myocardial fibrosis can be detected by cardiac magnetic resonance imaging (MRI).

Categorical data are presented as absolute values and percentages, and continuous data are expressed as mean values±SD. The significance of baseline differences was determined by the chi-square test, Fisher's exact test, or the unpaired t-test, as appropriate.

Pearson's correlation coefficient was used to measure the correlation between EF at baseline, during stress echocardiography and at follow-up. A p value of <0.05 was established as statistically significant. SPSS version 13 (SPSS Inc., Chicago, IL) was used for all analyses.

Nine women were studied. Mean age was 29.7±7.9 (range 16–38). All patients began to have symptoms of heart failure either prepartum (three patients, initial presentation: 3, 7 and 30 days before delivery) or postpartum (six patients, initial presentation: 4, 5, 9, 10, 12, and 30 days after delivery). Two presented with acute pulmonary edema, one with stroke and one with pulmonary embolism. Although four patients had a diagnosis of pre-eclampsia (44%), they had no previous history of heart failure, so the diagnosis of peripartum cardiomyopathy could be made.3,6 One patient had sickle cell trait.

At the time of DSE, all patients were on angiotensin-converting enzyme (ACE) inhibitors and diuretics. Three patients were on anticoagulants, two on beta-blockers and two on digoxin. There were no neonatal deaths.

Follow-up

Patients underwent a rest Doppler echocardiogram after 6 months of the clinical diagnosis of peripartum cardiomyopathy (mean follow-up 7.9±1.4 months, range 6–10). All patients, including the one in whom low EF persisted, were in class I or II, although the majority were still on medication (six patients were on ACE inhibitors, four on diuretics, two on beta-blockers and one on digoxin). One patient had pulmonary embolism a week after delivery and another had a cerebral stroke a month after delivery. No patient died.

Eight patients had an increase in EF, which normalized in seven patients. EF increased by a mean of 17.7±9.9% from baseline to follow-up (range 0–31%), mild systolic dysfunction persisting in one patient (patient 7, EF=51%) and severe dysfunction in another (patient 3, EF=24%). There was a strong correlation between EF at baseline and at follow-up (r=0.699, p=0.036) (Fig. 1).

Figure 1.

Correlation between EF at baseline and at follow-up.

(0,09MB).

The present study showed that EF at clinical onset of peripartum cardiomyopathy correlates with improvement in systolic function after six months. Although this has been previously suggested,22 several studies failed to demonstrate correlation of either LV size or LV function with outcome in peripartum cardiomyopathy.11,23,24

Similarly to our study, Fett et al. reported that LV echocardiographic features at diagnosis were unable to predict individually who would recover. They studied 92 patients in Haiti who were followed for five years and showed that LV echocardiographic features at diagnosis were not predictive of individual recovery. However, EF at diagnosis differed statistically between recovered and nonrecovered groups (28% vs. 23%, p<0.001).25 This is similar to what we observed in our study, in which a statistically significant difference in mean EF at diagnosis was seen between recovered and nonrecovered patients.

Although the US National Heart, Lung, and Blood Institute defines peripartum cardiomyopathy4 as heart failure within the last month of pregnancy or five months postpartum; absence of preexisting heart disease; no determinable etiology (the traditional definition) together with strict echocardiographic criteria of LV dysfunction: EF less than 45%, or M-mode fractional shortening less than 30%, or both, and end-diastolic dimension more than 2.7cm/m2, a cut-off value of 50% for EF was used in the present study. The reason for a higher cut-off value was that stress echocardiography was performed only when patients were stable and already receiving optimized treatment for congestive heart failure. However, all patients had objective signs and symptoms of congestive heart failure, appearing either in late pregnancy or in the puerperium (two patients had had acute pulmonary edema at presentation), and those who had EF >45% and <50% all had EF <45% at admission.

Six patients (67%) had mild or moderate systolic dysfunction at the time of stress echocardiography. The fact that EF was only mild or moderately reduced in six of these nine patients may be one of the reasons that eight of them improved at follow-up. In fact, the only patient in whom severe LV systolic dysfunction persisted had severe LV dysfunction at presentation. Witlin et al. demonstrated that women with severe myocardial dysfunction (LV end-diastolic dimension ≥60mm and fractional shortening ≤21%) resulting from peripartum cardiomyopathy are unlikely to regain normal cardiac function on follow-up echocardiographic study (severe dysfunction persisted in six out of seven), while in two of their patients with mild dysfunction, mild dysfunction persisted in one and the other recovered fully.22

Our finding of normalization of systolic function in seven of nine patients (78%) is in contrast to what has previously been reported, but supports more recent findings of a higher incidence of recovery of systolic function (around 75%) and better prognosis in patients with PC compared to other forms of dilated cardiomyopathy.26,27 The small number of patients with mild systolic dysfunction at the time of DSE may be one of the reasons that LV functional recovery occurred in the majority of our patients.

As in other studies, there was a history of hypertension during pregnancy in 44% of our patients, and this diagnosis should not exclude the diagnosis of peripartum cardiomyopathy.3

Lampert et al. used low-dose dobutamine stress echocardiography to demonstrate that nonpregnant women who had recovered from postpartum cardiomyopathy had lower contractility reserve than matched controls. The change in LV contractility over baseline values was significantly less in women with a normal rest echocardiogram who had recovered from peripartum cardiomyopathy than in a group of normal matched control subjects challenged in an identical manner. Although patients appear to recover clinically and by rest echocardiogram, there may be some residual subclinical abnormalities that can only be detected when the myocardium is subjected to a significant hemodynamic stress, such as the use of dobutamine stress echocardiography. The authors speculated that contractile reserve in these patients could offer information regarding cardiac performance in subsequent pregnancies, but this was not investigated by their study.28

In the present study, response of EF to dobutamine stress echocardiography was not predictive of EF at follow-up. Unlike our study, Dorbala et al. showed in six patients with PC that inotropic contractile reserve during dobutamine stress echocardiography accurately correlated with subsequent recovery of LV systolic function.17 However, a p value of 0.10 was used in their study for statistically significant findings.

Dobutamine was used in doses of 5μg/kg/m in the study by Lampert et al., 28 and up to 50μg/kg/m in the study by Dorbala et al.17 A dose of 5μg/kg/m may be insufficient to detect full contractile reserve, while 50μg/kg/m is too high to analyze it. Pelicka et al. demonstrated that stroke volume during dobutamine is commonly maximum at a dose of 20μg/kg/min,29 which is why this dose was used in the present study.

The presence of fibrosis, detected by endomyocardial biopsy, has been reported in five out of 11 African women with a diagnosis of PC. A “healed” pattern of myocarditis and fibrosis was found in five patients (three out of four with persistent heart failure presented this finding).12 In another study, endomyocardial biopsy was performed in a black woman with PC and no fibrosis, necrosis or evidence of inflammation was found. These authors speculate that normal endomyocardial biopsy findings during the acute phase of the disease may be predictive of recovery.30 However, the utility of endomyocardial biopsy is unclear, with studies involving a large number of patients showing no correlation of endomyocardial fibrosis findings with prognosis.23 Other less invasive methods are thus necessary.

MRI with hyperenhancement can detect myocardial fibrosis in dilated cardiomyopathy as a midwall and focal pattern, which differs from the subendocardial fibrosis seen in myocardial infarction.21,31 Fibrosis detected by MRI has been described in four out of 10 patients (40%) with peripartum cardiomyopathy.13 It occurred in only one of seven patients (14%) in the present study, and was minimal. It may be that patients with less severe involvement of the disease were included in our group. The mean time between dobutamine stress echocardiography and MRI was 16.3±11.3 months (range 2–29 months) with one patient having had MRI only two months after delivery. Although fibrosis has been shown to decrease from the acute phase to ten months later,14 the timing for development and/or regression of fibrosis in PC is unknown and its occurrence may have been different had MRI been performed earlier on during the disease, at least in some of our patients.

In patients with peripartum cardiomyopathy, LV ejection fraction at baseline was a predictor of LV functional recovery. Dobutamine stress echocardiography at presentation of the disease did not predict recovery at follow-up. Myocardial fibrosis as detected by cardiac MRI does not appear to be common in this disease.

The authors have no conflicts of interest to declare.