DISCUSSION:
PPCM is a rare cause of HF, and incidence changes according to region and ethnicity. PPCM diagnosis is made by exclusion of the other causes of cardiomyopathies. The etiology of PPCM is uncertain; predisposition and acquired factors may take part. (1) This retrospective designed study evaluated clinical, echocardiographic, and cardiac magnetic resonance imaging, survival and recovery characteristics.
In the US, population incidence ranges from 1 in 100 to 1 in 4000, this rate was highest in Nigeria with 1/100 patients, and 1/400 in Haiti reports. (3) In our HF cohort, the incidence rate was 1.6/100 patients. This incidence rate was high as compared to the literature. The higher incidence rate in our institute could be the tertiary and referral center for HF patients. 57.9% of patients were referred to our center assessment for resolved acute HF. Ethnicity was impressed in the studies that African-American women were more predisposed to PPCM. IPAC registry results showed that African-American women had low initial LVEF, worse prognosis, and lower recovery rate than white women (4). In our study, all patients’ ethnicity was caucasian white women; therefore, no comparison could be made regarding ethnicity.
Hypertension was present in six (16.2%) patients before pregnancy, and preeclampsia was developed in 4 (10.8%) patients before presentation with HF. Preeclampsia-developed patients did not reveal any differences regarding the primary end-points. However, in four patients, LVEF recovered during the follow-up course (p=0.016). This finding could be ascertained by abruptly increased afterload may have caused left ventricular dysfunction with a predisposition to HF and pregnancy-related hormonal or molecular imbalance that shares a similar mechanism. (1) Clinical and cardiac imaging findings did not relate significantly to primary end-points.
Haghikia et al. analyzed 115 patients in the German registry. 19 (16.5%) patients reported a positive family history of cardiomyopathy. Family history was accepted as positive when PPCM, DCM, sudden death, and arrhythmias in family first-degree relatives were present. The recovery rate was not influenced by family history. (5) In our study, four patients had a family history of non-ischemic HF; this finding did not affect the primary end-point or LVEF recovery. Genetic analyses could not be performed on these patients.
PPCM has been associated with higher LVEF recovery rates, primarily seen in the first six months. The IPAC registry analyzed outcomes and predictors of recovery in 100 patients. The enrolled patient’s diagnosis was consistent with idiopathic non-ischemic cardiomyopathy. In the early postpartum period, mean LVEF increased from 0.35 (0.09 to 0.55) in twelve months. For women enrolled later in the course of the disease, LVEF stayed the same for 12 months, but it was lower than early postpartum women. Follow-up LVEF differed by race at 6 and 12 months. (4)
ESC EORP PPCM registry 2020, 730 patients were assessed. Results showed that recovery (LVEF>%50) occurred in 46% of the women, and 23% of women’s left ventricular dysfunction persisted at six months. (6)
In our cohort, left ventricular systolic function recovery was accepted as an improvement by 10% and increasing over 40%. Thirteen (35.1%) patients’ left ventricle systolic function recovered during the follow-up. The median time to recovery was 281.5 (IQR [78.7-358]) days. In patients admitted to our hospital early in the course of the disease recovery rate was 46.7%, and in referred patients, the recovery rate was 27.3% (p=0.30).
In PPCM, known for reversibility of the left ventricular systolic function, mortality is less announced in the studies. In the IPAC registry, follow-up at one-year mortality ranged from 4% to 11%. Black women were more likely to face worse prognoses. (4) Karaye et al. stated outcomes in Nigeria and found that 18.7% of women died due to any cause in a median of 17 months. Maternal age below 20 years, hypotension at presentation, and tachycardia were positively related to mortality. Obesity and beta-blocker therapy was related to reduced risk of death. (7) In 2020, ESC EORP PPCM registry results showed six-months outcomes: overall mortality was 6%, and regional mortality differed. In European countries, the mortality rate decreased to 4% in six months compared to Middle Eastern countries, where the mortality rate was 10%. This ESC EORP PPCM cohort was assessed for composite end-point (all-cause mortality, LVAD implantation, and heart transplantation) in PPCM patients with extended follow-up time. The death rate was 10.8%, and one-year death-free survival was 97.1% (2.9 events per year). The survival probability without a primary end-point was 11.9% in one year. Our study’s event rate in one year was lower than the EORP PPCM registry. (8) This could be explained by the higher rate of referred patients. Mortality probably is higher in the course of initial acute HF hospitalization. The availability of advanced HF therapies in our center probably affects more favorable survival.
In our study, we analyzed the CMR data of this patient population. In the literature, few studies present CMR data. Mouquet et al. analyzed 8 PPCM patients with CMR. LVEF measured by CMR correlated with echocardiography. The authors did not observe morphological aspects of LVNC or any other cardiomyopathies. CMR was re-performed in five patients. Neither baseline nor six months control CMR demonstrated an LGE pattern. (9) Marmursztejn et al. reported CMR results of two cases where one patient exhibited LGE. (10) Ersboll and colleagues analyzed a nationwide Danish cohort of 28 PPCM women with CMR: the mean LVEF was 62%, and one patient had LGE. The CMR was not performed in the early phase of PPCM(11). Liang et al. showed three CMR parameters correlated with recovery in PPCM: T1, T2, and extracellular volume (ECV). This study was performed on 21 PPCM patients and 20 age-matched patients. Although these parameters were related to LVEF recovery, ECV was the only independent parameter that surmised LVEF recovery. (12)
In the study performed by Arora et al., ten patients were retrospectively recruited, and their medical records paved the way for diagnosing PPCM. LGE was presented in 4 patients, all of whom had HF exacerbations during delivery. In addition, four patients had future pregnancies, and two women had LGE. These two women had HF decompensation during the following delivery(13).
This study reveals 25 patients’ baseline CMR results and seven control CMR results. Patients who have reached the primary end-point did not show a difference compared to the patients in whom the primary endpoint did not occur. Three patients with LGE reached the primary end-point. Two had sub-epicardial late enhancement; the first patient had an arrhythmic event during a HF medication titration course. Although LV systolic function improved during the follow-up, she experienced sudden cardiac death on the 136 days. The second patient implanted with an ICD had an appropriate shock for ventricular tachycardia. The third patient with LGE needed heart transplantation. These limited data revealed that PPCM with LGE had a poor prognosis, and sub-epicardial LGE exhibited patients may more likely have ventricular arrhythmia. This information may indicate that patients with sub-epicardial LGE are more likely to have arrhythmic death, so early ICD implantation strategies could be life-saving.
A few studies in the literature show trabeculations in LV consistent with LVNC during pregnancy. These trabeculations are best studied with CMR. In PPCM patients, a non-compaction cardiomyopathy phenotype was reported in case reports. Rehfeldt et al. shared a case report that suffered resuscitated cardiac arrest without past cardiac history. The echocardiographic assessment showed very low LVEF (5%), and she had LVAD surgery. She had prominent trabeculations compatible with LVNC phenotype after LVEF recovered at three months, and LVAD was removed. The NC/C ratio was not stated. (14)
Patel et al. reported two case reports claiming LVNC phenotype in PPCM patients. In the first patient, the NC/C ratio was 3.2 according to Jenni’s criteria and regressed to 2.5. The second patient had an end-systolic NC/C ratio of 1.63; however, this ratio was inconsistent with LVNC criteria. (15) Lea et al. presented a single case report: the patient diagnosed with acute HF during pregnancy showed an enlarged left ventricle and hypertrabeculation coherent with LVNC. After delivery, the patient’s clinical situation worsened, and they needed heart transplantation surgery. Explant material revealed LVNC properties. (16) and Rajagopalan et al. identified five patients who met the clinical definition of PPCM and LVNC imaging criteria. CMR was assessed in all patients. Three of the five patients had sustained left ventricular dysfunction; one had heart transplantation surgery, second had LVAD bridge to transplantation. In two patients, LVEF improved by over 40%.(17)
The development of trabeculations in the LV during pregnancy is another issue of discussion. Gati et al. evaluated 102 pregnant women and performed TTE in the first and third trimesters. 26 (25.4%) patients developed increased trabeculations; eight women revealed sufficient trabeculations to fulfill the criteria for LVNC. During the postpartum period, 19 (73%) women demonstrated complete resolution of the trabeculations, and five showed a reduction in the trabeculated layer. (18)
In our cohort, 25 women diagnosed with PPCM CMR were performed, and the NC/C ratio was measured. Eighteen patients met the LVNC Peterson criteria. Seven patients’ NC/C ratio was lower than 2.3 and did not confirm other cardiomyopathies. In Nine of eighteen LVNC phenotype patients, LVEF recovered. However, three patients with this phenotype reached the primary end-point (two death and one heart transplantation). In Seven of the 18 patients, CMR was re-performed, and in four patients NC/C ratio was over 2.3, consistent with LVNC. Three of Four patients’ NC/C ratio decreased, and in one patient NC/C ratio increased.
Studies in the literature and our paper raise the question of whether patients with LVNC who presented with first-time HF during pregnancy were misdiagnosed with peripartum CMP. PPCM is unknown idiopathic cardiomyopathy and has to be differentiated from other causes of HF. Thus a careful examination of patients with PPCM should be done, and CMR should be performed to exclude other forms of cardiomyopathies. Further research on the development of trabeculations during pregnancy is needed.