The ESC guidelines recommend treating women with moderate- or high-risk pregnancies in specialized centres by an interdisciplinary pregnancy heart team. In a team-based approach, the pregnancy heart team was founded at University Hospital Heidelberg, in order to establish specific structures of communication and responsibility. Our interdisciplinary team consists of cardiologists, gynaecologists, anaesthesiologists, pediatricians, pulmologists and hemostasiologists. We report data on a group of high-risk patients, mostly women with mWHO > II-III and an elevated average age, who underwent pregnancy. We were faced with a heterogeneous patient collective—ranging from PPCM to spontaneous coronary artery dissection—and novel challenges such as IVF within the context of cardiac disease or pregnancy with congenital heart disease. Our data demonstrated a significantly higher number of maternal and fetal complications within this vulnerable patient collective and highlights the necessity of an interdisciplinary approach to adequately monitor these high-risk pregnancies in women with CVD. [10]

The cardiac medications commonly used (except for cardioselective betablockers and some diuretics) cannot usually be continued during pregnancy due to their teratogenic properties and must therefore be discontinued during pregnancy. Data suggests a negative effect following redrawal of heart failure medication in women with known heart failure. [12] However, within the context of pregnancy the data is insufficient. Furthermore, it is challenging to distinguish the symptoms of worsening or new-onset heart failure from the symptoms of physiological pregnancy. Therefore, a better understanding of the interpretation of cardiac biomarkers during pregnancy is needed.

Our main findings were: (1) Pregnant women with underlying cardiac disease are at a significantly higher risk associated with an elevated number of maternal and/or fetal complications. Therefore, they should be monitored by an interdisciplinary pregnancy heart team, depending on their risk profile. (2) We suggest that NT-proBNP primarily be used as an exclusion parameter for cardiac complications in pregnancy. For patients with pre-existing CVD, we recommend determining a baseline NT-proBNP value before pregnancy and performing follow-up tests each trimester or when symptoms progress. (3) Elevated hscTnT levels during pregnancy should always be a reason for further diagnostic testing. (4) Elevated cardiac biomarkers were associated with the occurrence of maternal and fetal complications and should result in close monitoring.

Our data showed an increase in cardiac biomarkers, particularly NT-proBNP, in the third trimester and after birth, regardless of left ventricular function. Previous studies have shown that cardiac output increases during pregnancy due to physiological changes such as an increase in LV mass or left ventricular end-systolic diameter. These results have been reproduced in a patient population with structural heart disease. However, in this specific patient group, a significant increase in diastolic dysfunction was observed during pregnancy, which persisted for up to 6 months after delivery and led to an overall weakened cardiovascular adaptation to pregnancy in women with structural heart disease. [1] In summary, systolic cardiac function alone is an unreliable parameter in a patient population with pre-existing CVD. This is consistent with our findings, which suggest the superiority of cardiac biomarkers in monitoring cardiac function during pregnancy—especially in a patient population of women with pre-existing CVD.

While NT-proBNP has a high negative predictive value during pregnancy and allows heart failure to be ruled out, its positive predictive value for heart failure is low. [14] Studies have reported elevated NT-proBNP concentrations in preeclampsia. [13] Furthermore, NT-proBNP typically increases in acute heart failure during pregnancy, especially in peripartum cardiomyopathy (PPCM). Interestingly, patients with lower NT-proBNP levels after diagnosis of PPCM showed a higher probability of recovery of left ventricular function than patients with higher NT-proBNP levels, suggesting an additional prognostic value of NT-proBNP in PPCM patients. [14] Since there is little data on specific cut-off values for NT-proBNP during pregnancy, and especially for women with pre-existing CVD, Sarma et al. suggest using an upper reference limit of 200 pg/ml for the first and second trimesters and an upper reference limit of 150 pg/ml in the third trimester. [8] Our data support this approach, as NT-proBNP values above 209 pg/ml (ROC-derived) were detected in acute complications such as acute heart failure or ischemic heart disease. We therefore support the approach of using NT-proBNP primarily as an exclusion parameter for cardiac complications and recommend further diagnostic measures in the event of an abnormal increase. For patients with pre-existing CVD, we recommend measuring a baseline NT-proBNP level before pregnancy and closely monitoring cardiac biomarkers during the trimesters.

hscTnT has also not been comprehensively studied in pregnant patients. Most studies show normal values during a physiological pregnancy, with a slight increase reported in patients with preeclampsia or pregnancy-induced hypertension. Elevated hscTnT values should therefore always be cause for further investigation. [8] In our patient population, elevated hscTnT levels were always associated with cardiovascular complications such as ischemic heart disease, perimyocarditis, or pulmonary embolism. Therefore, our data also suggest that further diagnostic investigations should be performed in cases of elevated hscTnT levels.

With regard to short-term maternal events, we were able to demonstrate a good discriminative value for both cardiac biomarkers, NT-proBNP and hscTnT, which exceeded that of LVEF. Adding the discriminative value of LVEF values in the logistic regression analysis did not significantly improve diagnostic accuracy (Figure S6, Supplementary material). Both highly elevated NT-proBNP and hscTnT levels were mostly associated with serious cardiac complications, both in women with pre-existing CVD and in women who developed CVD during pregnancy (Figure S7, Supplementary material). This often justified earlier delivery by CS and led to preterm births with subsequent fetal/neonatal complications. Our data highlight the particular vulnerability of women with underlying cardiac disease during pregnancy, with high rates of maternal and fetal complications. We registered some severe outcomes such as resuscitation and fetal death. Despite the ESC’s recommendation for interdisciplinary pregnancy heart teams, such structured approaches are not yet widely implemented. Greater awareness and adoption of these strategies are urgently needed to improve outcomes for this high-risk patient group. In summary, we are confronted with an increasing number of high-risk pregnancies, especially in pregnant women with pre-existing heart disease. A better understanding of prognostic outcome as well as the role of cardiac biomarkers is necessary for the appropriate monitoring and risk stratification of this vulnerable patient population.

This study is limited by its single-center design and relatively small sample size as well as heterogeneity, which may affect the generalizability of findings. As this was an all-comers patient population we naturally included a large number of women with a lower risk profile which limits our work’s clinical significance but nevertheless represents a typical patient population. Additionally, the majority of women presented during later stages of pregnancy – therefore the preceding biomarker and echocardiographic data could not be established. We also did not obtain speckle tracking data. The absence of a control group consisting of healthy pregnant women without CVD limits direct comparisons and hinders the ability to define “normal” trajectories of cardiac biomarkers in a comparable population. It is furthermore important to note that we focused some of our analyses on peak biomarker measurements which therefore limits interpretability regarding predictive factors. Long-term follow-up on maternal and neonatal outcomes was not performed. As participants were already under specialized care, clinical decisions—such as timing of delivery—were made within an interdisciplinary team and not blinded, potentially introducing bias in management strategies.