Study finds accounting for sex improves precision and prognostic performance of CMR biomarkers for heart failure

In a recent study published in the European Heart Journal Open, a group of researchers investigated the influence of sex on cardiac magnetic resonance (CMR) assessment of left ventricular filling pressure (LVFP) in heart failure (HF) diagnosis.

Background 

There is a critical need for distinct strategies to enhance the diagnosis and treatment of heart disease in women. HF is a growing global health issue, with over 64 million affected individuals.

Women disproportionately suffer from HF with preserved ejection fraction (HFpEF) due to factors like aging and hypertension. The ejection fraction (EF) method for classifying HF has limitations, leading to fewer therapeutic options for HFpEF compared to HF with reduced EF (HFrEF). Women are less likely to receive specialist HF care or optimal guideline therapy, resulting in lower quality of life.

Sex-specific differences in cardiac imaging and biomarkers have been noted. CMR-derived pulmonary capillary wedge pressure (PCWP) shows prognostic utility, but current non-invasive methods require refinement. Further research is needed to refine sex-specific diagnostic models and improve the accuracy and efficacy of HF diagnosis and treatment in women.

About the study 

The present study diagnosed HF in participants based on European Society of Cardiology guidelines, which require symptoms like breathlessness, clinical signs like peripheral edema, and evidence of cardiac dysfunction.

Two cohorts were included: a derivation cohort from the Assessing the Spectrum of Pulmonary Hypertension Identified at a Referral Centre (ASPIRE) registry in Sheffield, United Kingdom (UK), and a validation cohort from Leeds, UK.

The ASPIRE registry involved patients with suspected pulmonary hypertension who underwent imaging and right heart catheterization (RHC) within 24 hours. The validation cohort included patients with new HF diagnoses recruited between 2018 and 2020 and assessed with CMR.

Invasive studies used a thermodilution catheter to record PCWP. CMR studies involved capturing cine images with General Electric High Definition X (GE HDx) and Siemens Magnetom Prisma scanners. Blinded image analysis was performed using GE Advantage Workstation and Circle Cardiovascular Imaging version 42 (cvi42) software. Cardiac volumes, left ventricular mass (LVM), and maximum left atrial volume (LAV) were calculated.

The derivation cohort included 835 subjects, and a sex-specific CMR-derived PCWP equation was developed. The validation cohort, consisting of 434 subjects, applied this equation.

Outcomes were evaluated for major adverse cardiovascular events (MACE) and HF hospitalization. Statistical analysis included t-tests, X² tests, multivariable regression, Receiver Operating Characteristic (ROC) analysis, and Kaplan–Meier and Cox models. Statistical significance was set at P < 0.05.

Study results 

In the study, the derivation cohort included 835 participants, 60% of whom were female. The mean age was similar between sexes (66 ± 13 years, P = 0.84). Females had a significantly lower mean body surface area (BSA) than males (1.8 ± 0.2 vs. 2.0 ± 0.2 m², P < 0.0001) and higher systolic blood pressure (146 ± 28 vs. 140 ± 24 mmHg, P < 0.001). Females also exhibited a slightly higher mean heart rate than males (72 ± 15 vs. 70 ± 16 bpm, P = 0.02) and had a lower prevalence of chronic obstructive pulmonary disease (9% vs. 15%, P = 0.005). HFpEF was more common in females than males (62% vs. 40%, P < 0.001), while males had more HF with mid-range EF (7% vs. 2%, P < 0.001). There was no significant difference in invasive mean PCWP between females and males (14.0 ± 6 vs. 13.7 ± 6 mmHg, P = 0.52).

In CMR evaluation, females had lower left ventricular end-diastolic volume (LVEDV) and left ventricular end-systolic volume (LVESV), resulting in smaller left ventricular stroke volume (LVSV) and higher left ventricular ejection fraction (LVEF) than males. Males demonstrated higher LVM and LAV. In the right ventricle (RV), females had lower right ventricular end-diastolic volume (RVEDV), right ventricular end-systolic volume (RVESV), and right ventricular stroke volume (RVSV), but higher overall right ventricular ejection fraction (RVEF). Generic CMR-derived PCWP values were significantly higher in males (14.7 ± 4.0 vs. 13.0 ± 3.0 mmHg, P < 0.001) compared to females.

Sex, LAV, and LVM were used as input variables in stepwise multivariable regression to develop a sex-specific CMR-derived PCWP equation, with age as a weighted variable. The non-indexed equation was selected for better goodness of fit (R-value of 0.571) compared to indexed values. Internal cross-validation showed that the novel sex-specific model retained its independent association with invasively measured PCWP, while the generic CMR-derived model did not (beta = 1, standard error = 0.005, P < 0.0001, partial r = 0.57). The generic equation underestimated PCWP in females and overestimated it in males, whereas the sex-specific CMR equation showed no significant difference compared to invasive assessment.

During a mean follow-up of 2.4 ± 1.2 years in the validation cohort, 56 patients (12.3%) experienced MACE. Multivariable Cox proportional hazard regression indicated that only the sex-specific model was independently associated with MACE (beta = 0.92, standard error = 0.23, P = 0.001, HR 2.5, 95% CI 1.4–4.3). Kaplan–Meier analysis confirmed that sex-specific CMR-derived PCWP predicted MACE (X² = 11.4, P = 0.0007). CMR-derived PCWP remained predictive for MACE and HF hospitalizations, independent of LVEF, across all HF classifications.

Conclusion

This study demonstrated that a sex-specific CMR-modeled LVFP improves the precision of PCWP estimation and enhances prognostic performance in patients with HF.