Cardiac MR derived ventricular global function index is independently associated with clinical outcomes in patients with severe left ventricular systolic dysfunction

Background/Purpose

Left ventricular global function index (LVGFI) is a novel marker that incorporates LV structure in the assessment of cardiac performance. We hypothesized that CMR-derived LVGFI could provide additive prognostic value in patients with severe LV systolic dysfunction (LVSD), in addition to known CMR prognosticators (LVEF, LV volume, GLS, LGE, native T1, ECV, myocardial perfusion reserve index (MPRI)).

Methods

We prospectively enrolled patients with known severe LVSD who underwent cardiovascular magnetic resonance imaging. The CMR protocol was performed using a 1.5 T system (Siemens, Magnetom Sola). Gated CMR examinations were performed, including regadesonon induced stress imaging, and LVEF, mass, volumes, GLS, pre- and post- contrast T1 mapping, LGE (number of segments and % extent) and MPRI were measured with dedicated software.

The primary endpoint included all-cause mortality and heart transplantation. Secondary endpoints included the primary endpoint, heart failure (HF) readmission, myocardial infarction, revascularization and stroke. LVGFI (%), defined as LV stroke volume/LV global volume*100 where LV global volume was calculated as the sum of LV mean cavity volume ([LV end diastolic volume + LV end systolic volume]/2) and myocardial volume (LV mass/myocardial density of 1.05 g/ml)3, was the primary exposure variable. The association between outcomes and LVGFI was assessed using a multivariable model adjusted with confounders.

Results

24 patients were prospectively included in the study between December 2024 and January 2025 (87.5 % male; mean age: 68.4 ± 9.46 years; mean LVEF: 27.7 ± (10.3) %). All patients had history of ischaemic (20/24 / 83.3 %) or non-ischaemic (4/24 / 16.7 %) cardiomyopathy. 18/24 patients had MR compatible CIED (75 %, all ICD DDDR). During a median of 13 months [interquartile range: 5 to 17], 9 (37.5 %) composite events [3 deaths (1/3 CV death), 4 admissions with HF decompensation, 1 stroke, 1 revascularization) occurred.

On linear regression, LVGFI was strongly associated (all p < 0.001) with LVEF and LV SVi and modestly associated with LV massi (p = 0.004) and LGE positive number of segments (p = 0.006). On Cox regression analysis, LVGFI was significantly predictive for the primary (HR=0.86 (0.75-0.98) 95 % CI, p=0.022) and secondary (HR=0.84 (0.73-0.97) 95 % CI, p =0.016) endpoint. For HF outcomes, Harrell's C-statistic for LVGFI was comparable to LVEF (both h-C: 0.817(0.632-0.952)). On penalized spline, LVGFI cutoff < 19 % was associated with higher primary events (7/12 [58.3 %] vs. 2/12 [16.6 %], Kaplan-Meier log-rank p < 0.01).

Conclusions

LVGFI is an independent predictor of survival in patients with severe LVSD and provides incremental prognostic value to known CMR prognosticators.Figure 1For image description, please refer to the figure legend and surrounding text.