Discrepancies in Reporting of Left-Ventricular Late Gadolinium Enhancement Pattern, Distribution, and Burden on Cardiac Magnetic Resonance

Background: Late gadolinium enhancement (LGE) is the gold standard for detecting myocardial fibrosis and scarring. The pattern, distribution, and extent of enhancement are important prognostic indicators for arrhythmogenesis and sudden cardiac death (SCD).^1–3 Unfortunately, many important clinical findings are not consistently reported in CMR reports due to technical, methodological, and interpretive variability^4,5. This study aims to assess the consistency of LGE reporting in ischemic (ICM) and non-ischemic cardiomyopathy (NICM).

Methods: A total of 828 CMR reports from cases read between 4/18/1990 and 4/19/2023 were analyzed, 400 from Cleveland Clinic Main Campus (Ohio) and 428 from Cleveland Clinic Florida. Reports were annotated using MedTator software and labeled with LGE attributes, including positivity, pattern, distribution, and severity (Figure 1B). CMR studies indicated for the evaluation of ICM and NICM were extracted via keyword search. We identified the percentages of reports indicating LGE positivity that document pattern, distribution, and severity. We further examined the frequency of LGE pattern and distribution reporting in patients with moderate/severe LGE burden compared with mild to assess whether attributes are reported more frequently in high-risk patients. Statistical analysis was conducted using chi-squared for differences in LGE attribute reporting by indication and Fisher’s exact for assessing differences by LGE burden.

Results: There was no significant difference in the documentation of LGE pattern or distribution in the ICM or NICM subgroups. Interestingly, there was reduced reporting of LGE burden in CMR studies indicated for ICM (Figure 2).

Across all indications, the proportion of reports documenting LGE pattern was 69.7% in cases with mild LGE and 34.8% in those with moderate/severe LGE (p = 0.014) (Figure 3A), while reporting of distribution was 57.6% and 52.2%, respectively (p = 0.79) (Figure 3B). For non-ischemic cardiomyopathy, pattern reporting was 52.6% (mild) versus 18.8% (moderate/severe, p = 0.078) (Figure 3A), and distribution reporting was 47.4% versus 62.5% (p = 0.50) (Figure 3B). For ischemia/viability, pattern reporting was 77.3% (mild) versus 45.5% (moderate/severe, p = 0.12) (Figure 3A), and distribution reporting was 63.6% for both groups (p = 1.0) (Figure 3B).

Conclusion: Our findings reveal substantial inconsistencies in the reporting of LGE pattern and distribution, with particularly low rates of pattern reporting in moderate/severe cases—even among high-risk groups such as patients with non-ischemic cardiomyopathy and ischemia/viability. These gaps in documentation may limit the clinical utility of CMR reports for risk stratification and management, and pose challenges for retrospective research. Standardizing the reporting of key LGE attributes could enhance the interpretability and clinical value of CMR, supporting more consistent care.

Figure 1: A) Cohort description and B) Annotation schema for recording LGE attributes.
Figure 2: Reporting of LGE location, pattern, and severity in ICM and NICM. * p < 0.05
Figure 3: Reporting of LGE A) Pattern and B) Distribution by LGE in ICM and NICM cases with mild and moderate/severe LGE burden. * p < 0.05