Diffuse LV Myocardial Fibrosis and its Clinical Associations in Adults With Repaired Tetralogy of Fallot

We recently investigated left ventricular (LV) myocardial fibrosis using the extracellular volume fraction (LV-ECV) and sought associations with other signs of cardiovascular dysfunction such as systolic dysfunction, exercise intolerance, and arrhythmia to investigate its potential role as a risk factor for complications after tetralogy of Fallot (TOF) repair.

We prospectively studied 52 TOF consented adults (without implanted devices or renal insufficiency) and 22 control subjects with magnetic resonance (1.5-T) for LV and right ventricular (RV) volume/function and atrial volumes (area-length method). A midventricular Look-Locker sequence (16 to 21 phases) was prescribed perpendicular to the septum before and 3, 7, and 15 min after Omniscan contrast (Novation LLC, GE Healthcare, Princeton, New Jersey) injection (0.15 mmol/kg). LV myocardium and blood pool were contoured (QMass, Medis Medical Imaging Systems, Leiden, the Netherlands). Time-signal intensity curves were used to quantify T₁ for myocardium and blood pool through exponential fitting, and its reciprocal R₁. The linear slope of R₁ for myocardium versus blood defined lambda. Multiplying by (1 − hematocrit/100) gave LV-ECV. For 5 subjects, faulty baseline T₁ values required an imputation using the mean for the cohort.

LV-ECV was greater in TOF than in control subjects (28.2 ± 3.9% vs. 26.1 ± 2.8%, p = 0.026) and higher in women than in men (30.0 ± 3.8% vs. 26.5 ± 3.3%, p = 0.001 in TOF and 28.5 ± 1.4% vs. 24.4 ± 2.2%, p < 0.001 in control subjects). Other differences are shown in Table 1. Among TOF subjects, age correlated positively with LV-ECV in TOF subjects (r = 0.529, p < 0.001), but not in control subjects (r = −0.142). LV-ECV correlated with age at repair (r = 0.487, p < 0.001), QRS duration (r = 0.521, p = 0.002), 6-min walk distance (r = −0.417, p < 0.004), B-type brain natriuretic peptide (BNP) (r = 0.548, p < 0.001), left atrial (r = 0.363, p = 0.008), and right atrial volume (r = 0.326, p = 0.018).

Fifteen TOF subjects (29%, 95% confidence interval: 17% to 41%) had LV-ECV values ≥30% (Table 1). These differed by age, age at repair, number of previous surgeries, previous atrial arrhythmia, QRS duration, 6-min walk distance, BNP, and left atrial volume. There was no difference in ventricular size, function, or pulmonary valve regurgitation fraction.

Late gadolinium enhancement (LGE) showed small septal foci in 5 subjects, and in the inferior/inferolateral wall in 4 (excluding the RV/LV junction). Three of 5 subjects had septal LGE in the LV-ECV plane. All regions were very small and had no appreciable impact on the overall results, whether included or excluded.

Forty-eight TOF subjects were followed for 3.5 ± 1.5 years. Of those with LV-ECV ≥30%, 5 (33%) had significant clinical events (2 new atrial arrhythmia and 3 cardiovascular deaths) versus only 2 subjects (atrial arrhythmia) in those with LV-ECV <30% (p = 0.024).

Thus, LV myocardial fibrosis was found in >25% of subjects and associated with other adverse clinical markers and outcomes. These findings agree with other research showing fibrosis as a byproduct of adverse ventricular loading and contributor to clinical outcome (1,2). Therefore, efforts to prevent and treat fibrosis should be renewed, including earlier intervention and consideration of pharmacotherapeutic options that may attenuate fibrogenesis. Clinical management should focus equally on preventing and treating fibrosis as on anatomic defect repair.

A higher percentage of subjects with LV fibrosis were identified by LV-ECV than with LGE, consistent with findings of other studies (3). LV-ECV did not differentiate those with systolic dysfunction, also similar to others (2), implying that fibrosis precedes dysfunction. Other myocardial function metrics, including strain, may be more sensitive to early LV fibrosis. Chronic volume and pressure loading of the RV are undoubtedly inciting factors of LV dysfunction over long periods of time. Future longitudinal studies could determine gradual changes in response to RV pressure and/or volume loading.

Although it has been shown to be predictive of events in other populations, the potential of LV-ECV in surveillance of future events needs confirmation. Fibrosis in the RV by LGE (1) and T₁ mapping (2) has also been shown, but we did not quantify this in the thin-walled RV with our methods. Consistent with other studies (4,5), LV-ECV was higher in female subjects (both groups), meriting future investigation.

Therefore, LV fibrosis should be considered a significant contributor to an adverse natural history after TOF repair. With further validation, LV-ECV measurement may support future investigation of both the etiology and treatment of TOF subjects.

• American College of Cardiology Foundation