003 Precise phenotyping with CMR identifies moderate alcohol consumption as an important phenotypic modifier of titin cardiomyopathy

• Published in: Heart (British Cardiac Society) Vol. 103; no. Suppl 1; pp. A2 - A3
• Main Authors: Tayal, U, Newsome, S, Whiffin, N, Buchan, R, Walsh, R, Barton, PJ, Ware, JS, Cook, SA, Prasad, SK
• Format: Journal Article
• Language: English
• Published: 01.04.2017
• ISSN: 1355-6037; 1468-201X
• DOI: 10.1136/heartjnl-2017-311399.3

BackgroundTruncating variants in titin (TTNtv) are the commonest genetic cause of dilated cardiomyopathy (DCM). They are notable for variable penetrance and expressivity, suggestive of environmental or genetic modifiers.PurposeUndertake deep phenotyping by cardiac MRI (CMR) to evaluate alcohol and hypertension as phenotypic modifiers of TTNtv cardiomyopathy.MethodsProspectively recruited DCM patients underwent comprehensive clinical evaluation, CMR with late-gadolinium enhancement and TTN sequencing.ResultsOverall, 733 subjects, mean age 53.4±14.4 years, 476 men (65.0%) were recruited. TTNtv were found in 82 (11.2%) patients.Abstract 003 Table 1Regression analysis evaluating TTNtv and alcohol excess as predictors of LVEFUnadjusted analysisAdjusted analysis* VariableEstimate(change in LVEF,%)95% confidence intervalsP valueEstimate(change in LVEF,%)95% confidence intervalsP valueBaseline: No TTNtv or alcohol excess0--0--TTNtv, no alcohol excess0.06−3.0 to 3.10.97−0.3−3.2 to 2.70.85Alcohol excess, no TTNtv−2.1−4.7 to 0.50.12−0.6−3.2 to 1.90.61TTNtv and alcohol excess−11.8−18.6 to −5.0<0.001−15.4−21.9 to −8.8<0.0001TTNtv*Alcohol excess interaction§−9.7−17.5 to −1.90.02−14.4−21.9 to −7.00.0001Table shows unadjusted univariable and adjusted multivariable analyses of the effect of TTNtv and alcohol excess on baseline left ventricular ejection fraction.*Adjusted for gender, a family history of DCM, a history of atrial fibrillation and the presence of mid wall fibrosis (late gadolinium enhancement on CMR) and NYHA class.§ i.e. the effect of TTNtv and alcohol excess compared to either TTNtv alone or alcohol excess alone.Patients with a history of alcohol excess (n=114, 15.6%) were more likely to be male (n=107 (93.9%) vs n=369 (59.6%), p<0.0001), have lower biventricular function (left/right ventricular ejection fraction; LVEF 36.3% vs 39.5%, p=0.01; RVEF 47.4% vs 52.5%, p<0.0001) and showed a trend to increased mid wall fibrosis (n=49 (43.0%) vs n=207 (33.4%), p=0.05) compared to patients without alcohol excess. There was no difference between groups in age at recruitment.Patients with TTNtv and a history of alcohol excess (n=13) had a lower mean LVEF compared to patients with TTNtv without a history of alcohol excess (n=69) (27.7%±12.7 vs 39.5%±13.1, p=0.005) (Figure 1). There was no difference in LVEF between TTNtv patients with a history of hypertension (n=10) and TTNv patients without (n=72) (37.7% vs 37.6%, p=0.94).Abstract 003 Figure 1Boxplot showing the change in left ventricular ejection fraction by TTNtv and alcohol status. LVEF= left ventricular ejection fraction. TTN+/- = presence or absence of truncating variants in TTN. EtOH+/- = presence or absence of history of excess alcohol consumption.While TTNtv, a history of alcohol excess or hypertension in isolation did not predict LVEF, the combination of TTNtv and a history of alcohol excess was associated with a 15.4% reduction in LVEF (95% CI −21.9 to −8.8%) compared to a patient without either risk factor (p<0.0001), independently of other predictors of LVEF (gender, NYHA class, mid-wall fibrosis, atrial fibrillation and a family history of DCM) (Table 1). There was no significant interaction between TTNtv and hypertension (Table 2).ConclusionsThese data demonstrate the potential of using CMR to study cardiovascular endophenotypes of genetic DCM. We identify a novel gene-environmental interaction between TTNtv and alcohol that prompts genetic studies in DCM in the context of high alcohol intake.Abstract 003 Table 2Regression analysis evaluating TTNtv and hypertension as predictors of LVEFUnadjusted analysisAdjusted analysis* VariableEstimate(change in LVEF,%)95% confidence intervalsP valueEstimate(change in LVEF,%)95% confidence intervalsP valueBaseline: No TTNtv or hypertension0--0--TTNtv, no hypertension−1.6−4.7 to 1.50.31−2.4−5.4 to 3.60.11Hypertension, no TTNtv−0.4−2.5 to 1.70.711.7−0.3 to 3.60.10TTNtv and hypertension−1.6−9.4 to 6.20.701.4−5.8 to 8.70.70TTNtv*hypertension interaction§0.5−8.0 to 8.90.922.2−5.7 to 10.10.58Table shows unadjusted univariable and adjusted multivariable analyses of the effect of TTNtv and hypertension on baseline left ventricular ejection fraction.*Adjusted for gender, a family history of DCM, a history of atrial fibrillation and the presence of mid wall fibrosis (late gadolinium enhancement on CMR) and NYHA class.§ i.e. the effect of TTNtv and hypertension compared to either TTNtv alone or hypertension alone.