A major finding of the study is that the earliest marker of the carrier status in TTNtv-related DCM is left ventricular dysfunction defined as LVEF < 50% or LVE > 112% which preceded the development of overt HF and transient or persistent severe LVSD by approximately 10 years which, in turn, was followed by a variety of arrhythmia, both ventricular and atrial, as well as AVB. Of note, among screened relatives (n = 49) isolated LVE with normal LVEF was the first sign of cardiotitinopathy in 16% of subjects, reduced LVEF without LVE in 10%, and both abnormalities were detected simultaneously in 16% of subjects. LVE is known as the first sign of early DCM18–20. TTNtv were also associated with eccentric cardiac remodeling in analysis of cardiac magnetic resonance in healthy humans21. This is in agreement with a proposal of a new definition of DCM which recognizes 3 forms of preclinical phase of DCM, including isolated left ventricular dilation22. However, in a zebrafish model, serial echocardiography showed significant LVEF reduction in ttnatv/+ vs ttna+/+ zebrafish preceding LVE by 3–6 months23. Furthermore, comprehensive genomics-first studies on the impact of TTNtv on cardiac phenotype by Haggerty et al.8 and Pirruccello et al.24 showed that TTNtv carriers are characterized by lower LVEF but not larger left ventricular (LV) diastolic dimensions or volumes. Our DCM-dedicated study shows that both LVE and LVSD may be the first detectable abnormality.
To the best of our knowledge, circulating cardiac biomarkers in relation to either the disease penetrance or prognosis have not been reported in cardiotitinopathies. The role of circulating cardiac biomarkers in the detection of HF is widely recognized25. In community based studies, multiple cardiac biomarkers are detectable in ambulatory individuals and add prognostic value to standard risk factors for predicting mortality, overall cardiovascular events, and HF26,27. However, little is known about the significance of circulating biomarkers in early stage of DCM in humans26,28. We showed recently that elevated hs-cTnT is the earliest marker of the carrier status in cardiolaminopathies29 and might be a “red flag” in asymptomatic or mildly symptomatic carriers.
Our data show that determination of widely available serum biomarkers cannot replace echocardiography in detection of affected TTNtv carriers. NT-proBNP serum level is rarely elevated in subjects with mildly reduced LVEF or isolated LVE and it usually exceeds the normal range when HF symptoms and advanced LVSD are present. In contrast to cardiolaminopathies, elevated hs-cTnT concentration seems to be an indicator of end-stage phase of cardiotitinopathies.
In our cohort, AA was found during baseline evaluation in 23% of TTNtv carriers and VA in 40%. The frequencies are lower than in the largest to date study by Akhtar et al. who found AA in one third and VA in one-half of TTNtv carriers16. The difference can be explained by younger population and less advanced LVSD in our study. The timing of the appearance of VA seems similar in both cohorts, in conjunction with LVSD progression.
Authors of several recent papers report significant arrhythmic burden characteristic of TTNtv-associated DCM, most often found in a relatively advanced disease stage. In the study by Corden et al., having a TTNtv was associated with a higher risk of receiving appropriate ICD therapy in the group of 148 DCM patients with implanted ICDs (hazard ratio (HR) 4.9; p < 0.001)15. In addition, TTNtv was a risk factor for developing new persistent atrial fibrillation (AF)15. Tayal et al. found that patients with TTNtv are more likely to have a history of AA or VA at the time of DCM diagnosis14. In a Danish study on 115 TTNtv-related DCM patients with mean LVEF 28%, AF and MVA occurred in 43% and 23% of patients, respectively30. Of note, AF preceded the DCM diagnosis in 16% of pts, and in 11% the presenting symptom of DCM was MVA30.
In cardiolaminopathies, conduction disease and arrhythmias precede the onset of HF by seven years31–33. However, LMNA mutations are not a common cause of lone AF34. Arrhythmias are also reported in TTNtv carriers with normal cardiac function. In a case control study that included 2,781 participants with early-onset AF and normal LVEF, and 4,959 controls, there was a statistically significant association between TTNtv and AF (odds ratio, 1.76)35. Associations with arrhythmias, including AF, were also observed in the genomics-first study by Haggerty et al., even when conditioning on DCM diagnosis8. Among screened relatives in our study, AA or VA was found in 8/50 (16%) subjects and it was the earliest detected abnormality in 6 (12%) of them. This shows that in the course of cardiotitinopathies various clinical scenarios are possible: arrhythmias appear typically in the late disease stage but they can also precede the DCM diagnosis.
As shown previously by others7,15 and us3, LBBB is relatively uncommon among cardiotitinopathy patients and therefore CRT-D requirement is less pronounced.
Another major finding of our study was strong, independent association between NT-proBNP level ≥ 650 pg/ml and the occurrence of the composite endpoint of MVA and esHF among TTNtv carriers. Recently several studies defining prognostic factors in cardiotitinopathy have been published, however none of them included impact of circulating biomarkers7,14−16.
NT-proBNP serum concentration ≥ 650 pg/ml was also the best predictor of MVA in our study. Excellent prognostic role of NT-proBNP in patients with HF is widely recognized36. The association of raised levels of NT-proBNP and MVA in HF patients was shown previously in general HF cohorts37,38. NT-proBNP provides also information regarding the risk of SCD in a community-based population beyond other traditional risk factors39.
There is great need for identification of prognostic factors that may help in decision making with regard to ICD therapy. Our study suggests that NT-proBNP, a commonly available circulating biomarker, may be useful in the setting of clinically stable TTNtv carriers.
The hearts of patients with TTNtv and DCM have thinner LV walls and lower indexed LV mass compared to TTNtv-negative controls6. It results in higher LV wall stress and release of NT-proBNP, and is associated with increased risk of VA6,15,40. It explains why NT-proBNP may be a good predictor of both MVA and esHF in TTNtv-related DCM.
Unlike in cardiolaminopathies41, TTNtv-positive DCM patients have midwall replacement fibrosis detected in CMR with similar frequency as TTNtv-negative DCM controls15,40 but interstitial fibrosis is found at endomyocardial biopsy significantly more often40. We hypothesize that hs-cTnT leakage, detectable from early stages of cardiolaminopathy29, may reflect cardiomyocyte death and replacement fibrosis, prevalent in LMNA-related cardiac disease, but it may be undetactable in interstitial fibrosis, characteristic of cardiotitinopathy. This might explain why hs-cTnT level raises significantly only in the end-stage of TTNtv-related DCM whilst in earlier stages, when interstitial fibrosis and increased risk of life-threatening arrhythmias are already present, it remains within normal range.
Male patients were younger (37 vs 44 years, p = 0.04) and had more advanced HF, similarly as in earlier studies3,16. We found more AA among male patients (30 vs 12%, p = 0.04) but no significant differences with regard to nsVT. In largest to date study on phenotypic characterization of TTNtv carriers male probands had a higher prevalence of AF at baseline (20 vs 5%, p = 0.01), the difference with regard to nsVT was less pronounced (58 vs 42%, p = 0.046)16.
As Giudicessi et al stated42, the prevalence of TTNtv in the Genome Aggregation Database (1.8%) is more than 4 times higher than the estimated prevalence of DCM in general population (0.4%). This underlines the role of TTNtv as susceptibility variants and suggests that strong environmental effects or additional genetic factors contribute to the development of a cardiac phenotype43.
In this study we showed that pathogenic or likely pathogenic TTNtv in DCM patients were located in all domains of the gene and had high proportion spliced-in index, as in the study by Akhtar et al.44. Early studies2, including ours3, showed A-band location of TTNtv mutations as more specific for DCM patients. With many more TTNtv identified and more accumulated data, no statistically significant differences in baseline clinical phenotypes or LVSD penetrance associated with TTNtv location across different TTN bands were identified16.