Creating certainty out of uncertainty

This editorial refers to ‘Insights into sudden cardiac death: exploring the potential relevance of non-diagnostic autopsy findings’†, by H. Raju et al., on page 831. Certainty, certainty, certainty is all what close relatives want to obtain when a loved one has died suddenly and unexpectedly. Certainty about what happened, certainty about whether this could have been prevented, and certainty concerning whether immediate family members carry a similar risk. Without doubt, the sudden death of a young person is a very tragic event. It most often has a cardiac nature [i.e. sudden cardiac death (SCD)]. In the elderly, structural heart disease forms the basis of the ventricular arrhythmias responsible for SCD, but in the young (≤35, 40, or 50 in different series), various other causes contribute significantly to these dramatic events, and a substantial subset even occurs in the presence of a completely structurally normal heart.1 , 2 In the last age group, most causes of SCD are (potentially) inherited and the questions posed above are thus very reasonable.1 Obtaining certainty after the sudden death of a young individual requires active investigation of the deceased (i.e. a full and detailed autopsy) or, in cases where an autopsy is not performed, a cardiac investigation of the surviving relatives. Although autopsy after the sudden death of a young individual has been greatly advocated for many years, the actual number of autopsies performed in such situations in countries where an autopsy is not mandatory is disappointingly low, despite active interventions to increase their use.3 Yet, when autopsy is performed, the preferred certainty about the cause of death is, unfortunately, not always obtained. Not that long ago in a significant subset of autopsies, this uncertainty was actually due to a ‘non-diagnostic finding’, i.e. findings that though abnormal, do not match the diagnostic criteria for a specific condition; such a classification is in fact not at all uncommon in an autopsy of an individual who died unexpectedly (60% in the study which this Editorial accompanies; see below). Moreover, a detailed cardiac autopsy does require significant expertise, as has been shown in an important study from St. George’s (London, UK). In 148 out of 200 cases of young SCD, two pathology reports, one from a coroner’s pathologist and one from an expert cardiac pathologist, i.e. Dr Mary Sheppard, were available and, in a surprising 41%, there was a disparity in the cardiac diagnosis.4 The majority of these disparate cases related to the definition of a normal heart.4 In this issue of the European Heart Journal, at-risk relatives of unexplained SCD (strictly speaking sudden death) cases without prior cardiac disease were investigated.5 The study started with a systematic autopsy which was performed in all 98 probands (the deceased individuals) and the group was then subdivided into those with a completely normal heart at autopsy (i.e. sudden arrhythmic death syndrome: SADS) and those in which non-diagnostic findings were identified during autopsy. The age group was between 1 and 64 years, i.e. excluding sudden infant death syndrome (SIDS). As expected, the SADS group was younger than the group with non-diagnostic findings. Indeed, in an earlier study, primary inherited arrhythmia syndromes were the exclusive suspected cause of death in the age group below 15 years and, with increasing age, structural heart disease became apparent.6 Hence, SADS is expected in the younger age group. Importantly, in both groups, screening of family members revealed a familial cardiogenetic diagnosis in an approximately equal portion (± 25%). This essentially means that what used to be an uncertain pathological diagnosis in the deceased in fact requires a similar follow-up with respect to further investigation in the family. It should be noted, however, that in 3 out of 12 families for which uncertain cardiac pathology findings were established in the deceased proband, a drug challenge was needed to unmask a long QT syndrome (LQTS) or Brugada syndrome phenotype in the family, and for both disorders drug challenges might not be as specific as previously thought.7 , 8 These results reinforce an earlier study performed in the UK on the same topic, including the current first author.9 However, in that study, a systematic autopsy was not performed in all cases. In fact, in the UK study, coroner’s (i.e. non-expert) autopsies were an important source of the pathological data in many cases. In addition, the present study also included a matched ‘control’ group of non-cardiac deaths (witnessed drowning’s and road accidents). In the latter group, fewer non-specific findings were observed, and this observation may point to a potential role for these non-specific findings in the arrhythmic death. Importantly, within the control group, inherited arrhythmias may also be present and, if they are, the conclusion that these non-specific findings do play a role may be too strong. Unfortunately, these patients are not described in detail but are, apparently, witnessed drowning and road accident victims. Actually, particularly in the latter, detailed circumstances of the event are of importance as unilateral accidents can also relate to a malignant arrhythmia and, in the former, in particular when it occurs in the very young, LQT1 or catecholaminergic polymorphic ventricular tachycardia (CPVT) should be suspected. In these cases, screening of immediate family members is of equal importance and would actually have been of interest to complete this study. Diagnostic roadmap for sudden unexplained death. *Uncertain pathology findings include myocyte hypertrophy, myocyte inflammatory infiltrate, interstitial fibrosis, or non-critical coronary artery disease. **Guidelines are those published in 2013 (Priori et al., Europace 2013;15:1389-1406). Molecular autopsy (using a selected panel of genes commonly associated with SCD) may be considered in cases in whom no clinical signs for an inherited condition (including SADS) are identified as the underlying cause for sudden unexplained death. When no mutation is identified, which is expected in the vast majority of cases, trio-based exome sequencing may be considered and when this reveals no potential causal genetic variant an oligogenic disease, which is expected not always to be associated with a family history of young sudden death. Alternatively, a non inherited cause may be suspected and family members are also expected to have a low risk of SCD. BrS, Brugada syndrome; CPVT, catecholaminergic polymorphic ventricular tachycardia; IVF, idiopathic ventricular fibrillation; LQTS, long QT syndrome. Based on these findings, certainty about what happened can thus be obtained in ± 25% of families, and that includes families with non-specific, i.e. uncertain, findings during autopsy.5 In a literature review conducted in 2015, family screening after the death of a close relative (with normal autopsy) revealed a most likely diagnosis (i.e. certainty) in 32% of families with a SADS case (mean yield, range 18–53%).10 A molecular diagnosis was obtained in 23–47% of families in whom a clinical diagnosis was obtained.10 It should, however, be realized that studies prior to 2015 might have been too lenient in calling mutations [no American College of Medical Genetics and Genomics (ACMG) criteria as implemented nowadays were used at that time, and the minor allele frequency threshold used for calling a variant pathogenic/probably pathogenic was often too lenient].11 Combining family screening with a ‘molecular autopsy’ (post-mortem molecular testing) seems to increase the yield of unravelling the cause of death to a level of ∼40%.11 Some variants occur de novo in the deceased, and is also important to know that because it does reassure the surviving family members. Potentially de novo mutations are picked up more readily with a standard ‘trio-based whole-exome or panel-based molecular autopsy’, i.e. a whole-exome analysis in the deceased and his or her parents, as has recently been performed by the Mayo clinic group.12 Despite all these new developments still about 60% of cases remain unexplained and whether the majority of deceased individuals died from inherited or non-inherited causes is not known. An ultimate possibility could be the presence of more than one variant where variants conspire to tip the balance towards an arrhythmic substrate (i.e. an oligenic inheritance as has been shown for Brugada syndrome). The finding that during follow-up no further casualties occurred in such families, except in those already with more than one young sudden death (i.e. a definite genetic basis), suggests that the vast majority of these remaining cases might actually be related to non-inherited causes,13 or related to an oligogenic inheritance which is also expected not always to be associated with a family history of young sudden death. With a follow-up duration of ‘only’ 6.6. years (interquartile range 4.7–9.6 years), this result may not be sufficient to provide absolute certainty for the surviving family members. Longer term follow-up of the latter and other cohorts of families in whom SADS occurred, including this study in the European Heart Journal,5 may confirm the above findings and, ultimately, create certainty for all surviving family members. The author acknowledges support from the Netherlands CardioVascular Research Initiative (CVON), the Dutch Heart Foundation, Dutch Federation of University Medical Centres, the Netherlands Organisation for Health Research and Development, and the Royal Netherlands Academy of Sciences (Predict-2). Conflict of interest: none declared.