Reviewing hereditary connective tissue disorders: Proposals of harmonic medicolegal assessments

Clinical features and the major autopsy key points about MFS are shown in Table 1 [16‐24].

The pathologist who finds a spontaneous aortic dissection or aneurysms rupture especially in young individuals (< 60 years of age) needs to consider an underlying collagen-related diseases. Therefore, the practitioners should adopt a standardized postmortem protocol based on the collection of samples which can allow them to diagnose such disorders. Specifically, a full body forensic autopsy with an in-depth examination of aorta (Letulle method) including carotids, subclavian arteries, celiac artery, mesenteric, renal, and iliac arteries is recommended [25]. By adapting the recommendations of Sabatasso et al. [26], fixation and preservation of the whole aorta is preferable since it can be later examined by an expert cardiovascular pathologist. However, if retention of the whole aorta is not possible for any reasons, full thickness aortic samples should be taken from at least six different points (aortic annulus, ascending aorta, aortic arch, descending aorta, and abdominal aorta supra- and infrarenal). This approach is highly advisable since there may be local variations in the extent of the intimal and medial degeneration [26, 27]. According to de Boer et al. [25], comprehensive histological investigations with special staining methods (as reported in Table 1) of major organs, aorta, and major aortic branches are necessary to verify morphologically the presence of pathological alterations which can be therefore referred to HCTDs.

In these cases, the pathologists need to collect and store generous samples for future genetic analyses (i.e., blood → if absent psoas muscle, spleen, or kidney) [28]. If autopsy or extensive sampling for histological investigations cannot be performed, postmortem radiology may provide a feasible alternative to study the aorta, some of its major branches, and other organs [25]; accordingly, genetic samples could be taken from skin or nails [29, 30].

There are some MFS-like syndromes which share the high risk of sudden deaths related to spontaneous severe aortic events but bare different gene mutations. Among these, Loeys-Dietz syndrome (MIM # 609,192) shows mutations in other genes which are involved in the same pathway of FBN1 and include TGFBR1 (MIM * 190,181), TGFBR2 (MIM * 190,182), SMAD3 (MIM * 603,109), TGFB2 (MIM * 190,220), TGFB3 (MIM * 190,230) and SMAD2 (MIM * 601,366) [23, 25]. Clinical features such as cleft palate, bifid uvula, hypertelorism, and craniosynostosis are suggestive of Loeys-Dietz syndrome rather than MFS. Furthermore, Familial Thoracic Aortic aneurysm and dissection (FTAAD) and its variant with bicuspid aortic valve lack of marfanoid skeletal features, involve different genes (ACTA2 (MIM * 102,620), MYLK (MIM * 600,922), PRKG1 (MIM * 176,894), MYH11 (MIM * 160,745), MFAP5 (MIM * 601,103), MAT2A (MIM * 601,468) but show aortic enlargement and early onset of rupture by dissection or aneurysm formation [31], which is also more challenging for pathologists.

Still, a standardized postmortem protocol for spontaneous major aortic events in young individuals is recommended to guarantee equal diagnostic chances among living family members.

Finally, a very careful postmortem examination is needed in the event of a traumatic aortic or major arteries rupture/dissection among individuals suspected or suffering from MFS or MFS-like syndromes. Extensive photographic documentation and sampling with necessary histological investigations (lesion vitality) are required to assess the traumatic mechanism of death.

The relationship between sudden death and aortic wall rupture has been identified by several studies [32‐39]. Although known FBN1 mutations are spread throughout the gene, very few genotype–phenotype correlations exist in FBN1-related Marfan syndrome. The major existing correlations are presented in Table 1.

A recent review by Stengl et al. [39] reported exhaustively the evidence of strong genotype–phenotype correlations in MFS, highlighting the importance of improving the risk stratification of severe aortic events (TAADs) among family members. Nowadays, only few widely recognized correlations exist [39‐48]. Regarding the cardiovascular features, it has been documented that the missense mutations substituting a cysteine had a higher probability of ascending aortic dilation and mitral valve prolapse than mutations creating a cysteine [44]. Thus, not all missense mutations present the same risk of cardiovascular fatalities, and also raise the possibility that variants eliminating cysteine may cause malignant cardiovascular phenotype [40, 46, 47]. Such evidence may indeed suggest a crucial role of this amino acid in the protein structure of fibrillin-1.

Recently, the HTAD (Heritable Thoracic Aortic Diseases) Rare Disease Working Group have reported clinical criteria which identify cardiovascular conditions to be investigated with genetic testing [49].

By adapting these criteria in a pathological context (both clinical and forensic), we propose that postmortem genetic testing should be done in the following situations:

In cases where known pathogenic variants are found, then family members should be contacted to undergo genetic cascade screening.

Gago-Dìaz et al. [50] stressed the recommendation of postmortem genetic testing in sudden death cases due to thoracic aortic dissection. The authors first tried to demonstrate that incorporating the molecular diagnosis via massive parallel sequencing in TAAD autopsy cases is beneficial since these cases remain largely unexplored in the forensic field; among 17 cases, the authors identified two pathogenic variants, two likely-pathogenic ones, and six VUS. Although there are papers which focus on the importance of a molecular approach for sudden cardiac deaths among family members [51, 52], the research by Gago-Dìaz et al. (2017) was the first which analyzed its application on TAADs-related deaths from a pathological to a clinical-genetic setting [50].

However, this approach includes some disadvantages. The interpretation of VUS is more difficult when dealing with a frequently asymptomatic disease with an incomplete penetrance and variable expression – such as Marfan syndrome – which means that not all carriers of the causal mutation develop the clinical manifestations and that a wide range of clinical manifestations exists. To avoid issues of medicolegal liability and assess the best clinical management, Gago-Dìaz et al. proposed a practical way of proceeding [50]: a) identification of a putative causal mutation → b) support causality through in silico predictions of pathogenicity; → c) cascade screening; → d) functional studies so that to offer the segregation analysis to every willing family member.

This methodology provides one of the strongest types of evidence of causation, but it is unlikely to be feasible in a routine pathological context and this situation opens to the second problematic issue. Gene sequencing is seldom performed worldwide because of the lack of standard procedures existing for postmortem genetic analyses, high costs and necessary permission by the public prosecutor [35]. Thus, many prosecutors and pathologists currently consider a dissected aortic aneurysm as a definitive cause of death, without inquiring further into any underlying genetic disorders [36]. This status quo might miss the opportunity to screen at-risk relatives for hereditary diseases and establish appropriate preventative measures prior to complications.

According to Fellmann et al. [28], we propose a feasible methodology which can be applied in routine practice and depicted in the flowchart 1 (Fig. 3). Since it may take too long for the availability of the test results, family members can empirically start undergoing a precautionary cardiovascular follow-up.

Bild vergrößern

Important medicolegal claims could be raised among sudden deaths or severe aortic events of young adults suffering from MFS or Marfan-like disorders who were screened for athletic competitions and sport. [53]. In such cases, clinicians should follow specific recommendations for the detection of a possible cardiovascular history, which include key questions about exercise-related symptoms, a history of murmur or increased blood pressure, and a family history of premature death or disability from cardiovascular disease in a relative less than 50 years of age [54]. In addition, the Councils on Clinical Cardiology and Cardiovascular Disease in the Young recommend questions aiming to specific knowledge about the occurrence of Marfan syndrome [18]. Noteworthy, in an analysis of 134 athletes who died of cardiovascular causes, 115 athletes were found to have had a standard physical evaluation [55]. Of these, only four (3%) were suspected of having a cardiovascular disease, whereas one athlete (0.9%) was found to have a dilated aorta with a postmortem diagnosis of Marfan syndrome.

Therefore, clinicians need to report accurately anamnestic data and physical examination findings, knowing that further exams such as the echocardiography or the cardiac magnetic resonance are mandatory in the event of any cardiovascular abnormality (MFS is frequently associated to mitral valve prolapse) [56]. They do not necessarily provide significant results due to atypical variants with not specific physical features and without strong genotype–phenotype correlations but represent the best clinical practice to follow which is now available.

Clinical and genetic features about OI are shown in Table 2 [10‐12, 57, 58].

In clinical forensic medicine, child abuse should be evaluated by a multidisciplinary team, including pediatricians, radiologists, medicolegal consultants, geneticists, psychologists, and social workers [59, 60]. In this setting, it is crucial to collect anamnestic data regarding family members, and to perform blood test analyses, including complete blood cell count, serum calcium, phosphorus, alkaline phosphatase concentrations, and serum 25-hydroxy-vitamin D concentration in order to early identify potential metabolic disorders [61]. Therefore, physical examination shows a pivotal clinical role: a thorough skin exam looking for bruises and burns is indeed important when differentiating between OI and NAI (non-accidental injury) [62, 63]. Hence, children with bruises located on the back, ears, and genital area are more indicative of NAI [64]. Furthermore, from a forensic point of view, bruises showing markedly different patterns, shapes, and colors cannot be referred to the same traumatic episode.

The major radiologic features in OI and NAI are summarized in Table 2 [64‐70].

Kleinman [71] differentiated among fractures that appear to be:

Kleinman [71] and Ablin et al. [60] argued that the radiologic features of OI differ sufficiently from those of abuse to make the distinction straightforward in most cases. D’Eufemia et al. [67] also reported that rib fracture shows a predictive value of 95% for child abuse, especially in the posteromedial site and when occurring in children with less than three years of age. The characteristic mechanism that needs to be produced is a compression around the chest accompanied by the act of squeezing anteriorly and posteriorly the thorax. This type of fracture is extremely rare in patient with OI, and it occurred generally only in the most severe forms but on lateral sides.

A review by Pandya et al. [68] focused on which types of Osteogeneses Imperfecta were most confused with NAI. Specifically, they documented that from the studies reviewed, 25.5% (44 out of 172) of subjects with OI type IV (MIM # 166,220) were confused with NAI; conversely, 13.1% (77 out of 589) of subjects with other OI types were confused with NAI. The authors suggest that the difficulty in differentiating NAI from bone disease is complicated by marked variability in the manifestations of classic findings, as reported in Table 2.

Other essential features suggestive of NAI include intra-abdominal trauma resulting in duodenal, spleen or liver hematomas, traumatic brain/spinal cord injury, and bilateral retinal hemorrhages (both intraretinal and preretinal) [64, 72, 73]. For this reason, an ophthalmologic examination with optical coherence tomography and digital wide-field fundus photography could be beneficial since retinal hemorrhages represent the most common findings of abusive head trauma. Ophthalmologic consultation is recommended preferably within the first 24 h and ideally within 72 h [70]. Other rare ocular findings include retinoschisis, retinal folds, chorioretinal scars, and optic nerve sheath hemorrhages.

Unfortunately, from a medicolegal point of view, the literature which directly compares bone diseases with NAI is sparse, published between the 1980s and 1990s, and does not provide enough scientific evidence to differentiate OI from NAI. Furthermore, the methodology does not provide controls for bias confounding or chance since the papers were mainly descriptive in nature. For these reasons, clinical findings are at high risk of misclassification, where sometimes only the physicians’ own instincts make the diagnosis with potential forensic catastrophic consequences. However, genetic testing (typically molecular genetic testing of COL1A1/2 and IFITM5 (MIM * 614,757) genes) could be employed to screen potential victims of physical abuse [74]. Zarate et al. [75] conducted a retrospective study (2016) on 43 children with bone fractures in whom physical abuse was suspected and molecular testing for OI was performed. The research revealed only 2 pathogenic mutations on COL1A1/2 which were also associated to clinical features. The study was only based on the sequencing of type I collagen genes.

Noteworthy, at this time there are no consensus guidelines that identify the circumstances in which genetic testing for OI is indicated or when clinical and radiological evaluation alone can rule out the diagnosis of OI. As suggested by Pepin et al. [76], testing for OI can be completed by sequencing COL1A1, COL1A2, and IFITM5 simultaneously; if these analyses are negative, performing additional genetic testing to identify deletions or duplications (array-CGH or MLPA test). If these tests do not identify a pathogenic variant, in the absence of a strong clinical suspicion, no further genetic analyses are suggested. Genetic testing for the recessive form of OI is indicated only in the presence of other features, including consanguinity, prenatal or neonatal fractures with moderate to severe bone phenotype, congenital contractures, or specific clinical characteristics [76].

Genetic testing may pose uncertain results reporting variants of unknown significance [75‐77]. They are sequence variants of uncertain pathogenicity in genes known to cause inherited Mendelian diseases in which molecular genetic testing has a proven clinical validity and utility. The European Medical Genetics Quality Network’s best practice guidelines for the laboratory analysis of Osteogenesis Imperfecta recommend that VUS should be identified as unclassified until segregation within the family has been investigated and/or functional evidence becomes available [78]. This approach may be hard to carry out due to long-lasting follow-up or other practical issues. A recent study by Canter et al. [79] focused the attention on the clinical significance of VUS during an evaluation for potential child physical abuse, showing important forensic implications. In such conditions, the child with evidence of VUS may be moved into safe placement and then evaluated at 3 months (the outer limit of hard callus appearance), providing the best time window to document potential new or healing fractures [79]. Specifically, evidence of new fractures while the child is in foster care could suggest that there may exist genotype–phenotype correlations [76, 79]. Conversely, the absence of new fractures may indicate that the variant is not clinically relevant and a more diligent handling of the child. Promising future criteria include the possibility of performing proteomic studies to functionally investigate the collagen production, and the extension of genetic testing to the family members [77]. In this regard, OI is a disorder showing high penetrance, even within the same family, so that genetic results should always be compared to the family clinical history [79]. A proposal of a methodological approach to OI and child abuse is depicted in the flowchart 2 (Fig. 4). Finally, the identification of child abuse in individuals suffering from OI is very challenging, resulting in a topic almost totally neglected in the scientific literature [80].

Bild vergrößern

Prenatal diagnosis of OI is the result of the corroboration of many factors which include ultrasonography (US) pathological findings and prenatal specific genetic analysis [81].

Specifically, US examinations may reveal multiple intrauterine fractures, thoracic hypoplasia, bone demineralization, and shortening and deformity of long bones [82]. The more severe pathological features are found, the more OI type II (MIM # 166,210) (lethal form) is probable to diagnose. However, it is essential to highlight that prenatal evaluation sometimes leads to expectations that are not confirmed on postnatal evaluation, and that usually this is not automatically the result of a medical malpractice. Thus, as reported by Kidszun et al. [83], this way of differentiating lethal subtype from milder forms of OI is flawed deeply. In fact, OI subtypes were developed by using retrospective data such as radiograph findings, genetics, and clinical course. By these criteria, OI type II was diagnosed when the affected subject had died in utero or shortly after birth. The criteria were not developed to be predictive, and the sensitivity or specificity of different findings as predictors of prognosis has never been scientifically validated – which is also a recurrent issue in clinical forensic medicine with OI. Furthermore, these subtypes were created in 1979 without the consideration of currently available medical interventions which can deeply modify the prognosis.

Genetic testing results are also confounding and difficult to interpret since OI is a highly heterogeneous disorder in which reliable genotype–phenotype correlations are very sparse and contradictory in literature [83‐86]. Specifically, quantitative mutations (glycine substitutions and split site) of COL1A1 and COL1A2 are mainly responsible for OI type I (MIM # 166,200), whereas qualitative defects are significantly associated to OI types II–IV [85, 86]. Prenatal diagnosis and the prognostic value of OI based on US and genetic testing are very difficult because of the large number of skeletal dysplasia, their phenotypic variability, and overlapping features, which must be considered in the event of any medical claims. Overall, parents always need to be counseled very carefully and the limitations of prenatal diagnostic results need to be discussed in detail.

Clinical, pathological, and genetic features are shown in Table 3 [87‐99].

These syndromes are heterogenous and may show overlapping manifestations of MFS and OI with which they share similar medicolegal issues. EDS type IV (MIM * 601,468, vascular-type) shows poor average life expectancy due to spontaneous pneumothorax, pulmonary hemorrhage, aneurysms rupture, arterial dissection, and hollow organ rupture [87‐89]. Specifically, major thoracic and abdominal arteries are mostly involved in rupture and dissection [90]. Rupture of the uterus is more frequent within the last stages of pregnancy [91], whereas the sigmoid colon is the intestinal segment most affected by perforation [92]. The postmortem examination should be performed in harmony with the procedural approach that has been given for MFS and MFS-like syndromes (see “The approach of pathologists upon MFS and MFS-like syndromes”). However, this syndrome requires more caution rather than MFS since it may show spontaneous non-aortic artery rupture/dissection [90, 95, 96] and hollow organ rupture. Furthermore, histology findings are often less specific than the other vascular syndromes [90] and may include a) reduced dermal skin thickness; b) thin collagen bundles with large interstitial spaces in the reticular dermis; c) increased thickness of cutaneous elastic fibers; d) thinning of the adventitia of the aortic wall [98]. For those purposes, the pathologist should collect the whole organ/artery and specimens of skin for the histologic examination with samples for future genetic analyses. A proposal of a feasible methodology which can be applied in routine practice is depicted in the flowchart 3 (Fig. 5).

Bild vergrößern

EDS can be considered in the differential diagnosis of child maltreatment or abuse since injuries due to minor trauma could mimic typical maltreatment injuries. Subjects suffering from EDS have multifaceted skin manifestations due to hyperextensibility and fragility of the skin with increased bruising and atrophic scar formation [97, 99, 100]. Accidental injuries are primarily located over the knees, anterior tibial area, and areas with underlying bony prominences with small and single bruising [101‐103]. As reported for OI, the main aim of practitioners in such situations is to differentiate between an accidental trauma versus a NAI [104], which should be evaluated in a multidisciplinary setting.