Psychiatric disorders in C9orf72 kindreds

Discussion

The present study has identified an increased rate of ASD, FTD, ALS, and other major psychiatric disorders in kindreds with the C9orf72 expansion compared with non-C9orf72 kindreds, with evidence that these conditions coaggregate within susceptible kindreds. We report that ASD occurs frequently in family members of patients with FTD and the C9orf72 expansion. The explanation for this association is as yet unknown. The effect of environmental stress should be considered, and ideally, future work should be extended to include genotyping of family members. We propose that familial psychiatric disorders should be considered when determining whether a proband is likely to carry the C9orf72 expansion.

The relationship between ASD and FTD, particularly in the presence of the C9orf72 expansion, warrants further investigation. As yet, no studies have directly compared these conditions, which may be partly related to the markedly different ages at onset. The triad of behavioral change in ASD is similar to that experienced by patients with FTD and includes impaired social interactions, impaired communication, and restricted and repetitive interests and activities.¹⁴ Theory of mind deficits are present in FTD, ALS with cognitive involvement, and ASD and may account for some of the abnormal behaviors in each condition.^15,16 It is plausible that the proportion of patients with a family history of ASD is underestimated, as none of the relatives identified with ASD were older than 15 years at the time of this study. Some of the older relatives may also have the syndrome, but it has simply not been formally recognized.

The discovery of psychotic symptoms in approximately 40% of C9orf72 carriers highlighted the potential link between FTD, ALS, and psychiatric disorders.^2,3 Furthermore, negative symptoms of schizophrenia and core behavioral features of FTD, ALS, and C9orf72 overlap, including poor personal hygiene, loss of volition, and reduced empathy. Although the cognitive difficulties experienced by patients with FTD and ALS are relatively homogenous and progressive (e.g., socioemotional, executive, and memory deficits), those in schizophrenia are variable, may be static, and are probably influenced by antipsychotic medications.¹⁷ Nonetheless, a similar pattern of cognitive dysfunction exists in patients with schizophrenia and FTD in that each group shows deficits in probabilistic learning, theory of mind, and sarcasm detection.^18,–,20 Theory of mind deficits in bvFTD occur early in the disease process and cause patients to lose insight, empathy, and their ability to sympathize.^15,21 Similarly, within schizophrenia, these deficits are pervasive and are associated with poor social functioning and outcome.²² Moreover, a wealth of evidence exists for memory and executive deficits in schizophrenia.²³

We were careful to include cases diagnosed as schizophrenia by a psychiatrist and not cases that the relatives reported to be “odd” or “mad.” We are therefore confident of the diagnosis in all cases labeled as schizophrenia. Of note is the fact that the age at onset of schizophrenia was very similar in C9orf72 and non-C9orf72 kindreds and fell within the typical age range of schizophrenia and that survival rates did not differ between C9orf72 and non-C9orf72 kindreds.²⁴ The majority of these cases remained in psychiatric care for many years with no suggestion of atypical features, all of which suggests that typical schizophrenia is more common in C9orf72 kindreds. It remains possible that the schizophrenia cases in C9orf72 kindreds are in some way atypical and would show neuropathologic and neuroimaging changes of the type seen in patients with the C9orf72 expansion manifesting as bvFTD. Further studies with detailed brain imaging are clearly required. Attempts have been made to determine whether patients with schizophrenia harbor the C9orf72 expansion, which has been found in some but not in many. These studies, however, have not specifically considered C9orf72 families. It is possible that results may have been diluted by a heterogeneous cohort of patients with schizophrenia or, as in one study, in a population in which C9orf72 expansions are rare.^25,–,28 Nonetheless, the evidence does not exist for the C9orf72 expansion as a cause of schizophrenia. Another possibility is that other genetic or epigenetic factors related to the C9orf72 expansion are responsible. As yet, these questions remain unanswered.

In contrast to family members with a diagnosis of schizophrenia, those with “psychosis” in the absence of a schizophrenia diagnosis were much older with a mean age of 50 years at the onset. It is possible that these individuals had a late-onset psychiatric illness or they may have experienced psychotic symptoms as the prodrome to bvFTD. Psychosis in relatives was associated with psychotic symptoms in the proband, and it suggests that a psychosis phenotype exists within C9orf72 FTD-ALS. Moreover, a positive association was found between psychotic probands and mental health disorders in their families, leading us to suggest that proneness to mental health disorders occurs in a subset of C9orf72 families.

The association between C9orf72 repeat expansions and mood disorders is less strong and difficult to disentangle in families with significant comorbidities. However, the results from this study suggest that mood disorders are also more common in C9orf72 kindreds than in non-C9orf72 kindreds. In fact, many patients with bvFTD experience depressive symptoms before the onset of their illness.²⁹ The existing literature is mixed, however, as to whether depression and anxiety are risk factors for dementia or merely appear in the prodromal phase.^30,–,32 A previous population-based study also found a link between suicide and C9orf72 ALS,⁶ which we have extended to include FTD. This association may reflect the consequences of poor mental health, which, in some cases, leads to suicide and highlights the central issue of appropriate genetic counseling and access to support services for family members of gene carriers.

This study has limitations. Although we have identified a higher rate of psychiatric disorders in kindreds of C9orf72 carriers compared with noncarriers, this is, however, not a population-based study, and we cannot compare the rates of psychiatric disorders with those found in the population. This raises the possibility that nongenetic FTD and ALS are protective against primary psychiatric disorders, whereas genetic FTD and ALS are not. This seems unlikely, particularly in view of recent reports of an increased prevalence of psychiatric disorders in patients with ALS years before the diagnosis of ALS and reports of depression occurring in the prodrome to FTD.^29,33 Nevertheless, we can draw conclusions in terms of comparison between C9orf72 expansion carriers and noncarriers, and specifically, this removes the potential environmental confounder of stress related to living with a family member with FTD or ALS. It was not possible, however, to address the issue of stress associated with being a member of a family with a strong history of hereditary neurodegenerative disease. Moreover, this study has focused on the C9orf72 expansion, given its high prevalence in this cohort; however, as yet, it is not clear whether other genetic mutations would also show an association with psychiatric disorders. If other genetic mutations were considered in future studies, this could serve as a control for the stress of living in a FTD/ALS genetic family and may confirm whether these stresses influence the development of psychiatric disease. Furthermore, controversy surrounds the C9orf72 expansion, as researchers debate the exact number of repeats that represent a pathogenic mutation, with some advocating a higher cutoff of up to 50. Other nongenetic factors may also be relevant, including socioeconomic status and environment, and should be considered for future studies. Recall bias is another confounder that should also be considered, as it is likely that family members in genetic families will be more aware of and more likely to research their family history in detail. Referral bias could also have influenced the results of this study. The research center has a reputation for an interest in familial FTD and ALS, which is reflected in the high proportion of C9orf72 carriers in the studied cohort. Ideally, we would have obtained genetic data on each family member within these cohorts; however, this was not feasible, given the large sample, and would have likely led to bias, as it is very unlikely that each member would have consented or been amenable to testing.

This study suggests that a link may exist between the FTD-ALS spectrum and psychiatric disorders, findings which potentially have important clinical implications, particularly in terms of genetic counseling. At a scientific level, these results indicate that close collaboration between neurology and psychiatry should be encouraged to better understand the shared features of these neurodegenerative and psychiatric disorders and determine whether they share other neurobiological or genetic features. Furthermore, given the overlap between these conditions, cross-disciplinary treatments could be considered in the future. Future work needs to determine an explanation for the findings of this study and, subsequently, the exact risk to family members of gene carriers. Ultimately, future work will need to establish whether the C9orf72 expansion has a role in increasing proneness to mental health disorders.