Characterization of the Pattern of Cognitive Impairment in Myotonic Dystrophy Type 1

Background  Central nervous system involvement occurs in most patients with myotonic dystrophy type 1 (DM1): mental retardation characterizes congenital forms, while a mild cognitive impairment has been described in adult patients with classic DM1. Neuropathological studies documented neurofibrillary tangles and an aberrant tau-protein expression in brain tissues of patients and animal models of DM1.

Objectives  To characterize the pattern of cognitive dysfunction occurring in DM1 and to analyze genotype-phenotype correlations in patients with DM1.

Methods  We assessed the results of a detailed neuropsychological study, including Mini-Mental State Examination, memory, linguistic level, praxis, attentional and frontal-executive tasks, in a group of 70 patients with DM1, including 10 congenital and 60 classic forms. Statistical analysis of data was performed using analysis of variance for multiple tests.

Results  Our study documented 2 distinct patterns of cognitive impairment in DM1: in particular, we confirmed the presence of a cognitive pattern characteristic of mental retardation in congenital cases, whereas in adult forms we documented an aging-related decline of frontal and temporal cognitive functions. No correlations were found between cognitive impairment and (CTG)_n in leukocytes or severity of muscle involvement.

Conclusions  Adult patients with DM1 frequently develop, with aging, a focal dementia: such findings agree with recent studies documenting an abnormal tau-protein expression in the brain tissues of patients with DM1. Cognitive decline may represent the only relevant clinical manifestation of DM1 in patients carrying very small (CTG)_n expansions in leukocytes.

Myotonic dystrophy (DM) is a multisystem disorder that affects, beside muscle, various tissues including the central nervous system. Myotonic dystrophy has been genetically associated with 2 distinct loci. An abnormal CTG expansion in the 3′ untranslated region of DMPK gene on chromosome 19q.13.3¹ is present in about 95% of cases. These forms are classified as “myotonic dystrophy type 1” (DM1).² Recently, a second form, named “myotonic dystrophy type 2” (DM2),² has been associated with an abnormal CCTG expansion in the first intron of the ZNF9 gene on chromosome 3q.³

In DM1 the pathological CTG expansion appears both meiotically and mitotically unstable, biased toward amplification. This explains both the “anticipation phenomenon” observed in DM1 pedigrees and the variable clinical expression among affected individuals.⁴

Analysis of genotype-phenotype correlations showed a significant correlation between (CTG)_n in leukocytes and the age of onset or the severity of muscle symptoms.⁴ The pathogenic mechanisms of DM1 are still debated.⁵ DMPK protein is only expressed in the muscle and the heart; therefore, haploinsufficiency of DMPK gene might contribute to the pathogenesis of muscular and cardiac damage.

The abnormal (CTG)_n expansion in the DMPK gene could also cause chromatin condensation, consequently affecting in cis the expression of 2 flanking genes, S1 × -5 and DMWD. This mechanism has been proposed for the pathogenesis of cataract, cognitive symptoms, and infertility in DM1.⁵ However, the most reliable hypothesis, supported by several research studies and by the evidence of phenotypic and molecular similarities between DM1 and DM2, suggests a toxic effect of the RNA containing abnormal CUG expansions. This RNA, assuming a double-stranded secondary structure, would accumulate in the nucleus sequestering CUG-binding proteins involved in the regulation of RNA-splicing mechanisms, ultimately affecting in trans the expression of other genes. The presence of a CNS involvement in DM1 has been observed since the first descriptions of the disease⁶: this can range from a condition of mental retardation, characteristic of congenital cases,⁶ to behavioral changes⁷ and “avoidant personality”⁸ frequent in adult forms.

In the early 1980s neuropsychological studies^9,10 documented a mild cognitive impairment, especially involving abstraction and visuospatial abilities. However, because of the limited number of patients included in the sample and the different test batteries used, their results did not show a definite pattern of cognitive impairment in DM.

Later on, neuroradiological studies showed nonspecific pathological findings such as ventricular enlargement, cerebral white matter lesions, or atrophy¹¹; neuropathological studies documented the presence of neurofibrillary tangles in the frontal and temporal lobes.¹² Recently, a specific abnormal pattern of tau protein expression was found in the brains of patients with DM1¹³ and transgenic mice.¹⁴

Following the genetic characterization of DM1, several studies analyzed genotype-phenotype correlations with specific regard to cognitive features, obtaining contrasting results.^15-17 Therefore, we carried out a detailed neuropsychological study on a large group of patients with DM1, to characterize the pattern of cognitive dysfunction occurring in DM1 and to analyze genotype-phenotype correlations.

Seventy patients with DM1 from 49 unrelated families were included in this study. They were selected from among 250 patients with DM1 followed up in the Institute Department of Neurology, Catholic University of Rome and at the Unione Italiana Lotta alla Distrofia Muscolare (UILDM) Sezione Laziale, Rome, Italy, using a multidisciplinary approach, also including cardiac and pneumologic evaluations: none of them had severe respiratory involvement, obstructive sleep apnea syndrome, hypersomnia, hypothyroidism, diabetes mellitus, hypertension, or evidence of cerebrovascular disorders. Selection criteria included a clinical psychiatric evaluation to rule out the presence of significant mood changes.

The group included 10 congenital forms (9 from maternal and 1 from paternal transmission) and 60 juvenile-adult onset forms (27 males and 33 females) (Table 1). Patients were arbitrarily divided into 4 groups according to the (CTG)_n detected in leukocytes (Table 1). The mean educational level did not significantly differ between the groups (Table 1), since even congenital cases could attend and complete primary school education with the assistance of a supporting teacher, as prescribed by the Italian educational program for mentally challenged children.

The severity of muscle weakness was assessed by the Myotonic Dystrophy Rating Scale.¹⁸ Genetic analysis for DM1 was performed using DNA extracted from leukocytes by a long-polymerase chain reaction protocol, as previously reported.¹⁹

The neuropsychological examination included Mini-Mental State Examination (MMSE) and a neuropsychological test battery exploring in detail all cognitive domains. Verbal abilities were also explored by 1 subtest of the Wechsler Adult Intelligence Scale specific for the evaluation of general linguistic knowledge. To exclude a bias due to muscle weakness, most tests did not require manual skills.

The test battery included the following: memory tests (Rey Auditory Verbal Learning Test and Forced Delayed Recognition,²⁰ Rey-Osterrieth figure recall,²⁰ Digit Span forward and backward), constructional praxis test (Copy of Rey-Osterrieth figure),²¹ language tests (Phonological Word Fluency and Semantic Word Fluency),²⁰ vocabulary subtest of the Wechsler Adult Intelligence Scale, level test (Raven Colored Progressive Matrices),²⁰ frontal and executive tasks (Temporal Rule Induction²² and Stroop Color-Word Interference test [shortened version]),²³ and a demanding test of visual attention (Dual Task [Trail Making B]).²³ Performances were considered either pathological or normal according to cutoff scores obtained by standardization studies, which also provided age and educational level–adjusted scores.***{xref ref-type="bibr" rid="REF-NOC40078-20 REF-NOC40078-21 REF-NOC40078-22 REF-NOC40078-23 REF-NOC40078-24"/>

Mean scores were reported for all groups of subjects divided according to either (CTG)_n or age classes. To avoid any confounding effect owing to the different mean ages of the 4 groups of patients with DM1, the number of pathological scores obtained according to the cut-off provided for age and educational level was computed for each patient on each neuropsychological test. Therefore, the group distribution of pathological performances on each neuropsychological test was analyzed using the Pearson χ² test. Data are reported as mean (SD) unless otherwise indicated.

A significant global cognitive impairment was documented in patients with DM1 belonging to the E1 group (Table 1), who showed lower scores in all long-term memory tasks and executive functions. To exclude a bias owing to the mean older age of patients from E1 group (Table 2), we not only performed an analysis of the rough scores, but also an analysis of the frequency distribution of pathological test scores in all groups studied, according to each patient’s age and educational level. Such analysis confirmed that a significantly higher percentage of patients belonging to E1 group were impaired in memory tasks, attention demanding–dual task, and tasks exploring frontal abilities (Table 3).

On the other hand, patients from the E4 group obtained significantly lower scores in tasks exploring verbal attainment as well as in level tasks exploring general intelligence (Table 2). Statistical significance of the data was further confirmed by the analysis of distribution of pathological scores, which also evidenced impairment in executive functions in E4 group patients (Table 3).

Patients with DM1 belonging to the E2 and E3 groups obtained pathological performances only at frontal and executive tasks, without a significant tendency to get lower scores for patients included in the E3 group (Table 2 and Table 3). To evaluate if performances obtained at some tests, including visuospatial and attention demanding–dual tasks, could be biased by muscle weakness, we analyzed correlations between the degree of muscle involvement assessed by the Myotonic Dystrophy Rating Scale and (CTG)_n in leukocytes. In fact, there was a strong positive correlation between genotype and severity of muscle involvement (Table 1). Although these data could partially account for the difficulties manifested with the dual tasks test by patients from the E4 group, they clearly indicated that the poor performances in executive functions and dual tasks obtained by patients belonging to the E1 group (Table 3), who generally had normal muscle strength, were actually due to the presence of a cognitive defect.

Our study was focused on the evaluation of cognitive abilities of a significant sample of patients with DM1: different from previous studies,^15-17 which included general intelligence tests (Wechsler Adult Intelligence Scale) and a gross evaluation of cognitive abilities (MMSE), we used a neuropsychological battery capable of exploring in detail every single cognitive function.

Results of our study highlight the occurrence of 2 distinct patterns of cognitive dysfunction in patients with DM1 (Table 2 and Table 3): patients from the E4 group (Table 1), which included all congenital forms, showed a severe impairment in all measures of general intelligence, verbal attainment, and frontal and executive abilities. Patients from the E2 and E3 groups were impaired only in frontal and executive tasks, with a trend toward lower scores in the E3 compared with the E2 group. Finally, the E1 group, including patients with the highest mean age and the smallest CTG expansions (Table 1), manifested a significant impairment in all measures of long-term verbal and visual memory.

Such data confirm that congenital forms invariably show a global cognitive impairment indicative of mental retardation, while classic forms with juvenile-adult onset as well as oligo or asymptomatic carriers frequently develop a cognitive decline, irrespective of the (CTG)_n harbored in leukocytes (Table 3); this is first characterized by a focal frontal dysfunction followed, with aging, by an impairment of temporal lobe functions.

Indeed, we detected an involvement of frontal functions in all adult patients with noncongenital DM1, while the presence of a mnemic defect suggestive of temporal lobe dysfunction was documented only in the E1 group (Table 3), which included the oldest patients with DM1, most of whom harbored small CTG expansion in leukocytes. (Table 1). These data were not influenced by the concurrence of mood or personality disturbances.

Our results agree with previous studies: Rubinsztein et al,¹⁶ studying 36 patients with DM1, have pointed out that adult forms harboring the smallest CTG expansions manifested as a group an impairment of memory function in a background of normal general intelligence. Wilson et al²⁵ described in a single patient affected by DM a cognitive decline resembling a focal dementia, initially involving frontal and executive and then temporal functions, suggesting that this pathology could be part of the clinical manifestations of DM. Neuropathological studies show neurofibrillary tangles in various brain regions of patients with DM1, especially in the hippocampus.¹² A specific aberrant expression of tau proteins has been recently documented in human DM1 brain¹³ as well as in DM1 transgenic models.¹⁴

Our study did not find a significant correlation between the (CTG)_n in leukocytes and the degree of cognitive impairment in adult-onset DM1 forms. This may be explained by the presence of somatic mosaicism in tissues of patients with DM1: to this regard, we may speculate that the (CTG)_n in leukocytes could conversely correlate with muscle involvement because these tissues would segregate similar genotypes because of common embryonic mesodermal origins.

Accordingly, Sergeant et al¹³ documented the presence of very large CTG expansions in brain tissues of several patients with DM1 characterized by different clinical severity and size of (CTG)_n in leukocytes. Interestingly, these authors also documented a somatic mosaicism in brain tissues of adult patients with DM1 that was instead absent in a congenital neonatal case, where one single very large expanded allele was detected.

Based on these data, we would like to speculate about the occurrence of different patterns of cognitive impairment in congenital vs adult patients with DM1: in particular, we may hypothesize that the early presence in brain tissues of congenital cases either of alleles containing very large CTG repeats or of their corresponding CUG RNAs could alter the expression of genes regulating later stages of brain development. Indeed, Westerlaken et al²⁶ recently documented that the expression of the DMWD protein, possibly involved in the pathogenesis of the central nervous system damage in DM1, is developmentally regulated in brains of mice models, increasing during early neonatal stages in neurons localized in central nervous system areas with high density of synaptic connections.

Conversely, the increased levels of very large CUG DMPK RNAs, caused by a progressive expansion of (CTG)_n trait in brain tissues of patients with adult-onset DM1, could specifically affect the correct splicing of tau proteins in neurons of frontotemporal cortex. According to this hypothesis, recent studies documented a progressive increase of CTG expansion, possibly related to specific tissue factors involved in DNA repair mechanisms, in neurons of DM1 transgenic mice.²⁷

We have characterized 2 distinct patterns of cognitive impairment in patients with congenital vs adult DM1. In particular, our data indicate that aged adult patients with DM1 frequently develop a focal dementia. Nevertheless, longitudinal neuropsychological studies would be useful to further confirm these findings. We would like to emphasize that, according to our data, the development of a significant cognitive impairment may be the only relevant clinical feature in patients with DM1 otherwise oligo or asymptomatic.

Correspondence: Camillo Marra, MD, PhD, Institute of Neurology, Catholic University of Rome, L. go F. Vito, 1 00168-Rome, Italy (cmarra@rm.unicatt.it).

Accepted for Publication: March 30, 2004.

Author Contributions:Study concept and design: Modoni, Silvestri, and Marra. Acquisition of data: Modoni and Pomponi. Analysis and interpretation of data: Modoni, Silvestri, Grazia Pomponi, and Marra. Drafting of the manuscript: Modoni, Silvestri, and Marra. Critical revision of the manuscript for important intellectual content: Mangiola and Tonali. Statistical analysis: Marra. Obtained funding: Modoni. Administrative, technical, and material support: Mangiola and Tonali. Obtained funding: Modoni. Study supervision: Silvestri and Marra.

Funding/Support: This study was partially supported by grants from the Italian Ministry of Scientific Research (MIUR COFIN 2003), Rome.

Acknowledgment: We thank Catherine Cullen for revising the English form of the manuscript.