Myotonic dystrophy type 2 (DM2) - Neuromuscular Disorders

Available online at www.sciencedirect.com

Neuromuscular Disorders 21 (2011) 443–450 www.elsevier.com/locate/nmd

Workshop report

Myotonic dystrophy type 2 (DM2) and related disorders Report of the 180th ENMC Workshop including guidelines on diagnostics and management 3–5 December 2010, Naarden, The Netherlands B. Udd ⇑, G. Meola, R. Krahe, D.G. Wansink, G. Bassez, W. Kress, B. Schoser, R. Moxley Neurology Department, Tampere University and University Hospital, Folkhalsan Institute of Genetics, Dept. of Medical Genetics and Haartman Institute, University of Helsinki and Vasa Central Hospital, Finland

1. Introduction Myotonic dystrophy is characterized by autosomal dominant progressive myopathy, myotonia and multiorgan involvement. Two distinct entities are currently known: Myotonic dystrophy type 1 (DM1, Steinert’s disease) [1], and Myotonic dystrophy type 2 (DM2) [2–4]. The first ENMC workshop on PROMM/DM2 in 1997 established clinical diagnostic criteria [5]. Before the second ENMC Workshop in 2000, Ranum et al. published mapping of the new genetic DM2 locus on chromosome 3q21 [6]. Subsequently the current nomenclature was adopted renaming the locus for myotonic dystrophy/Steinert’s disease as DM1 and the PROMM locus as DM2 [7]. Most of the previously reported families mapped to the DM2 locus [8], and the workshop adopted the term Myotonic dystrophy type 2 (DM2) for all the progressive myotonic multiorgan disorders linked to the DM2 locus. Prior to the 3rd workshop in 2003, the mutation underlying DM2 was discovered [9]. The mutation is a large (CCTG)n microsatellite repeat expansion in the first intron of ZNF9 gene on chromosome 3q21 and is thus similar to the untranslated (CTG)n repeat causing DM1. The 4th workshop reviewed epidemiology, evaluated molecular diagnostic DNA-tests, the range of clinical presentations and new findings on muscle biopsy, imaging and brain examinations in DM2 patients. The workshop discussed the molecular pathomechanisms based on the uncovered toxic RNA-gain-of-function with secondary effects on target genes and learned about the first DM2 mouse models [10]. ⇑ Corresponding author. Fax: +358 6 3232888.

E-mail address: Bjarne.Udd@pshp.fi (B. Udd). 0960-8966/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.nmd.2011.03.013

The clinical course of DM2 was considered to be usually more favourable compared to DM1. Families with DM2 do not have the severe congenital form of illness that occurs in DM1. Abnormalities in the social and cognitive abilities of adults with DM2 are typically mild or absent, and there is no prominent late weakness of the respiratory, facial and bulbar muscles [10]. In DM2 the manual skills largely remain intact and complications during general anesthesia have not been reported. However, DM2 has been reported with very severe variants of the disease: severe fatal cardiac complications have occurred as well as severe muscle weakness and disability [11]. The aims of this 5th ENMC Workshop on DM2 and related disorders were: (a) To evaluate the current DM2 diagnostic tests, their availability and their indications. (b) To learn more about the phenotype and its variations. (c) To review and update the Management guidelines from 2006. (d) To thoroughly review and discuss the current understanding of the molecular pathomechanisms. (e) To issue a statement on the position of DM2 related to ongoing preparations for therapy readiness in DM1. 2. DM2-mutation: Nature, molecular diagnosis and epidemiology Wolfram Kress from Wu¨rzburg, one of the major diagnostic centers for myotonic dystrophy in Germany, showed statistics indicating similar numbers of positive diagnostic

444

B. Udd et al. / Neuromuscular Disorders 21 (2011) 443–450

tests for DM2 and DM1. Both disorders are diagnosed in about 100 patients each year in the lab, with a positive hit rate of around 25% for samples referred for DM2 testing. Stan Vohanka from Brno reported on the first results of DM2 genetic testing in the Czech population. The total number of inhabitants in the Republic is about 10 million and currently two different laboratories carry out DNA testing for DM2. Since the start of DM2 genetic testing 87 patients have been diagnosed with DM1 and 152 patients with DM2 disease. Because clinical ascertainment for DM2 is still highly variable the numbers obtained suggest DM2 might be more frequent than DM1 in the Czech population. The overall clinical presentation of the DM2 confirmed patients was within the range of previously reported phenotypes with proximal lower limb weakness in all patients, muscle pains in 42%, and myotonia in 80%. They had also identified one family with a homozygote DM2 patient presenting with pronounced myotonia at age 20. A registry is established in the country providing more detailed information in the future. Bjarne Udd (Tampere, Finland) described recent population studies on the frequency of DM1 and DM2 mutations performed by Tiina Suominen in Finland. The basic cohort consisted of 4532 samples from healthy blood donors and was supplemented by additional cohorts of 827 patients diagnosed with other non-myotonic neuromuscular diseases and their 127 healthy relatives. The results show a surprisingly high frequency 1/1830 of the DM2 mutation in the Finnish population, which is >4-fold higher than any previous estimates of the prevalence of the disease prevalence for DM1 and DM2 combined [12]. The study raised many questions, such as how well the Finnish population may reflect the mutation frequency in other European populations and whether the mutation in fact is 100% penetrant in all circumstances. The results suggest a prevalence of more than 1000 patients in Finland, whereas to date only some 300 have been identified. 1.1. Genetic testing – Diagnostic methods Claudio Catalli from Rome, Italy gave a detailed description of the tetranucleotide repeat primed PCR (TP-PCR) method used in their laboratory for DM2 diagnostic purposes and the minor changes to the protocol they have adopted. The technique has been very reliable in the studied patients and individuals and is able to detect both short and very large expansions with no false negative results obtained so far. Charles Thornton (Rochester, NY) showed a new minimal Southern blotting technique for identification of DM1 and DM2 mutations which was developed in his laboratory to avoid radioactive steps and to make use of smaller sample sizes in the order of 100–500 ng of DNA. The higher sensitivity was obtained by applying multiple digestions closer to the repeat and use of an LNA-modified probe. The expansions usually appear as a smear and the

lower limit for detection of DM1 or DM2 repeat copies has not been established; DM1 in muscle has been detected down to 600 repeats. Whether other (CCTG)n repeats in the genome might be included in the assay has not been clarified. The sensitivity for small expansions might be enhanced by modifications of the protocol. Wolfram Kress reviewed the molecular genetic tests available and currently used for DM2 diagnostic and concluded that all different methods have their innate advantages and disadvantages. Any diagnostic laboratory should not rely on just one methodological approach since a marginal number of samples will be difficult to conclusively determine with any single method. Different versions of repeat primed PCRs are most widely used in combination with the first step PCR across the region containing the mutation to determine the sizes of alleles in the normal or premutation range.The laboratory should have access to complementary techniques such as modified Southerns, FISH or CISH, and even genotyping the haplotype of the mutant allele to clarify uncertain results. Muscle tissue DNA may provide a more reliable results if leukocyte DNA results remain uncertain. Richard Moxley (Rochester, NY) heads the initiative for a North-American registry of myotonic dystrophy aiming to provide a more detailed natural history of both DM1 and DM2. Currently 700 DM1 patients and 117 DM2 patients are registered. How well these numbers reflect the proportions of diagnosed DM1 and DM2 patients is not known. One of the already available findings based on the registry data is the wide range of previous diagnoses and the very long delay until final diagnosis in the DM2 patient group. This fact provides direct evidence that the diagnostic accuracy and awareness of DM2 among neurologists is still far too limited to the harm of the DM2 patients. 3. DM2 – The spectrum of clinical expression Giovanni Meola (Milan, Italy) updated the clinical and molecular information in a peculiar DM2 family with juvenile onset myotonia in the daughter of the proband mother. Both carry a verified DM2 mutation with larger expansion in muscle than leukocyte DNA. The mother has a common phenotype whereas the daughter shows clinical and EMG myotonia only. Histopathology in the muscle biopsy of the daughter is also within normal range despite the occurrence of ribonuclear foci in similar numbers and sizes as with the findings in the mother’s biopsy. Extensive studies of other genes causing myotonia such as CLCN1 and SCN4A have so far not identified a second gene defect accounting for the myotonia. Benedikt Schoser (Germany) reported on the experience in Munich based on a new cohort of 134 DM2 patients collected in a registry during the last years. The average age of onset was 40 years (±14 years) and the age of diagnosis 47 years (±14 years). The first symptoms consisted of proximal lower limb weakness in 40% of the patients, myalgia


stiffness and cramps in 20%, and others were referred for diagnostics because of elevated CK values or a positive family history. At diagnosis the range of symptoms was similar, however with the addition of findings in clinical examinations showing arrhythmia, cataracts, dysphagia, obesity, COPD, thyroid abnormalities, associated autoimmune diseases like psoriasis or rheumatoid arthritis, and GI cancers or constipation in 20–25% of the patients. Around 10% were eventually wheel-chair-bound and a minority needed respiratory devices or pacemakers (4% and 3%). The educational level of the diagnosed DM2 patients was in keeping with the normal distribution. As reported there is no increased risk of severe anesthetic complications (0.6% equals normal distribution) associated with DM2 disease. In the previous cohort of 480 patients collected by late Ken Ricker, there were four cases of sudden cardiac deaths, six males having heart transplantation because of cardiomyopathy before DM2 diagnosis, three cases of dilatied cardiomyopathy in the follow-up, and four strokes. The overall life expectancy was normal despite cardiac arrhythmia in 35%, coronary heart disease in 11%, myocardiac infarcts in 6% and pacemakers in 3%. Guillaume Bassez (Paris, France) complemented with data from the French registry showing that heart conduction defects were more prevalent in DM1 than DM2, even if the arrhythmia incidence was equal and strokes in DM2 patients frequently were associated with atrial fibrillation. Alide Tieleman from the Netherlands detailed the recent studies performed on the Dutch DM2 patient cohort regarding GI-abnormalities, dysphagia symptoms, sleep and other management problems usually less well recognized and addressed. Dysphagia was mild but occurred in more or less all patients. 24% used continuous pain medication. Despite experienced fatigue there was no day time sleepiness in DM2 patients. Sleep disturbances were usually associated with myalgic pains and caused increased sleep latency. A new subtype of DM2 related myopathy has been identified in a Finnish family showing late onset dominant myalgia cramps and mild proximal lower limb weakness but without myotonia even on EMG. The family was identified based on similar muscle biopsy pathology with frequent highly atrophic type 2 and nuclear clump fibers. Further molecular diagnostics showed just one normal sized allele over the mutation but no ribonuclear inclusions as antisense oligo FISH/CISH proved negative on muscle sections. Further investigation by Southern, repeat primed PCR and sequencing revealed a very short expansion of 55 uninterrupted repeats on a different microsatellite haplotype than all other Finnish DM2 patients, and apparently causing a mild myopathy without the abnormal splicing defect characteristic of DM2. 4. Molecular pathomechanism of DM2 Linda Bachinski (Houston, TX) started off the session by describing recent results on alternative splicing events

445

occurring also in other muscle diseases than the myotonic dystrophies. The group had investigated and compared Becker, Duchenne and Tibial muscular dystrophy samples with DM1 and DM2 samples, and observed alternative exon 5 splicing of MEF2A and MEF2C in all dystrophies [13]. Despite considerable methodological difficulties, the approach based on microarray screening and RT-PCR verification was able to identify a number of new genes showing splicing abnormality in DM1 and DM2 such as inclusion of exon 2 in MBNL2. The conclusion was that DM1 and DM2 are rather similar regarding splicing abnormality and that qualitative splicing differences seem not to account for phenotype differences of the two diseases. Nicolas Charlet-Berguerand (Illkirch, France) described their extensive studies on Amphiphysin (BIN1) missplicing and tubular defects in DM1 and DM2. Missplicing was not identical in unaffected vs. affected muscles in DM1 and apparently less in DM2 compared to DM1, nevertheless leading to secondary effects on tubulation in the muscle possibly over abnormal phosphoinositide signalling. The results were confirmed in mouse model studies and might be related to both the central nucleation and weaknesshypotonia of congenital DM1 and to the increase of internal nuclei in the adult myopathology of both DM2 and DM1. Charles Thornton summarized the current understanding of the splicing abnormality caused by mutant transcripts in myotonic dystrophies. The effects of many splicing events on the clinical phenotype may be moderate or difficult to show. In different models exon skipping events are easier to monitor correctly, whereas overexpression is always more difficult, as shown by the conflicting data on exon 11 inclusion in insulin receptor. The sequestered splicing factors in DM are involved in many other molecular processes and the phenotype of the MBNL1 KO-mouse is caused also by other dysregulations than splicing. The splicing events, however, serve as good biomarkers of the pathology and can be easily monitored. Lubov Timchenko (Houston, TX) discussed the observations of increased CUGBP1 in DM and any differences between DM1 and DM2. The increase in DM1 consists of both active and inactive forms particularly in congenital DM1 whereas in DM2 only the active form is increased. The regulation of CUGB1 activity involves phosphorylation at S302 and the active form binds eIF2 on polysomes affecting translation. If phosphorylated, eIF2 binds inactive CUGBP1 resulting in arrest of translation. Increase of inactive CUGBP1 may also be involved in delayed myogenesis in congenital DM1. Chantal Sellier (Illkirch, France) reported on new observations on FOX1 and 2, which bind (CCUG)n but not (CUG)n and compete with MBNL in binding to CCUG repeats. This may in the end affect the amount of free MBNL1 in DM2 vs. DM1. In the absence of Olayinka Raheem, Bjarne Udd related her recent results indicating an altered expression of ZNF9 in human DM2 muscle tissue. Previous studies, mostly on myoblast or lymphocyte culture and not on patient muscle

446


tissue, indicated that the mRNA and protein expression of the ZNF9 resident gene were not affected in DM2. However, new results using three different ZNF9 antibodies and a larger sample collection including DM1 muscle tissue for comparison show that there is a significant decrease of ZNF9 protein in DM2, based on inappropriate processing of mutant pre-mRNA transcript [14]. Andrew Link (Nashville, TN) introduced new results regarding ZNF9’s function in internal ribosome entry site (IRES)-mediated cap-independent translation and its role in DM2. ZNF9 binds directly to the IRES of the human ornithine decarboxlyase (ODC) mRNA and promotes cap-independent translation initiation. ZNF9 associates with actively translating human ribosomes as shown with polysome profiling analysis. Overexpression of ZNF9 increases IRES-mediated translation initiation in human cells. DM2 myoblasts have significantly reduced cap-independent translation, which was partially restored by overexpressing ZNF9 [15]. Finally, the S. cerevisiae gene Gis2 is a yeast ortholog of ZNF9, which will provide an opportunity to use genetic and molecular approaches to dissect the functional role of ZNF9. Ralf Krahe (Houston, Texas) summarized their extensive experience with expression profiling and pathway analysis of DM biopsies, indicating that the net effect of the DM2 mutation is more or less globally affecting transcriptional levels of many genes (2000) with 80% showing increased expression. Significantly dysregulated pathways included among others, the large field of Ca2+-handling genes, actin dynamics, upregulation of TGFB genes. Seventy-one of 386 genes specifically involved in RNA metabolism were dysregulated, with 80% showing increased expression. Exon specific array profiling for the assessment of splicing events showed dysregulation of 225 genes in DM1, 530 genes in DM2, and in a comparison between DM1 and DM2 181 dysregulated genes were identical between both diseases [16]. However, dysregulation also occurred in 256 genes shared with other non-DM neuromuscular diseases studied, apparently many of which are due to secondary compensatory mechanisms in disease. 5. Animals models to understand DM2 Ralf Krahe updated the findings in the transgenic mouse model of DM2 disease called DM2-HSAtg-121: ribonuclear foci of (CCUG)n transcripts are present in muscle mainly nuclear but some also in the cytoplasm. Expression was found also in brain and smooth muscle despite the skeletal muscle actin promotor used. The mouse shows a mild muscle pathology and also myotonia features on EMG despite normal CLC-1 and no abnormal splicing of the gene. MBNL1 co-localizes with the ribonuclear foci, with total protein levels unchanged, while CUGBP1 was increased. Abnormal splicing was neither observed for SERCA1, RYR1, ANKRD2 or TNNT3. Andreas Bergmann (Houston, TX) reported on a transgenic DM2-Drosophila fly model that he has developed in

collaboration with Ralf Krahe (Houston, TX). Expression of (CCTG)106 under control of the eye-specific GMR promoter causes ribonuclear and cytoplasmic foci, increased apoptosis, as well as irregular numbers and positions of photoreceptor neurons and cone cells. Inhibition of apoptosis using the Caspase3-inhibitor P35 reduces apoptosis and ameliorates the phenotype. Reduction of MBNL worsens and overexpression rescues the phenotype, whereas increasing or decreasing CUGBP1 does not make any difference. A screen for compounds having a beneficial effect on the phenotype has provided very promising results. 6. Therapy and practical management Rick Wansink (Nijmegen, NL) gave an overview of ongoing approaches towards molecular interventions in DM1 and their possible practicability in DM2. The Nijmegen lab focusses on complete elimination of toxic repeat transcripts in DM1 models. Powerful and stable AONs have been identified – most of them complementary to the (CUG)n expansion – that reduce products from the expanded DMPK allele and diminish ribonuclear foci in model systems without reducing much of the total DMPK protein. Apparently by binding to the repeat expansion, degradation of harmful expanded RNA is induced. Details of the underlying cellular pathway still need to be elucidated, however. Local AON delivery to muscle in DM1 mouse models has been completed with good results, but systemic treatment requires further optimization. The project is in good progress towards clinical Phase 1 trials in a few years time if all preclinical studies will develop fine. For DM2, related specific AONs directed at the CCUG expansion are easy to design and only few other genes with a long CCTG stretch exist in the human genome. Whether repeat-complementary AONs will be effective remains to be established, as the DM2 repeat segment is contained within an intron, which normally has a nuclear fate entirely different from that of mRNA. Small compounds that specifically bind C/CUG structures to counteract MBNL1 sequestration have been developed by other labs [17], but there remains some concern about their cellular toxicity. Charles Thornton continued on the same theme. Twenty years of experience with using AONs has provided a single licensed molecule for therapeutic usage and the combined experience sofar indicates overall little toxicity. However, by systemic i.v. delivery most of the molecules end up in the liver or kidney with much less in muscle tissue. Conjugates to increase uptake in muscle are searched for but even without such components the experience with tail vein morpholino injections in a mice model makes an effect on reduced myotonia in 4–5 weeks. Guillaume Bassez (Paris, France) presented a comprehensive overview of which components could be included as clinically meaningful, sensitive and robust outcome measures in any future therapeutic intervention trial for DM2. A favorable clinical outcome would need assessment of


pain and QoL by reliable questionnaire method, assessment of weakness probably with a DM2 specific scale despite new reliable devices used in DM1, and a long enough pretreatment natural history study along the established parameters used for sporadic IBM (six MWT, climbing four steps, rising from supine). MRI could possibly be used depending on the rate of change over time; correlation between weakness and T1 fatty degenerative changes in Tibialis Anterior muscle have been shown in DM1. Heart assessment could be similar to DM1 and conduction abnormality can be used as outcome measure. 7. Outcome of the initiatives from the previous workshop 1. Retrospective study of the outcome of general anesthesia in DM2 patients to overcome the lack of reliable and confirmed data. The German group of scientists have performed and published the study showing no increase of severe complications of general anesthesia in DM2 [18]. 2. A web site for general information and Guidelines was supported. This is still under consideration and will probably be set up as a link on the ENMC homepage with direct access to the published reports of the workshops. 3. Development of reliable clinical measures of myotonia. Assessment methods of myotonia have been published by the Rochester, NY group [19]. 4. Questionnaire instrument to describe and quantify pain in its specific locations. Chronic pain questionnaire tools have been applied in a Finnish study and will be reported. 5. Documentation of frequency of EKG and cardiac ultrasound abnormalities and heart attacks in DM2. This is part of the registry parameters in Germany, France and the USA. 6. Documentation of changes in CK and any toxicity during statin treatment. Studies on the increased statin adverse reaction susceptibility are initiated in Germany and in Finland. 7. Therapeutic trial readiness in DM2? Since DM2 usually does not end up with severe disability, dysphagia or respiratory insufficiency, the current baseline for trial readiness is to improve natural history and outcome measures, but any collection of patients will await positive outcome from trials in DM1 with shown efficacy in model systems of DM2.

8. Guidelines and recommendations 8.1. Guidelines #1: indications for genetic DM2 testing It is important to note that the overall clinical appearance of DM2 is different from DM1. DM2 needs to be con-

447

sidered in many other patients than those who have myotonia. In the single patient, none of the common clinical features: proximal weakness, myotonia, cataracts, elevated CK-values or established family history is absolutely mandatory for DM2 disease. – Based on clinical findings, the threshold for genetic testing should generally be lower in DM2 than the findings required for DM1 genetic testing because myotonia and cataracts are frequently absent. – One common clinical presentation is late or very late onset of mild proximal weakness and stiffness with or without myotonic findings on EMG. – Other presentations may involve prominent myalgia in the 30–50’s as the most disabling symptom associated with one or more other feature of muscle disease. – Findings typical for DM2 on muscle biopsy may be another indication for genetic testing [20,21]. – A considerable number of patients with a verified DM2 mutation have had very unusual phenotypes such as cramps, radiculopathic or chest pains, fluctuating weakness etc., which requests a high degree of awareness by the clinician regarding DM2 disease. – Presymptomatic genetic testing, as with other genetic disorders, requires appropriate genetic counselling. – The ENMC consortium provides a platform for exchanging expertise and quality control for lab diagnostics. 8.2. Guidelines #2: indications for genetic DM2 testing in fibromyalgia Screening everyone with fibromyalgia symptoms may not be practical despite the many overlapping symptoms fatigue, myofascial pain, stiffness, GI problems, etc. Fibromyalgia cannot explain many of the other symptoms that may occur in DM patients and these findings should alert towards testing: Check CK first! More and less important findings: Some family history demonstrating a dominant pattern of inheritance. Proximal muscle weakness. Myotonia, or even just increased insertional activity on EMG. Muscle atrophy or calf hypertrophy. Cardiac issues: sudden death, syncope and presyncope, cardiac conduction defects, cardiac arrhythmias. Endocrine disturbances including insulin resistance, diabetes, male hypogonadism. Elevated CK or liver enzymes, hypogammaglobulinemia. Cataracts: posterior subcapsular iridiscent.

448


Slight ptosis or facial weakness. Tremors. 8.3. Guidelines #3: Recommendations for management Clinical myotonia is rarely severe or disabling but since myotonic stiffness is mixing with myalgic pain a trial of mexiletine, and alternatively flecainide (2 50 – 2 100 mg) or propafenone (2 150–2 300 mg), can be considered. Baseline and serial monitoring of blood count, electrolytes, and liver gamma glutamyl transferase level and EKG are recommended. The consortium recommends serial monitoring of muscle weakness and suggests use of supervised aquatic therapy and land based resistive training to ameliorate fatigue and to maintain strength. No specific medication for muscle pain in DM2 has shown consistent benefit. Referral to a pain specialist may be necessary if completely disabling pain persists. Cataracts are treated by surgical removal as necessary. Cardiac fatalities have occurred, but the frequency is very low. However, prompt referral for cardiology consultation if EKG abnormalities or clinical signs suggestive of cardiac dysfunction occur. While manifest cardiomyopathy is rare, coronary heart disease in DM2 patients is not. Management of cardiac problems needs to be co-ordinated with the primary care physician and the cardiologist. Atherosclerosis and related hyperlipidemia are frequent in DM2 and seem to correlate closely with the insulin resistance that develops in patients. Serial monitoring of the lipid profile and intermittent assessment of oral glucose tolerance testing (including measurement of serum insulin values at 0, 30, 60, 120, 150, and 180 min time points after glucose ingestion). Statin treatment in patients with DM2 can pose a most challenging problem, because it is difficult in DM2 patients with myalgia to determine if they are developing a statin-related myopathy. Serial measurements of creatine kinase before and after treatment are helpful in deciding if statin toxicity has developed. Not all DM2 patients have pronounced or severe adverse reactions to statins, but if they develop, discontinuing treatment indefinitely is recommended. Moderate elevations of cholesterol (