Abstract
INTRODUCTION
Myotonic dystrophy type 1 (DM1) is the most common adult muscular dystrophy. It has been suggested that missplicing of various genes, secondary to expansion of an unstable trinucleotide repeat in the 3’ untranslated region of the DMPK gene, explains the multi-system pathophysiology [1, 2]. There is a correlation between the severity of disease and the number of repeats; patients with larger repeats have earlier onset of disease and more clinically severe forms. The repeat is unstable which gives rise to anticipation, with children having more severe disease than parents.
Besides muscle weakness and myotonia, the multi-systemic expression includes involvement of brain, heart, eyes, bowels and respiratory dysfunction. Breathing is affected both in sleep and wakefulness with many contributors: neuromuscular weakness causes nocturnal hypoventilation and lung function impairment, sleep fragmentation and daytime respiratory failure; central carbon dioxide insensitivity exists along with increased rates of respiratory infections and oro-facial dysmorphism [3, 4]. One study has shown a positive correlation between CTG repeat length in people with myotonic dystrophy and the degree of ventilatory failure [5].
Excessive daytime sleepiness is a common problem for people with myotonic dystrophy, affecting between 33 and 80% of people in studies [6–9]. This may be an intrinsic feature of the condition and a central nervous system pathophysiological mechanism is proposed [10]. It can also be caused by poor sleep patterns, obstructive sleep apnoea (OSA), respiratory failure, periodic limb movements during sleep, or narcolepsy features.
Treatments for sleep disorders, such as continuous positive airway pressure (CPAP) for OSA, non-invasive ventilation (NIV) for nocturnal hypoventilation associated with neuromuscular weakness and modafinil for excessive daytime sleepiness associated with narcolepsy are well-established, evidence-based treatments being used in other patient groups [11–15]. These have not all been rigorously studied in people with myotonic dystrophy, and treatment responses have not been quantified. Specifically, anthropometric features, or aspects of muscle impairment caused by myotonic dystrophy have not been studied in relation to any treatment response. There are no large series looking at outcomes from CPAP and NIV therapy, but two small studies showed beneficial results [16, 17]. Whilst modafinil has been trialled for daytime sleepiness, variable results have been found [18–21].
In this large cohort study, we performed detailed prospective overnight sleep and respiratory assessments of 120 people with myotonic dystrophy and daytime sleepiness and provided stratified treatment according to their study outcomes. Their treatment responses were assessed.
METHODS
From May 2011, all patients with DM1 seen at annual review by the John Walton Muscular Dystrophy Research Centre, part of the Northern Genetics Service, with an Epworth Sleepiness Score of greater than nine, and/ or with symptoms of excessive daytime sleepiness, were referred to a single centre for overnight assessment [22]. The DM1 diagnosis was based on clinical signs and family history, and confirmed by genetic testing.
The overnight assessment included: in-patient multi-channel sleep study with video, limb electromyography and respiratory polygraphy (Embla S4500), daytime arterial blood gas measurement, lying and standing spirometry (wedge bellows spirometer, Vitalograph), maximal respiratory pressures generated at the mouth after full inspiration and full expiration (PI max and PE max) (Micro Respiratory Pressure Meter, Micro-Medical Ltd) and Sniff nasal inspiratory pressure (SNIP). SNIP and PI max are non-invasive measures of inspiratory muscle function, and PE max of expiratory muscle function. A >15% supine fall in forced vital capacity was taken to indicate significant diaphragmatic weakness. On the overnight sleep study, a 4% or greater oxygen desaturation dip rate/index (ODI) or an apnoea-hypopnoea index (AHI) of five or more per hour are widely accepted as indicating mild sleep apnoea, with an ODI of 10–19 or AHI 15–29 per hour indicating moderate sleep apnoea and ODI≥20 or AHI≥30 per hour indicating severe sleep apnoea. All patients were reviewed by a sleep physician the morning after the study. A history of daytime sleepiness, sleep symptoms and sleep patterns was obtained. With these results, patients were allocated to one of four possible options in a pre-defined treatment algorithm: People with hypercapnia defined as PaCO2 > 6.2kPa, with nocturnal hypoventilation were offered nocturnal non-invasive ventilation (NIV). This was provided by the Newcastle Home Ventilation team as an out-patient after the sleep study. People found to have significant OSA with ODI 10 or more per hour (or AHI 15 or more per hour) were offered continuous positive airway pressure (CPAP) treatment. This was provided by the Newcastle Regional Sleep Service the day after the overnight study for the patient to take home. People found to have daytime sleepiness, with no other sleep disorder found on sleep study, were offered an appointment with a Sleep neurologist to complete sleep diaries and receive a prescription for modafinil, a wakefulness agent. People with normal sleep studies and/or with symptoms who did not want to pursue therapy were discharged back to the Northern Genetics Service.
Each patient receiving treatment was followed up in the appropriate clinic (for NIV, CPAP or modafinil) to assess treatment response. If a patient declined treatment according to their original group allocation, this was noted. Some patients transferred groups in order to receive a non-first line treatment (e.g. modafinil in someone declining CPAP for OSA). Data from routine annual review on the Muscular Impairment Rating Scale (MIRS) was collated [23]. The more symptoms and impairment a person has, the higher the MIRS score, with 1 being no muscular impairment, 2 minimal signs, 3 distal weakness, 4 mild-moderate proximal weakness, 5 severe proximal weakness.
Statistical analysis
Differences between groups were assessed with Analysis of Variance, ANOVA testing, with post hoc Bonferroni test for multiple comparisons. Independent t-test was used to compare means of responders and non-responders to treatment. A p value of <0.05 was considered to be statistically significant. Analysis was performed with SPSS version 22 (IBM Corp, USA).
RESULTS
There were 350 patients on the Northern Regional Genetics database with myotonic dystrophy type 1. During the 4 year period of this study (May 2011-May 2015), 140 met the criteria and were referred for sleep and ventilation assessments. Of these, 20 (14%) cancelled the assessment or did not attend, despite further appointments being offered. Their mean age was 48.4 (SD 14.4, range 22–72) and was not significantly different from those who attended. The data described therefore are for the 120 people studied. Of these, 10 were referred for repeat sleep assessments during the study after subsequent annual review by the Clinical Genetics service, which highlighted ongoing symptoms of daytime sleepiness; only two attended these. The data from their first assessment is included here.
Baseline characteristics
The baseline characteristics of the cohort are shown in Table 1. The Forced Vital Capacity was below 1 litre in three people. In 19 people (16%) FVC was less than 50% predicted. There was a >15% supine fall in FVC in 14% of people. One or more of PI max, PE max or SNIP were below 50% predicted in 87% of people. The Muscular Impairment Rating Scale (MIRS) was not available in 14 patients. In the remainder: none were grade 0 or 1, 5% were grade 2, 22% were grade 3, 57% were grade 4, 16% were grade 5.
Treatment groups according to sleep study data
According to the results of the sleep study and the pre-defined protocol, people were grouped into four outcome groups and offered treatment accordingly; Table 2 shows their baseline data. Figure 1 shows a flow chart of these patients.
The NIV group was significantly older than the no action group. The CPAP and NIV group had significantly higher BMI than the Modafinil group. The standing FVC was significantly lower in NIV group than the CPAP and Modafinil groups. The level of sleepiness (ESS) was similar in all three intervention groups, but significantly different to the no action group. This may have accounted for why they chose to not have any further treatment interventions, even though the mean AHI and ODI were abnormal. There were no significant differences in the MIRS score across these treatment groups. The scoring subsets did show a difference in wheelchair use, which was significant (p = 0.02) comparing the NIV group to the Modafinil group, and pacemaker or ICD insertion, which was significant (p = 0.007) comparing the NIV group to the CPAP group and the No action group. Vision problems scoring was significant (p = 0.01) comparing the Modafinil group to the no action group.
Treatment success according to baseline group
Respiratory failure and NIV
37% of those in whom this treatment was indicated have continued this with beneficial effects.
OSA and CPAP group
33% of those in whom this treatment was indicated have continued this with beneficial effects.
Normal study and modafinil group
78% of those in whom a sleep neurologist clinic review was indicated attended this. 33% had beneficial effects from modafinil. Two who received modafinil had both daytime sleepiness and delayed sleep phase syndrome and received melatonin also. For four of the patients who did not respond to modafinil, methyphenidate was commenced as second-line stimulant therapy trial, but none reported clinical improvement.
Transferred from original group in order to receive a non-first line treatment
Four people from the CPAP allocation group were keen to try modafinil after discontinuation of CPAP and therefore were seen for this. Of those using modafinil as a “second line” therapy, three found it beneficial and continued. Five people from the NIV group were also keen to try modafinil; two had not found NIV beneficial, three had opted not to start. Of these, two had no benefit from modafinil and the other three did not attend appointments made to receive a modafinil prescription.
Total treatment response
Of the 120 people with myotonic dystrophy who attended sleep and ventilation assessment, 35 (29%) had benefit from either first line or second line therapy. This was 25% of the total cohort referred. Comparing the means of treatment responders to non-responders showed no significant difference in any variable, except for ESS, 15.9 in responders vs.11.9 in non-responders, p < 0.0001.
DISCUSSION
This detailed, prospective cohort study of sleep and ventilation in people with myotonic dystrophy is the largest described in the literature. The data show a heterogeneous cohort, but 75% of the cohort had a spectrum of sleep and ventilation problems which warranted a targeted treatment intervention. Overall 29% of the total cohort studied gained benefit from CPAP, NIV or modafinil and continued these therapies. The only variable which differed significantly between the treatment-intervention responders and non-responders was excessive daytime sleepiness, measured by the ESS. This was a mean of 16 in responders and 12 in non-responders, but raised in both groups (normal is 9 or less). No other variable was identified as significant in predicting response to treatment. Clinically it would be helpful to identify any predictors for treatment success to help determine who to target for detailed assessments and interventions.
We would suggest that detailed initial assessment with a standard proforma to classify people with myotonic dystrophy into four different sleep-disturbance phenotypes is beneficial (Fig. 2). It allows specific targeted therapies which are more likely to be associated with treatment success than a “one size fits all” approach. Additional sleep disorders (restless legs syndrome, delayed sleep phase syndrome) can be detected, which may be a reason why ventilatory support or modafinil alone for some patients is less successful. Follow up to assess treatment response is also part of our model.
We think this stratified approach is superior to models in which patients are referred to CPAP, NIV or modafinil teams according to their predominant symptom, without single visit detailed physiological and sleep studies and clinician review. We also think this approach gives improved care to patients, is associated with high levels of satisfaction (postal patient satisfaction survey completed after 50 overnight studies showed 69% of respondents rated the assessment experience as “good”, 23% as “excellent”) and is potentially more cost effective than patients attending multiple visits to different departments. The 29% overall response rate however shows that further trials of different therapies are required in this patient group.
This data is similar to that described in a previous study, where 69% of a cohort of people with myotonic dystrophy (n = 71) had sleep apnoea, mostly obstructive [24]. The proportion of responders in our cohort is lower than would be expected in other patient groups using these therapies for their specific indications. For example, adherence to CPAP in patients with OSA selected from a sleep clinic is 81% at 5 years, with CPAP adherence predicted by the severity of sleep disordered breathing rather than sleepiness [25]. This compares to 33% adherence in this cohort. NIV is well established in some conditions of progressive neuromuscular weakness, such as Motor Neurone Disease and Duchenne muscular dystrophy, where it increases life expectancy and quality of life [26, 27]. In a cohort of 65 people using non-invasive ventilation (3% with neuromuscular disease), people used NIV for more than four hours a day on 97% of days [28]. In our cohort, 36% continued with NIV, but nightly adherence data is not available with the machines used. In a meta-analysis of modafinil in narcolepsy, a mean drop in ESS of 2.7 was reported, compared to placebo [14]. When using modafinil for other neurological causes of sleepiness, a lower and more variable treatment response was seen [29]. This again highlights the variable causes of sleep disturbance in those with myotonic dystrophy, as seen in our cohort.
It is unclear why these people with DM1 gained less benefit as a group overall. Many of them have significant comorbid disease which potentially made attending hospital assessments difficult. This may have led to the high initial non-attendance rate. Those who had sleep studies may not have been able to attend follow up and access support to adjust to the physical therapies of CPAP or NIV. Myotonic facies and pharynx may make the effective application of positive pressure to the upper airway via face mask harder or difficult to tolerate. They may have found the overnight sleep assessment demanding or found it geographically difficult to visit the regional centre and therefore made decisions not to proceed with further treatments which would necessitate travel. Individuals with congenital or infantile onset myotonic dystrophy may have intellectual impairment, so the standard delivery of information about CPAP or NIV used for other patients may not have been optimal for this cohort [30]. Specific information packages may need to be designed for people with myotonic dystrophy to ensure they are able to make informed decisions about their health choices.
A retrospective study of 42 patients with DM1 referred for NIV showed this was started in 25 people for daytime hypercapnia or hypoventilation symptoms [31]. Compliance was poor in seven (28%), associated with hypercapnia and inspiratory positive airway pressure setting, but overall NIV improved symptoms and nocturnal hypoventilation. We could find no studies in the literature regarding OSA or CPAP and myotonic dystrophy. One study identified the presence of OSA in 76% of a small cohort of 17 asymptomatic patients with myotonic dystrophy [32]. The authors recommend routine polysomnography for patients with myotonic dystrophy, but our data suggests this is unlikely to be helpful if patients do not have daytime sleepiness. In our cohort, it was noted that six people had abnormalities of awake breathing as well as in sleep, corresponding to central apnoeas (measured by air flow, chest/abdominal bands and video). This has been described previously [33]. Many aspects regarding ventilation in this condition are still poorly understood. In one study, the reduced ventilatory response to carbon dioxide in patients with myotonic dystrophy appeared independent of lung function impairment and respiratory muscle weakness, suggesting a central cause of carbon dioxide insensitivity [3]. Hypercapnia is likely to be multi-factorial, with central hypoventilation and altered carbon dioxide sensitivity appearing to be more stable, not implying the inevitable progression of respiratory failure found in other neuromuscular diseases like Duchenne Muscular Dystrophy [27, 34]. It could be postulated that hypercapnia is normal in some people with myotonic dystrophy, which might explain why NIV is not effective or tolerated. Multicentre randomised controlled trials of NIV are needed to guide practice in DM1, as there is little published about the optimal care of people with DM1 when they develop respiratory failure, or OSA and/or daytime sleepiness. Clearly some patients gain benefit from CPAP or NIV, as shown in our data (37% of those in who it was indicated) but without randomised controlled trials, whether patients receive this or not is likely to be variable.
Excessive daytime sleepiness can be disabling for patients with myotonic dystrophy. Modafinil is effective at treating this sleepiness in some and a UK national audit found 10–15% of patients used modafinil in the UK centres surveyed (personal communication). A registry-based survey among patients suggested a strong preference for continuation [35]. There have been randomised controlled studies of modafinil, but these have not been conclusive, possibly because of variable entry criteria into the studies [18, 36–38]. Future studies should seek to identify the characteristics of the “modafinil-responders” and conduct a randomised controlled trial of modafinil in them. In this cohort, there were 46 people in whom modafinil was indicated after other conditions had been excluded or specific treatment tried, of whom 33% responded. A meta-analysis of the previously conducted randomised controlled trials might allow more exploration and clarification of the features of this responder group. Careful assessment to treat comorbid sleep disorders may also help stratify responders. Although one previous publication suggested that methylphenidate was also effective for daytime sleepiness [39], we did not find it offered any clear advantage in people who had not responded to modafinil. Our numbers were small and methylphenidate was not used first-line.
Overall, the multi-systemic nature of DM1 results in different pathological causes for disordered breathing and sleep. This explains in part the difficulties in choosing the appropriate therapy for individual patients. Around a third of patients benefit clinically and continue on the appropriate intervention or medication.
CONCLUSIONS
This large study illustrates myotonic dystrophy is a heterogeneous disorder, with varying body mass, daytime sleepiness, muscular and respiratory impairment. A detailed assessment of sleep and ventilation is feasible and allows targeted therapy, which may be beneficial to some. Epworth Sleepiness Score is the strongest predictor of response. Further research is needed to establish which other features predict response so that appropriate treatments can be targeted to provide benefit to people with myotonic dystrophy.
FUNDING
This work was supported by the Newcastle upon Tyne Research and Capability Funding. The myotonic dystrophy registry in the UK has been funded by Muscular Dystrophy UK formerly the Muscular Dystrophy Campaign, and the Myotonic Dystrophy Support Group from 2012–2016, and is part of the TREAT-NMD Alliance (www.treat-nmd.eu).
Hanns Lochmüller receives funding from the Medical Research Council as part of the MRC Centre for Neuromuscular Diseases (reference G1002274, grant ID 98482), and by the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement No. 305444 (RD-Connect) and 305121 (Neuromics) and 305697 (OPTIMISTIC). Funding to support this work has also come from the National Institute for Health Research Rare Diseases Translational Research Collaboration (NIHR RD-TRC) for PHENODM1. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.
Footnotes
ACKNOWLEDGMENTS
We are grateful to Mrs Joan Furness, Regional Sleep Service administrator and the Lung Function team, Freeman Hospital for their input.