Assessment of respiratory function and need for non-invasive ventilation in a cohort of patients with myotonic dystrophy type 1 followed at one single expert centre

Introduction: Respiratory insufficiency is one of the main causes of death in myotonic dystrophy type 1 (DM1). Although there is general consensus these patients have a restrictive ventilatory pattern, hypoventilation, chronic hypercapnia and sleep disturbances, the prevalence of respiratory disease and indication for and effects of non-invasive ventilation (NIV) need to be further explored. Objectives: We aim to describe the respiratory function and the need for NIV at baseline and over time in a cohort of adult patients with DM1. Methods: One hundred and fifty-one adult patients with DM1 were subjected to arterial blood gas analysis, sitting and supine forced vital capacity (FVC), peak cough expiratory flow (PCEF), nocturnal oximetry, maximal inspiratory


| INTRODUCTION
Myotonic dystrophy type 1 (DM1) is a multisystem disorder characterized by multiple organ involvement. 1 Respiratory failure, together with cardiac involvement, is the main cause of death in adult-onset DM1 and is associated with increased morbidity and poor quality of life perception. [2][3][4] Respiratory function in DM1 patients is characterized by progressive inspiratory and expiratory muscle weakness which causes a restrictive respiratory syndrome. 4,5 Several authors have specifically addressed respiratory dysfunction in DM1. 6 Most studies conclude that there is evidence that older age at onset and overweight are negative predictors of lung capacity and respiratory function. 7,8 Less clear is the relationship to CTG size: in some patients, large expansions correlate to a decrease of vital capacity over time 9 or to expiratory muscle strength and oxygen saturation, 10 whereas in some others these values are reported as independent predictors of ventilator support. 11 Muscles of the trunk also play a role, and especially trunk flexor weakness has been associated with lower lung volumes. 12 Although respiratory protocols for respiratory muscle tests are well defined 13,14 as are respiratory management protocols for patients with neuromuscular conditions in general, 15,16 the specific approach to respiratory treatment in DM1 is still challenging and management protocols for respiratory care in these patients vary. 4 This may be due to several reasons. Firstly, there are limited data on the natural history of changes in pulmonary function over time in patients with DM1. 4,10 Secondly, the clinical presentation of respiratory involvement varies between patients: in some patients, the muscle weakness may predominate, whereas in the majority, it is the CNS component that may present as sleep-related breathing disorders including central apnoeas 17 and excessive daytime sleepiness, which are independent of breathing issues and may not be alleviated by respiratory interventions such as CPAP, BPAP or respiratory muscle training. Lastly, although there are several protocols recommending to use non-invasive ventilation (NIV) to correct for chronic respiratory insufficiency in DM1, 4,18 the criteria to launch NIV in DM1 are variably applied, the effects of NIV on clinical symptoms vary, [19][20][21] and overall patients' adherence is limited. 4,11,19 We present data on respiratory function in a large cohort of adult patients with DM1 followed at a single dedicated multidisciplinary neuromuscular centre to analyse the prevalence of respiratory disease at baseline and over time and the indication for and effects of noninvasive ventilation (NIV).

| Patient population
Demographical and medical records of 151 adult patients with genetically confirmed DM1 having at least one neurological and respiratory assessment were retrospectively reviewed from 2009 to 2019 at the NEMO Clinical Center in Milan, a specialized multidisciplinary clinic for patients with neuromuscular disorders and specifically with DM1. No a priori selection based on respiratory symptoms or signs was taken into account to enroll patients except that patients who were already on NIV (n = 24) were not included in the study cohort ( Figure 1). Patients were seen as out-patients or in-patients at the NEMO Clinical Center. As out-patients, patients are seen every 6 months in a 1-day visit during which patients are subjected to neuromotor function tests, respiratory assessments and psychological interviews. According to the patients' needs additional nutritional and swallowing studies, cardiac monitoring and orthosis assessments may be included. As in-patients, patients are usually admitted for 2-3 week neurorehabilitation programmes, including respiratory assessments, as part of their care, and never during acute illnesses, of any kind.
Patients with a clear congenital (symptom onset within 1 month from birth) or paediatric onset (prior to 18 years of age) were excluded. Disease onset was taken as the age at which symptoms clearly consistent with DM1 (myotonia, muscle weakness, and cataracts) were documented in the charts. Patients with no signs or symptoms of myotonic dystrophy including no referred muscle weakness nor myotonia and a genetic diagnosis of DM1 were also included. It is common experience that weakness and myotonia may be present when examined and this is the reason for this inclusion criteria. Only 2 FERRARI AGGRADI ET AL.

R e t r a c t e d
patients with a clinically stable condition and especially no referral for acute pulmonary disease within the month prior to admission at the site were included.

| Respiratory assessments
All patients were subjected to a complete battery of respiratory assessments at baseline (n = 151) and after a median of 3.85 years (n = 116). All pulmonary function tests (PFT) were carried out by the respiratory physiotherapists in compliance with the American Thoracic Society/European Respiratory Society guidelines 13,22,23 and the percentages of predicted values (%), age, weight and height were recorded by the respiratory physiotherapist in occasion of every admission for follow up. The PFTs were measured using a Cosmed Quark PFT (Cosmed srl Rome -Italy) spirometer and the decision whether to use the mouthpiece with nasal clip or the mask was based on the evaluation of each single patient, specifically on the eventual presence of facial muscle weakness and potential air leaks. The PFTs included were the following tests: • FVC in the seated and supine positions (FVC and sFVC, respectively): expressed both as absolute value, in litres, and as a percentage of the predicted value (FVC% and sFVC%, respectively) 22 All respiratory tests were then manually reviewed by a pulmonologist with expertise in neuromuscular disorders and especially in DM1. Nocturnal cardiorespiratory monitoring was performed according to standard procedures, 24 using the device Nox T3 (Nox Medical, Reykjavík Iceland, complete monitoring, eight channels), inclusive of a wrist saturimeter with a finger clip tip, either at patient's home the night before the visit (out-patients) or at the centre itself (in-patients). The data considered relevant for clinical interpretation were the mean oxygen saturation (SpO 2 ), the time spent with oxygen saturation <90% expressed as a percentage (T90), the oxygen desaturation index (ODI) during sleep and the apnoea-hypopnea index (AHI). The exam was considered abnormal either if the ODI was >15 events/h, the AHI exceeded 5/h, a sustained oxygen desaturation (SpO 2 ) ≤ 88% for five

| Excessive daytime sleepiness assessment
Excessive daytime sleepiness was self-reported using the Epworth Sleepiness Scale 25 (ESS). Sleepy patients were considered as such if scores were >10. This was available in 76 of 151 charts only.

| Neuromotor assessments
Muscle disease severity was rated using the Muscle Impairment Rating Scale (MIRS), a ordinal 5-point rating scale, able to track the clinically recognized distal to proximal progression of the muscular involvement in DM1, partly implementing the manual muscle testing of 11 muscle groups. 26 Additionally, the distance performed during the 6-m walk test (6MWT) and the pre-and post-test dyspnoea scales were collected for ambulatory patients.

| Coping strategy assessment
The COPE-25 questionnaire 27 was administered to see whether coping strategies could at least in part come into play in determining adherence to management protocols in respiratory care.

| Statistical analysis
Shapiro-Wilk test and Levene's test were used to assess the normality of the distribution and the homogeneity of the variance. Data are summarized as median and interquartile ranges for continuous variables, and as numbers and percentages for categorical ones. The non-parametric Mann-Whitney U test and the χ 2 /Fisher's exact test were used to compare patients without NIV indication and patients with NIV indication in terms of continuous and categorical demographic, clinical and respiratory features, respectively.
The correlations between respiratory features and demographic and clinic characteristics were investigated using the Spearman's rank correlation coefficient.
To standardize the trend of decline between groups over time, only patients who attended at least two visits within 6 years were considered. The decline for each variable was calculated for each patient from the time of the first recording to the last recording, assuming a linear progression. The average annual decline for each variable was then calculated and compared between patients without NIV indication and patients with NIV indication at baseline using the nonparametric using the Mann-Whitney U test. The same analysis was also used to compare the average annual decline between NIV non-adherent and NIV adherent patients.
The Mann-Whitney U test was also used to compare the COPE-25 scores between patients without NIV indication and patients with NIV indication at baseline, and between NIV non-adherent and NIV adherent patients.
Finally, all the comparisons between patients with and without follow-up data were investigated using the χ 2 /Fisher's exact test when the IVs were categorical, and the Mann-Whitney U test when the IVs were continuous.
All tests were two tailed and a p-value <0.05 was considered as statistically significant. All the analyses were performed using SAS 9.3 (SAS Institute Inc.) 3 | RESULTS

| Demographic and clinical features at baseline
Demographic and clinical details are provided in Table 1. Both men and women were equally distributed (56% women) with BMI values ranging from normal to overweight ranges. The majority were mildly to 4 FERRARI AGGRADI ET AL.

R e t r a c t e d
moderately affected patients (MIRS 3 and 4), with CTG ranging from moderate to large expansions and with, at least, a 6-year history of disease. Only four patients were sitters, whereas 82 of 103 (79.6%) patients in whom the 6MWT was available in their charts were able to walk for more than 300 m. Table 2 describes the general respiratory features at baseline of the entire cohort. Sitting FVC was below normal (≤80%) in about half of our patients (77 of 151) and 67 (44.4%) had an indication for NIV at baseline according to one or more of the aforementioned criteria. Demographic and general clinical features were compared between patients having normal respiratory baseline assessments (n = 84, cohort 1 in Table 3) and patients with indication to use NIV at night (n = 67, cohort 2 in Table 3). Results are summarized in Table 3.

| General respiratory features and NIV indication at baseline
Significant older age, more severe muscular impairment and longer disease duration characterized the patients who received an indication to start NIV. This was in line with data of patients who were already on NIV on first assessment at our site (data not shown and not considered in this analysis). Three of 84 patients (3.6%) with no indication for NIV at baseline, 10 of 67 (14.9%) receiving an indication to use NIV had a PM or an ICD. Table 4 specifically compares the respiratory features of patients with no indication for NIV at baseline and of those  Table S1. Obstructive sleep apnoeas (OSA) were recorded in a minority (4.6%). Excessive daytime sleepiness was reported by 24.2% of patients (n = 16 of 66 available ESS scores). Of these 16 patients, nine patients reported daytime sleepiness, none reported morning headaches or snoring during the collection of medical history specifically addressing respiratory symptoms, one patient reported both dyspnoea at rest and exertional dyspnoea, and two patients reported subjective fatigue, with some symptoms overlapping in some patients.
Twenty-three of 151 patients (15.2%) fulfilled criteria for secretion management support (air-stacking or cough assist) based on Peak Cough Expiratory Flow (PCEF) ≤ 270 L/min, as suggested in the most recent consensus on respiratory management of DM1 patients. 18 Respiratory and demographic details for each patient are provided in Table S1.

| Follow-up of the whole cohort
The median follow-up period between first and last recording was 44.90 months [26.32-59.20 months, 3.74 years]. Thirty-five patients of 151 were lost to follow-up as indicated in Figure 1. These were patients mainly having a MIRS of 2. No other demographic features distinguished these patients lost to follow-up compared with the rest of the cohort. Considering respiratory and motor assessments, patients lost to follow-up reported a significantly higher ODI, had a significantly lower pH, FEV1/FVC ratio, PCEF, MEP and the distance they were able to walk during the 6MWT was significantly less compared with the rest of the cohort.
Eleven of the 116 patients (9.5%) available for followup lost their ability to walk. None developed cardiac symptoms or signs requiring pace-maker or implantable cardiac defibrillators of other cardiac pharmacological treatment during our observational period.  R e t r a c t e d Nine of the 59 patients (15.3%) who had normal respiratory tests and pulmonary assessments at baseline fulfilled criteria for NIV indication at night on follow-up. No specific clinical and demographic features characterized these nine patients, who were similar to the remaining patients maintaining normal respiratory function during the same observational period. NIV was prescribed in these patients due to abnormalities registered in those parameters that the latest guidelines and patients. Additionally, being the overall respiratory status of each subject analysed by a pulmonologist specialized in DM1, two patients whose respiratory assessments were considered 'borderline' were initiated to NIV in a more proactive and preventive way. Respiratory details of these nine patients are provided in Table S1.

| NIV effects and adherence
Mean SpO 2 and ODI on oxymetry and pCO 2 on ABG improved in the patients who started NIV based on symptoms and respiratory test results (cohort 2, Table 6). Regarding the min SpO 2 , pO 2 and HCO 3 no significant differences were observed among patients who had indication for NIV and those with normal respiratory function on first assessment. Only 17 of 57 with indication for NIV however used their ventilator regularly. Comparing the results between the 17 adherent and the 40 non adherent patients in cohort 2 (fulfilling criteria for NIV at baseline) we observed that there was a statistically significant improvement in both min SpO 2 and pO 2 in the adherent cohort (Table 7). General clinical features including sex, age, disease duration, BMI and MIRS scores were similar between the two groups. Details of the correlations between respiratory parameters and demographic and clinical features are provided in Table S2. Follow-up data prior and after NIV use was available in only seven patients of the nine who had shifted from normal respiratory parameters to NIV at night. When we compared the first available evaluation of patients who shifted from normal respiratory parameters to NIV during the follow-up period with the respiratory parameters of the group of patients who reported a normal respiratory function during the entire follow-up period, the patients requiring NIV proved to have a significantly longer disease duration (18. [7.44-7.49], p = 0.0194; respectively). The nine patients who shifted to NIV reported a feeling of well-being and reduced tiredness on clinical history which was the reason why they were able to use their ventilators at night continuously for 4 h or more when we discussed the results with these patients.
No difference was observed in the coping strategies assessed by the COPE-25 questionnaire (seeking social support for instrumental reasons, behavioural disengagement, positive reframing, active coping, turning to religion) between NIV-adherent and non-adherent patients.

R e t r a c t e d 4 | DISCUSSION
Our results provide additional data on respiratory parameters in a large cohort of well-characterized adult patients with DM1, and suggest that respiratory function, in general, may have a slow rate of progression. Only a minority of our patients developed the need for NIV during the observational time frame of the study. Our results also confirm that respiratory dysfunction is frequent in DM1, that there are distinctive demographic and clinical features in patients requiring NIV compared with those with normal respiratory function at baseline, that referral for typical respiratory symptoms like shortness of breath or orthopnoea is unusual and that adherence to NIV is a limiting factor.
There is general consensus that DM1 is a slowly progressive disorder, 1-4 but the rate of decline of the multiple organs involved still needs to be defined. During our observational period, respiratory parameters overall did not show a clinically meaningful change in the cohort taken as a whole, in line with other reports. 5,6 Only 9 of 84 (11%) patients who had normal respiratory parameters at baseline (cohort 1) developed the need for NIV at night during our observational period. We could not detect any distinctive clinical or respiratory features in these nine patients that could help to predict a potential break-point between normal and dysfunctional ventilation in patients with DM1. Unfortunately, the follow-up period and sample size did not allow to appreciate any significant improvement or delay in progression.  Note: Pre-and post-test dyspnoea were calculated using the modified dyspnoea Borg scale. This scale asks the patient to rate the difficulty of breathing from 0 (no difficulty) to 10 (maximal difficulty) at the very beginning of the 6MWT and immediately after. Abbreviations: IQR, interquartile range; SpO 2 , oxygen saturation; ODI, oxygen desaturation index; pH, potential of hydrogen; pCO 2 , partial pressure of carbon dioxide; pO 2 , partial pressure of oxygen; HCO 3 À, bicarbonate; FVC, forced vital capacity; FEV1, forced expiratory volume in 1 s; PCEF, peak cough expiratory flow; MIP, maximal inspiratory pressure; MEP, maximal expiratory pressure; 6MWT, 6-min walking test; NIV, non-invasive ventilation.

R e t r a c t e d
In line with previous reports describing the restrictive pattern of respiratory involvement and the predictive factors of respiratory involvement in this disease, 6 we provide additional data on the prevalence of respiratory disease in DM1. The majority of the patients in our cohort were mildly to moderately affected, more than 60% could walk for more than 300 m in 6 min and only 11 were sitters. Yet half of the patients when first assessed at baseline had an FVC below normal and had an indication for nocturnal non-invasive support. The patients with an indication for NIV were significantly older, had a longer disease duration, were more severely disabled (higher MIRS scores) and had higher BMI scores compared with the ones with normal respiratory function (cohort 1), supporting previous evidence. 6 Of note, that none of our patients, including those with FVC below normal, referred to our site for shortness of breath or orthopnoea. Less than a quarter reported excessive daytime sleepiness on the Epworth Sleepiness Scale, but this was not a reason for referral, nor was choking in those patients with weak peak cough flow, which is usually associated with secretion management difficulties. This is not unexpected considering the passive attitudes and apathy of many patients with DM1, 28,29 but it also emphasizes the complexity of respiratory dysfunction in this disorder, where central nervous system related problems like sleep apnoeas, sleep-related breathing disorders and central fatigue and sleepiness play a predominant role. 30,31 When we consider the changes over time in the patients who received an indication to start NIV at night at baseline (cohort 2), we can observe that, as expected,  R e t r a c t e d respiratory parameters improved. Patients reported a general feeling of well-being with no specific referral to respiratory symptoms who were never the cause for referral nor the main and only reason to start NIV. This is in contrast with the reports in 12 patients on and off NIV after 1 month 32 in whom no changes in respiratory parameters were observed and the patients experienced very little benefit in symptoms or quality of life. The observational period in the report of the 12 patients was however very short (1 month). In fact, Boussaid and coworkers 8 describe an increase in the risk of major events (invasive ventilation and death) in a large cohort of patients who should have used NIV at night in a 10-year observational period. In our cohort of patients who regularly used NIV, pCO 2 increased. We can speculate that this rise may be due to a more severely impaired central drive for hypercapnia and hypoventilation control, which progresses despite the strength of the respiratory pump, which is what NIV mainly acts on. Despite improvement in respiratory parameters, adherence to NIV prescription was limited to a minority of patients in agreement with previous reports and clinical experience. 4,8 We can assume that the limited compliance was not due to the type of ventilator used, the type of interface or the leaks because several attempts were made by therapists and pulmonologists with longlasting experience in neuromuscular disorders and in DM1, and an in-patient setting was also provided for the most difficult cases. We were unfortunately unable to detect specific distinctive features between adherent and non-adherent patients. 4,7,33 Considering previous studies describing major events in the long-term for patients who should have used NIV but were not adherent to this prescription, 8 we recommend caution in amending the current recommendations to initiate NIV in these patients. Further studies are needed to determine whether additional strategies (home care to monitor NIV use where specialized nurses or PTs could also play an educational role on improving patients sleep-wake cycle, refresh training and provide more explicit and accessible information on the effects of NIV adherence) could actually improve adherence in this patient population.
No quality of life data and disease burden perception were captured in this study and this represents a major limitation to the interpretation of compliance to NIV and to respiratory dysfunction perception. Although we included the COPE questionnaire in the study, no detailed neuropsychological testing was performed to define the cognitive and behavioural profile of our study population which could provide information on potential additional management strategies we could have adopted. Moreover, the retrospective nature of the study did not allow to include sleep studies to identify those patients with predominant sleep related breathing disorders in whom the CNS component predominates and NIV has minimal effects on the central fatigue and sleepiness. Another limitation due to the retrospective nature of the study was the variable availability of the data and missing information between baseline and final assessments led to the need to consider progression as linear, but this may not be the case for all patients. 5 Our observational period was also probably not long enough to capture long-term major events of patients who should have used NIV but were not adherent to this prescription. Considering previous studies describing major events in the long-term for patients who should have used NIV but were not adherent to this prescription 8 we recommend caution in amending the current recommendations to initiate NIV in these patients. Further studies are needed to determine whether additional strategies could actually improve adherence in this patient population (home care to monitor NIV use where specialized nurses or PTs could also play an educational role in improving patients' sleep-wake cycle, refresh training and provide more explicit and accessible information on the effects of NIV adherence).
In conclusion, we present data on respiratory function over a relatively long period of time in a large cohort of well-characterized adult patients with DM1 and confirm that respiratory dysfunction is frequent and complex. Although we confirm that the older and more disabled patients with a longer disease duration are the ones usually requiring NIV, we provide evidence that patients may develop the need for nocturnal NIV support no matter the lack of respiratory symptoms and their degree of muscular disability. Although a minority are adherent to respiratory treatment and hypercapnia persists, the correction of the oximetry profile following NIV introduction and use, may, at least in part, contribute to the general improvement of well-being described in the patients available in this study. In line with other reports, 29,33 despite the limitations with adherence and the complexity to assess benefit given the role of the central component, we emphasize the need to recommend respiratory tests on first assessment and to closely monitor nocturnal oximetry and hypercapnia in the older, weaker and with higher BMI patients, despite their symptoms.

ACKNOWLEDGMENTS
We appreciate the time and availability of all patients and caregivers during the tests and procedures. We are very grateful to the whole NEMO team for the time and effort provided with passion and professionality during the clinical assessments. All data generated or analysed R e t r a c t e d during this study are included in this published article (and its supporting information files). Preliminary data were presented as a poster at the Italian Muscle Association Meeting in 2021.

CONFLICT OF INTEREST
On behalf of all authors, the corresponding author states that there is no conflict of interest. None of the authors and corresponding authors have nothing to disclose regarding the data and publication of this article. There has been no financial support to the study.

ETHICS STATEMENT
The study was conducted after IRB approval and all patients gave their informed consent to participate.

AUTHOR CONTRIBUTIONS
Carola Ferrari Aggradi has collected the data and seen patients. Elisa Falcier, Elisabetta Roma, Fabrizio Rao, Antonino Iannello, Elisa De Mattia, and Sara Lupone have performed respiratory assessments. Jascopo Casiraghi and Cristina Italiano have performed psychological tests. Elena Carraro and Valentina Gatti have performed neuromuscular tests. Luca Mauro has coordinated the study from a database and planning point of view. Alice Zanolini and Andrea Barp have recruited patients. Andrea Lizio and Alice Pirola have worked on the statistical analysis. VA Sansone has written the protocol, supervised the study and drafted and revised the paper.

DATA AVAILABILITY STATEMENT
The data that supports the findings of this study are available in the supplementary material of this article.