Report of the third outcome measures in myotonic dystrophy type 1 (OMMYD-3) international workshop Paris, France, June 8, 2015

Use of the Cambridge Brain Sciences online tests

Prior to the meeting, a SIG member (LR) examined the feasibility of using part of a computerized battery of tests in order to assess cognitive and executive functions among a sample of 25 participants with adult-onset. The protocol was composed of four tests from the Cambridge Brain Science Web-based battery to evaluate visuospatial working memory (Monkey Ladder), attention and inhibition abilities (Double Trouble), planning (Hampshire Tree) and spatial memory (Spatial Search). These tests are game-like, easy to administer and require no special staff training. When necessary, test instructions are easy to translate and the tests involve no language component, except for three words used in the attention test. The protocol takes approximately 20 minutes to complete by patients and since no learning effect is involved, test-retest procedures are possible. The instructions’ quality and understandability, and the ease of transition from instructions to tests were assessed. The participants were offered the option to use a mouse or a touch screen if they experienced difficulty manipulating a mouse. All results will be submitted for publication soon, but they are promising regarding the use of this cognitive test battery in the DM1 population.

Phenotype

A Delphi process took place in 2014-15 and mobilized 14 clinical and research experts. The process resulted in an array of suggested adjustments to existing definitions. During the OMMYD-3 meeting, a major focus of the participants was to seek consensus on phenotypes. The SIG discussions led to (1) the adoption of the criterion of “lumpers” instead of “splitters” for disease classification, and (2) a preliminary consensus on the five phenotype classification of DM1 considering the age of onset and most frequent presenting symptoms of each form: congenital, childhood, juvenile, adult, and late-onset. Frequent symptoms come from natural history studies and from the analysis of the largest registries of the world. Concerning the size of CTG repeats, it was decided to not include it in the phenotype classification because it is not always determined accurately, consistently and can be confounding. It can be a useful tool in research but to ensure consistency between research centers, this item was abandoned from the classification. The SIG also highlighted the need to have more data on cognitive impairments according to each phenotype. The detailed classification will be published after having done a larger consultation and having the agreement of the whole OMMYD community. A session independent of OMMYD should take place to get there.

Conclusion and recommendations

The suggested phenotypic classification needs to receive a general consensus among other DM1 experts and the members and chairmen of the OMMYD-3 experts through a congress or a workshop devoted to this specific topic of phenotypic classification. Other recommendations for future studies include evaluation of cognitive impairment in late-onset form and its clinical relevance. Moreover, longitudinal studies that can deepen knowledge on the progression of subtle cognitive deficits from baseline could be very helpful. Another issue to cover will be the specificity of congenital group that can be split in “severe” or “mild” form. This could be helpful concerning prognosis, because the mild subgroup may not experience respiratory difficulties in comparison to the severe one which face greater challenge living with more impairing difficulties including respiratory problems.

Myotonic Dystrophy Health Index

The Myotonic Dystrophy Health Index (MDHI) is a disease-specific patient reported outcome measure designed specifically for use in myotonic dystrophy clinical trials [6, 7]. It has been identified during OMMYD-2 [3] as potentially interesting and was translated into French and Italian at this time [8, 9]. Given the prior validation of this instrument, it has been identified by the NIH’s Common Data Elements group as one of only four outcome measures that are highly recommended for use in DM1 studies and the only outcome measure that is highly recommended as an outcome/endpoint to measure multifactorial patient reported Burden of Disease. One limitation identified was the need to document responsiveness to change for this questionnaire.

Congenital and childhood Myotonic Dystrophy Health Index

The Congenital and Childhood Myotonic Dystrophy Health Index (CCMDHI) is a proxy reported outcome measure designed to measure the level of disease burden experienced by children with myotonic dystrophy. The instrument was derived from patient and parent qualitative interviews and later validated in a multinational survey. This instrument is completed by a parent or proxy on behalf of a congenital or childhood onset DM1 patients. As part of the CCMDHI initiative, additional age specific instruments were created to measure disease burden in children with DM1 directly using a child’s perspective. These instruments were designed to be administered to specific age groups including patients ages 5–7, 8–11, and 12–17. Children complete questionnaires with assistance from a parent or proxy. For the 5–7 year-old age group, a picture based questionnaire is used. The instruments have been validated in a multinational disease progression study. Specific attention has been given to the weight of the proxy and pediatric responses in relation to other objective outcome measures.

ACTIVLIM

ACTIVLIM is a measure of activity limitation designed for children and adults with neuromuscular disorders [10]. The instrument was designed using Dutch and French neuromuscular patients. In the development process, 17 of the 369 patients (4.6%) had myotonic dystrophy and no children with myotonic dystrophy were included in the development of the instrument. The questionnaire was developed using Rasch methods and includes 14 base items, 4 child-specific and 4 adult-specific items. Intra-rater reliability was high among a neuromuscular population (ICC = 0.93).

LIFE-H

The short-version of Assessment of Life Habits questionnaire (LIFE-H 3.1) covers 12 generic domains of social participation with 77 questions including housing, responsibility and leisure. The total score has been reported to have an ICC of 0.86 for intra-rater reliability and 0.90 for inter-rater reliability among a DM1 population. Limitations of the LIFE-H include its completion time; it reportedly takes 20–60 minutes to complete. Additional versions of this instrument are being planned which may limit the amount of questions and the participant/examiner burden associated with this instrument.

DM1-Activ

The DM1-Activ is a disease-specific activity and participation scale developed for myotonic dystrophy type-1 using Rasch methods [11]. The questionnaire contains 20 items and was developed using DM1 patients recruited through the genetic register of the Maastricht University Medical Centre and through the Dutch neuromuscular patients’ association. A strength of this instrument is that it was designed specifically using and for DM1 patients. The instrument has an excellent ICC of 0.93 for intra-rater reliability. The instrument is currently being utilized in the Optimistic trial as one of the clinical outcome measures. Limitations of this instrument’s potential use for DM1 clinical trials were discussed and include the use of items that are seasonally and geographically tied (e.g. gardening), and the unknown relevance of some of the included items to the DM1 population (e.g. open toothpaste).

QoL-gNMD

The Qol-gNMD (formerly QoL-NMD) is a measure of health-related quality of life for adults with genetic neuromuscular disorders [12]. The instrument was designed using French patients with a variety of neuromuscular diseases; 44 of the 159 test sample had DM1. An English translation is available. The questionnaire was formed using item response theory (IRT) methods and includes 26 items and 3 domains. It showed adequate psychometric properties and met IRT assumptions. One limitation identified is the need to validate the questionnaire in on an independent sample of DM1 patients.

SF-36

The SF-36 is a generic health-related quality of life questionnaire. It was observed in previous studies that DM1 patients from Serbia had lower SF-36 scores compared to other countries showing the importance of transcultural studies utilizing patient-reported outcome measures. Analysis of quality of life changes in patients with DM1 during a five-year follow-up period, including assessment of responsiveness of the SF-36 questionnaire, was also presented. All mental subdomains, role physical and total SF-36 scores significantly improved after five years in 62 DM1 subjects. Unexpectedly, worsening of muscular weakness from mild to severe was in association with improvement of SF-36 scores.

Standardized administration procedures

The group discussed general recommendations when administrating tests. As fatigue is commonly present, the time of the day when tests are performed need to be considered. The group proposed that it would be preferable to perform tests between 10:00am and 4:00pm. Further, that fatigue/tiredness might be a confounding factor when capacity tests are performed, and that perceived fatigue/tiredness could be rated before and after a capacity test with the Borg category ratio scale with ratings from 0 to 10 (Borg CR10 scale ^®) or the Borg rating of perceived exertion scale (Borg-RPE-scale ^®) which ranges from 6 (no exertion at all) to 20 (maximal) [13]. When repeated tests are performed, in for example clinical trials or reliability studies, tests should be performed under the same conditions and at the same hour of the day. An important factor to take into consideration when administrating a capacity test, is the possible cognitive difficulties. This implies that the administrator of tests must make sure that the instructions are well understood, and the language should be adapted to the cognitive level of the persons to be tested. Short sentences and not too many instructions at the same time are recommended. As there is so far no consensus in the literature on how many trials to perform for each test, the suggestion from the group is that two trials are performed and that the best is used as the result.

Group members shared their experience of administering, scoring and interpreting results from various capacity tests. This led to a consensus on specific instructions for the administration of the 6MWT, 10mWT, 30s-CST, and 9HPT in adult DM1 patients. Complete administration guidelines are presented as supplementary file as well as development process.

Balance outcome measures

The group discussed different balance tests, such as the Berg balance scale, the mini-BESTest the walk in figure-of-eight test, stand on one leg test, and the step test. The Berg balance scale has been used in Canada and they found that it had major ceiling effects. The mini-BESTest will be used in a forthcoming longitudinal study, and for the next OMMYD meeting there will be research/clinical experience for this scale. The stand on one leg test and the walk in figure-of-eight test were considered to be of less value in DM1 as large within-subject variations and measurements errors have been reported for these tests. The step test has been recommended to be a reliable dynamic balance test, but members in the group had limited experience with this test and it was decided not to recommend this test as a “must” test at the moment.

Since OMMYD-2, new data were obtained regarding the use of a Wii balance-board to measure balance and was presented to the group. Further validation and development of the test and data on dynamic balance are necessary before taking a decision for this test.

Outcome measures for children

The SIG recommends that members who have expertise or experience on functional capacity outcome measures for children with DM1 be invited. A battery could thus be developed at the next meeting.

Revision and discussion of the consensus adopted during OMMYD-2

SIG report from OMMYD-2 was revised and discussed, more specifically the desired features of the muscle impairments assessment protocol (see Table 1 of the second report). The first domain revised was the general statement, the assessment of muscle impairments. One issue in that domain that was discussed is the value of strength that should be retained for the final analyses according to the number of trials performed. After discussion, it was deemed appropriate to use, depending on the study, either the mean peak value of all trials or the highest peak value among all trials. However, it must be clearly understood the impact of each choice. With weaker patients, although it may be appropriate to use the peak force, caution is advised when using the peak force with a HHD with weak patients as it is easy to induce a break test and therefore not reporting the true isometric value. With regard to the assessment of both sides, as we cannot assume that both sides are equally affected, it remains the judgement of the clinician/researcher to decide but the assessment of both sides should be favoured as asymmetry of muscle impairments is often observed. And finally, the issue of assessing muscle endurance (defined as handling an isometric contraction between 50 to 70% of the maximal force for as long as possible) in this population was discussed. However, considering the negative past experience of some participants with this variable (presence of pain and difficulty to concentrate enough to maintain the targeted high level of contraction for a long period of time) and the subjectivity of this measure with DM1 patients in addition to the necessity to reach a state of exertion, it was agreed to not add this item to the main list of outcomes.

Also, a few clarifications were suggested. In the exclusion criteria, the terms congenital and infantile phenotypes should be well defined. It must be clear that this protocol focuses on adults. Also, the cognitive impairment as an exclusion criterion is determined by clinical judgment. Participants also mentioned that excessive daytime sleepiness should be better described by the SIG working on that issue.

With regard to the recommended muscle groups (see Table 2 of the second report), the overall consensus previously adopted remains but the long finger flexors were suggested as a valid indicator to detect early on the onset of the disease. However, as no specific protocol exists yet for measuring the muscle strength of this specific muscle group, the assessment process of this additional muscle group will need to be defined and validated. The consensus adopted on the number and distribution of muscle groups also remains but it was pointed out that in any study, at least one muscle group of each lower limb joint (hip, knee, ankle) should be assessed and additionally, the anti-gravity muscles should be favoured.

The previous consensus was maintained regarding the method to assess muscle strength (until the transition to QMT is completed, the use of MMT should favour the modified MRC scale/functional assessments graded from 0–10 in clinical research; however, QMT must be favored for muscle groups that have shown good to excellent psychometric qualities) but comments from representatives from the pharmaceutical industry reinforced the fact that quantitative measures may be favoured over manual muscle testing as the use of an ordinal scale may require a higher sample size to detect the effect of a therapeutic intervention.

The use of myotonia as a significant biomarker was also revisited. This variable seems to be perceived as an important one by the experts but the way myotonia is measured between centers seems to vary considerably. One suggestion was to gather the psychometric qualities of both measures and present the results to the next OMMYD. This may help to compare the methods and explore the possibility to keep both or recommend the use of one. It was also suggested that researchers who have data on myotonia further analyze their data to better understand the profiles of change, comparing isometric grip testing with isotonic testing and needle EMG testing. This will provide additional arguments in favour of using this variable or not.

Comparison of the OMMYD’s consensus and the Common Data Element published by the NIH

The NIH held a meeting of the National Institute of Neurological Diseases and Stroke and suggested the creation of categories of neurologic diseases, with common elements that would enable benchmarking. The NIH report [15] that was recently posted and submitted for public consultation provides succinct recommendations but it is unclear how these recommendations were obtained and especially how the literature review that supports these recommendations was conducted. Specifically, several participants were very surprised by the recommendations on muscle testing as they look more like generic guidelines and some are even in opposition to what was recommended in the last OMMYD meeting. For example, from what was published in the past decade, there are no supportive arguments in favour of the use of break test. And in the same line of thoughts, while manual muscle testing is still used, there are several and strong arguments in the literature in favour on the use of quantitative muscle testing [14]. The bibliography used in the NIH report for the section on muscle strength seems incomplete and does not reflect the current level of knowledge on the assessment of muscle strength in DM1.

Discrepancies between protocols used by research groups in their DM1 studies

A general discussion took place around the possibility to develop/use the equivalent of an ISO norm to assess muscle function. Some tools/protocols used to assess muscle function are very responsive but require specific knowledge and skill sets. Thus, evaluator training is of utmost importance. While there is no ISO norm per se yet to measure muscle function, four basic parameters were identified: evaluators should be well-educated regarding muscle function; the best available equipment should be used; evaluators should receive high quality, standardized training; and training should involve a certification process.

The SIG also discussed the use of normative values that is a term that should be used with caution. The use of normative data to compare results may at some point have to struggle with regional differences within countries. The use of the word ‘Reference values’ was suggested as it implies the limitations associated with the use of such data. Reference values should be collected from various geographical areas and each population or group of patients must be clearly described to ensure reproducibility of studies.

Fatigue and Daytime Sleepiness Scale (FDSS)

Some concerns were raised about FDSS [17] items 9, 10, and 11 which were derived from the Epworth Sleepiness Scale and slightly modified (see supplementary material). It relates mainly to the necessity of providing examples to patients. The SIG advises to talk to the authors of the FDSS to debate these issues and help reach an agreement. For the meantime, the SIG reached a consensus to simply deliver rating scales to patients without giving examples, so not to bias the results and ensure better comparability between neuromuscular clinics. However, the SIG acknowledges the relevance of developing a guide with standardized examples to be given to patients whose verbal abstract reasoning skills have been previously raised.

A sample of 143 patients from MRC Neuromuscular Centre (Newcastle) filled out the FDSS. Mean±SD score was 44.8±16.4 and female patients exhibited higher FDSS scores than male (p < 0.05). Also, FDSS scores were significantly associated with body mass index (p < 0.05), Muscular Impairment Rating Scale (MIRS) score (p < 0.001), QT interval (p < 0.05), and sitting forced vital capacity (FVC) (p < 0.05). On the other hand, FDSS scores did not vary with age, walking distance, lying FVC, blood pressure, heart rate, PR interval, and QS duration. Moreover, reliability of the FDSS was documented on 20 patients (7 from the Newcastle neuromuscular center and 13 from the Clinique des maladies neuromusculaires, Jonquière). Internal consistency and intra-rater reliability were good (Cronbach’s alpha = 0.82; ICC = 0.82, p < 0.001).

Multiple Sleep Latency Test and Maintenance of Wakefulness Test

Some SIG members suggested that the Multiple Sleep Latency Test (MSLT) characterizes the intensity of DM1 patients’ sleepiness while the Maintenance of Wakefulness Test (MWT), by asking them to stay awake, both tracks sleepiness as well as capacity/motivation to stay awake. Hence, if the MWT is improved, both sleepiness and motivation may be improved whereas improvement regarding the MSLT may solely reflect improved sleepiness per se. Results yielded by these two polysomnographic techniques should be compared in patients in order to clarify their relationship and increase our understanding of the reportedly various excessive daytime sleepiness (EDS) phenotypes noted. Also, studies should explore whether some personality features influence MSLT results. SIG members feel that an important issue is to better characterize the habitual sleep/wake cycle of DM1 patients. Are they all long sleepers? What is their napping behavior? Is their habitual sleep/wake cycle irregular? Does it vary with employment status or activity level? Actigraphic studies could namely help clarifying these issues.

Apathy Evaluation Scale (AES)

The three versions of the AES (self (AES-S), clinician (AES-C), and informant (AES-I) (significant other, e.g. personal or professional caregiver)) [18] were administered at the Clinique des maladies neuromusculaires (Jonquière). It was noted that patients had difficulty understanding some AES items. The advantages of the AES are that it can be rapidly administered (about 10 min for the clinician version and 3 min for the self and informant versions) and that there is no significant difference in the median score obtained from the clinician, self, and informant versions (n = 20). Preliminary results regarding test-retest reliability of the AES in patients (n = 13) and caregivers (n = 9) suggest good to excellent reliability (0.70 < ICC < 0.85).

Lille Apathy Rating Scale (LARS)

Results concerning the use of the Lille Apathy Rating scale [19] to measure apathy in 87 patients with classic adult and late-onset phenotypes were presented. It was found that the 4 dimensions of apathy tracked by the LARS (intellectual curiosity, emotion, action initiation, self-awareness) were significantly higher in DM1 patients with the classic adult phenotype than in those with the late-onset phenotype (p < 0.01). Of note is that about 1 of 4 patients (24%) with the classic adult phenotype met the criterion for apathy while none of the late-onset patients did.

Face validity of commonly-used fatigue rating scales

SIG members first examined the subjective fatigue experience dimension of the Checklist Individual Strength (CIS20-R) [20], a well validated and widely used multidimensional self-report measure that aims to assess subjective experience of fatigue, concentration, motivation, and physical activity. Some SIG members expressed the view that the 8 items of the fatigue subscale seemed to evaluate peripheral or muscle fatigue consecutive to physical activity rather than central fatigue per se. Others were of the opinion that the subjective fatigue items appeared to have been devised for healthy people rather than for individuals with chronic neuromuscular conditions. In addition, SIG members expressed some concerns regarding the content validity of the fatigue subscale, most items seemingly focusing on physical fatigue and not sufficiently documenting aspects of mental fatigue or weariness that many patients typically feel upon morning awakening. Moreover, many items are not plausible for DM1. For example, no clinician would ask DM1 patients whether they “feel fit” (item 3). Similarly, it may not be relevant to ask a patient whether s/he feels, physically, in an excellent condition. Also, the timeframe “in the last two weeks” is not relevant for DM1 patients. Results from the OPTIMISTIC study on fatigue and exercise therapy shall permit to further document the validity of this rating scale, namely the convergent validity.

The 9-item Fatigue Severity Scale (FSS) [21] was originally developed for patients with multiple sclerosis and systemic lupus erythematosus. Its reliability was previously determined in DM1. As regards item 5 “Fatigue causes frequent problems for me”, it was noted that patients often ask to which problems this refers to. In order not to bias patients’ answers, SIG members advise to say “any kind of problem that you may think of”. On the other hand, it has been generally easy for patients to fill out this scale. Finally, SIG members particularly appreciated the fact that the FSS addresses fatigue’s effects on daily functioning and that patients can easily refer to their daily life situations.

The Chalder Fatigue scale [22] was designed to measure the severity of fatigue in adults. Its 14 items examine symptoms of both physical and mental fatigue. This scale has been shown to be reliable in DM1 patients. It can be administered both by interview and through the use of a pencil-and-paper test. SIG members believe that the response options are not adapted to such a chronic condition as DM1, patients oftentimes having been drowsy or fatigued for decades. SIG members do not routinely use that scale because they hold the view that the content validity remains questionable. For example, items 7 and 8 “Do you have less strength in your muscles?” and “Do you feel weak?” respectively appear unsuitable for DM1.

The Brief Fatigue Inventory [23] was devised to assess the severity of fatigue in cancer patients. It is based on the same principle as a visual analog scale for fatigue. SIG members believe that it may not be relevant to inquire about fatigue and its effects during the past 24 hours in the context of a chronic, progressive neuromuscular disease. Moreover, this rating scale does not evaluate symptoms of fatigue and their repercussions on patients’ lives. Even if there may be a bit of fluctuation in the daily levels of fatigue that patients experience, it is not of particular interest to document peaks and troughs from day to day. On the other hand, circadian variation of fatigue levels may be worth studying, but other scales such as the Stanford Sleepiness Scale and the Karolinska Sleepiness Scale should preferably be used.

The Piper Fatigue Scale (PFS) [24] is a self-reported instrument measuring cancer-related fatigue. The scale has well established reliability and validity in American women with breast cancer. SIG members first consider that the 27-item PFS is a bit long and that the answer choices are too numerous (1 to 10). Almost half of the scale’s items ask patients to which degree they are “now” feeling strong, refreshed, relaxed, lively, etc. Other items deal with the degree to which fatigue experienced “now” interferes with various aspects of patients’ lives (eg. work and school sexual activity, social relationships, etc.). Some SIG members have the impression that some items tap sleepiness more than fatigue. In all, this fatigue rating scale does not seem applicable to DM1.

The Fatigue Symptom Inventory (FSI) [25] was developed with women undergoing treatment for breast cancer and aims to assess the severity, frequency, and diurnal variation of fatigue, as well as its perceived interference with quality of life. Also, the Multidimensional Fatigue Symptom Inventory – Short Form (MFSI-SF) [26] was developed to capture the full spectrum of the fatigue symptom profile in cancer patients. SIG members consider that those scales are too broad and unspecific as regards DM1.

Respiratory Symptom Checklist questionnaire

The RESPICHECK questionnaire was drafted during the 207th ENMC workshop on the diagnosis and management of chronic respiratory insufficiency in myotonic dystrophies. It includes 9 domains: orthopnea, dyspnea, sleep, headaches, apnea, cognitive performance, excessive daytime sleepiness, fatigue, and chest infection. Data was presented from the experience at the NEMO Center and at the Neurological Institute Besta in Milan. The initial questionnaire was administered twice within a 4-week interval to 25 patients to assess test-retest reliability (correlation coefficient 0.02; r = 0.82; p = 0.65; ANOVA = 0.25). The RESPICHECK questionnaire was then correlated with the Epworth Sleepiness scale (ESS) and other respiratory parameters. It showed to have a good convergent validity with the ESS (r = 0.52, p = 0.02) and, after adjusting for age and sex, showed to have a direct correlation with the pC02 level (p < 0.05), and an indirect correlation with the sitting forced vital capacity and the forced expiratory volume (p < 0.05).

This checklist was reviewed item by item first with the respiratory SIG and then with the Sleepiness/Fatigue/Apathy SIG. Wording was reviewed and questions related to excessive daytime sleepiness and fatigue were modified by the respiratory SIG. A new version of the questionnaire was agreed upon (see Table 2) and validation of this version is currently ongoing.

Existing respiratory parameters used in clinical trials

An overview of respiratory outcome measures used in other trials in pulmonary diseases was given as well as programmes to improve adherence and compliance in other pulmonary diseases. The SIG ultimately concluded that a multidisciplinary approach is needed to ensure the best compliance. Experience from other trials seems to show that preliminary psychological interviews on motivation and educational training sessions on the importance of non-invasive ventilation and respiratory involvement in general may be a good option to improve compliance and adherence to prescriptions in this population. There is a lack of standardized respiratory outcome measures which is known to be sensitive to change in this population group and this needs further investigation.