Year in Review 2019: Neuromuscular Diseases
Abstract
Neuromuscular cardiopulmonary medicine is entering a new and exciting phase, with studies that assess the respiratory effect of emerging genetic and molecular therapies. In this year’s neuromuscular Year in Review, we focus on Duchenne muscular dystrophy (DMD), reviewing studies that evaluate the respiratory effect of eteplirsen, the cardiopulmonary effects of ataluren, and a study comparing the use of spironolactone to eplerenone for the treatment of DMD-related cardiomyopathy.
The neuromuscular respiratory medicine section of last year’s Year in Review 1 highlighted articles related to genetic therapies affecting the respiratory system. That area of inquiry is expanding rapidly and this year some of the first larger scale studies were published regarding the respiratory effect of genetic therapies for Duchenne muscular dystrophy (DMD).
DMD is characterized by absence of functional dystrophin protein, resulting in damage to and inflammation of muscle tissue, with subsequent fibrosis and progressive loss of skeletal muscle function, including respiratory function.2 Genetically, DMD is an X- linked disease that is usually caused by deletions or duplications in one or more exons of the DMD gene, shifting the reading frame, producing premature stop codons, and interfering with dystrophin production. 3 However, in 10-15% of patients, DMD is caused by nonsense mutations. Ataluren is an oral drug that suppresses premature stop codons, enabling full-length protein production from genes disrupted by nonsense mutations.
Ataluren is approved for use in nonsense mutation DMD (nmDMD) in the European Union as well as in several other countries.
In a study from Mercuri et al.,4 the efficacy of ataluren in treating nmDMD was assessed by comparing two populations: patients from the Strategic Targeting of Registries and International Database of Excellence (STRIDE) registry, matched to those from an historical control registry, the Cooperative International Neuromuscular Research Group Duchenne Natural History Study (CINRG DNHS).
STRIDE is an ongoing multicenter post-approval data base on ataluren safety and use in nmDMD in clinical practice. Patients selected were enrolled between 2015-2018. CINRG DNHS was a prospective longitudinal multicenter natural history study of 400 DMD patients followed between 2006-2016.
Patients who received Ataluren plus standard of care in the STRIDE registry were compared to those who received standard of care alone in the CINRG registry. 181 male patients within the STRIDE registry with a confirmed genetic diagnosis of nmDMD and who had received at least one dose of ataluren were eligible for participation in the effectiveness outcome study. Subjects from the historical control pool, CINRG DNHS, were required to have a diagnosis of DMD confirmed by either genetic testing, a positive muscle biopsy, or a combination of creatinine kinase (CK) levels that were at least 5x normal and a relevant X linked pedigree. CINRG DNHS subjects were excluded if they had received investigational drugs in previous trials. The authors used propensity score matching to seek comparable groups between the two registries. Criteria for the matching of groups included: age at first clinical symptoms, age at initiation of corticosteroids and duration of corticosteroid, including deflazacort.
Age at loss of ambulation was used as the primary endpoint of treatment benefit. Secondary endpoints included age to specific timed motor declines in standing and stair climbing, age to predicted FVC of less than 60%, 50% and 1 liter; and age to left ventricular ejection fraction (LVEF) less than 55%, or shortening fraction (SF) <28%. Of the 181 propensity score matched pairs in the study, the median age at loss of ambulation was 14.5 years for the STRIDE group and 11.0 years for the CINRG DNHS, with a statistically significant difference in favor of ataluren treatment (p<0.0001).STRIDE patients also performed better in regard to delayed onset in motor declines for standing and stair-climbing with p values statistically significant in all timed tests of motor function. The younger age of the STRIDE cohort, coupled with a relatively short three year observation period and resultant paucity of advanced clinical events in the treated group made it premature for the authors to draw meaningful conclusions from the pulmonary and cardiac data. Specifically, the mean age at last assessment was just 11.23 years in the STRIDE group, compared with 15.73 years in the CINRG control group (p < 0.0001).However, there was a trend towards delayed worsening of pulmonary function (FVC < 60% predicted, FVC < 50% predicted and FVC < 1 liter) and the comparisons were statistically significant. One ancillary area of investigation from the STRIDE nmDMD cohort was the degree of correlation between DMD genotype and phenotype. The latter was designated by disease severity, specified by the authors as age at symptom onset. STRIDE has the advantage of being the largest registry of nonsense mutation DMD patients. The authors found no relationship between disease severity and the exon location of the nonsense mutation out of 181 patients, with an impressive level of phenotypic variability among patients with identical dystrophin mutations. The authors conclude that patients receiving ataluren plus standard of care in the STRIDE registry had significantly delayed age for the following parameters: loss of ambulation and worsening of performance on timed tests, with a trend toward delay in worsening of pulmonary function in comparison with historical controls receiving standard of care alone. Changes in cardiac function did not reach statistical significance. No correlation was observed between DMD genotype and either DMD phenotype or treatment benefit for patients with nmDMD. In another notable study, Khan et al. 5 studied the respiratory effects of eteplirsen. Eteplirsen is a drug that acts to promote dystrophin production by restoring the translational reading frame of DMD through specific skipping of exon 51 in defective gene variants. Approximately 14% of patients with DMD mutations are eligible for eteplirsen therapy. This report describes three trials which assessed annual percent change in percent predicted FVC (% FVC) in eteplirsen-treated DMD subjects compared to 3 cohorts from the Cooperative International Neuromuscular Research Group (CINRG) Duchenne Natural History Study (DNHS) registry, forming the historical control (all three CINRG control cohorts were treated with glucocorticoids, but differed as to genetic profile, consisting of: no genetic testing, genetically confirmed mutations, and mutations amenable to exon 51 skipping). Eteplirsen was administered as weekly intravenous infusions. The mean age of the subjects at baseline was approximately 10 years to 13 years in the eteplirsen treated group and approximately 11.75 years in the 3 historical cohorts. Eteplirsen study 201 was a 28 week single-center, randomized, double-blind, placebo- controlled study of 12 DMD patients with mutations amenable to exon 51 skipping.Treated patients had an FVC ≥ 50% predicted and were receiving treatment with glucocorticoids on a stable dose for at least 24 weeks prior to study entry. Study 202 was an open label extension study in which the patients who were originally randomized to eteplirsen in trial 201 continued on the same dosage (30 or 50 mg/kg/week), and patients originally randomized to placebo in trial 201 were re- randomized to receive either eteplirsen 30 mg/kg or 50 mg/kg weekly for the last 4 weeks of study 201 and for the duration of trial 202 (284 weeks). FVC was measured every 24 weeks through 216 weeks. Study 204 was an open-label 96 week trial that followed 20 DMD subjects with mutations amenable to exon 51 skipping. Patients were on a stable dose of glucocorticoids and were categorized as minimally ambulatory or non-ambulatory. Eteplirsen at 30 mg/kg was administered weekly through 96 weeks. FVC was assessed every 12 weeks for a year and then every 24 weeks until study end. Study 301, an ongoing, open-label, multicenter study followed 42 DMD subjects who received weekly infusions of eteplirsen at a dose of 30 mg/kg. Subjects had gene mutations amenable to exon 51 skipping. FVC was assessed every 12 weeks for one year, then every 24 weeks for 2 years, with an interim analysis performed at 96 weeks. Results: The eteplirsen treated patients in studies 201/202 had the best baseline % predicted FVC (with a mean value 96.9% predicted, compared with a mean value of 79.6% predicted in the control cohort with the best baseline FVC), even though the eteplirsen treated patients were on average just 1.5 years younger than the control subjects. Using a rather complex mixed model for repeated measures (MMRM) statistical analysis, the authors found that there was a statistically significant attenuation in the annual decline of % predicted FVC in the eteplirsen treated patients compared with the 3 historical control cohorts. Among the treated patients, the % predicted FVC declined 2.19%/yr for the subjects in studies 201/202, 3.66%/yr for the subjects in study 204, and 3.79%/year for the subjects in study 301. By comparison, % predicted FVC declined 5.56% to 6%/yr among the subjects in the 3 historical control cohorts. The authors conclude that significant, clinically meaningful attenuation of the expected annual decline in % predicted FVC was observed in eteplirsen-treated patients when compared with the CINRG DNHS controls. The studies of ataluren and eteplirsen, summarized above, are important because they are among the first to examine cardiopulmonary outcomes for emerging DMD therapies, comparing data from treated subjects to a carefully selected historical control population. The results suggest that the studied therapies may have cardiopulmonary benefits.However, it is crucial to consider the limitations of these studies, including the potential for cardiopulmonary phenotypic variability and discordance to confound their results. Phenotypic variability is defined as divergent pulmonary function or cardiac function among patients who share a common dystrophin mutation. 6 For example, a DMD patient with a favorable pulmonary phenotype (i.e., an FVC that peaks at a high absolute value and then declines slowly) may have a brother with a detrimental pulmonary phenotype (i.e., an FVC that peaks at a much lower level and declines much faster), despite the brothers’ shared dystrophin mutation. Phenotypic discordance is defined as diametrically opposite pulmonary and cardiac function in a single patient. For example, in one study, prolonged survivors of DMD (mean age 34.3 years) had unexpectedly excellent heart function (mean ejection fraction [EF] 42.2%) despite their advanced age and need for 24 hour/day assisted ventilation due to respiratory failure (all patients had FVC of 0 mL). Conversely, DMD patients who experienced early death (at mean age 21.7 years) had unexpectedly poor heart function (mean EF 29.2%) despite their younger age and relatively good pulmonary function (mean FVC 804 mL).Phenotypic variability and discordance create a group of “phenotypic outliers”--- patients with unexpectedly favorable or detrimental cardiac or pulmonary function that cannot be predicted by the patients’ dystrophin mutation. These “outliers” have the potential to confound studies based on aggregate data, like the studies on ataluren and eteplirsen summarized above. For example, favorable cardiopulmonary results attributed to the benefits of the therapy could simply be due to the fact that a significant portion of the study population had an unexpectedly favorable cardiac or pulmonary phenotype. In one of our recent studies, cardiopulmonary phenotypic discordance was surprisingly common, affecting one-third of our patients aged 18 years and older. 8 These results suggest that DMD therapies should be evaluated in the context of each individual patient’s particular cardiopulmonary natural history, and not with aggregate data, as was done in the ataluren and eteplirsen studies. The results also suggest that grouping patients by identical dystrophin genotypes does not assure phenotypic homogeneity in a DMD population that is exposed to a particular therapeutic intervention. Phenotypic variability and discordance have other implications for study design. Favorable and detrimental cardiopulmonary phenotypes are defined by natural histories that evolve over a period of years. For example, in one study of cardiac function in DMD, the favorable phenotype was characterized by onset of left ventricular dysfunction (LVD) at or after 18 years of age, while patients with the detrimental phenotype had onset of LVD before 18 years of age.9 After the onset of LVD, both groups had progressive worsening of cardiac function, leading to congestive heart failure, at which time median survival was just 8 months. The survival between the groups did not diverge until the patients reached their late teens to early twenties; cardiac function was usually normal in the youngest patients, regardless of ultimate phenotype. The implications of the prolonged timeline over which favorable and detrimental phenotypes evolve is that studies of cardiopulmonary function should include older patients, in their mid-teens to early twenties, and an extended duration of observation. Instead, subjects in the studies of ataluren and eteplirsen were generally young at baseline (10-13 years old) and were mostly observed over a time period of 2 to 4 years. Additionally, as expected in younger cohorts, pulmonary function was generally well-preserved; for example, the patients in eteplirsen sub-group 201/202 had a mean baseline FVC of > 96% predicted. It is hazardous to ascribe long-term efficacy to therapies studied in younger patients with good baseline pulmonary function for relatively short periods of time. While such studies may show statistically significant short-term respiratory benefits, they do not assess the ability of new therapies to protect patients from the morbidity and mortality associated with long-term respiratory decline.
The ataluren study uses an FVC < 50% predicted as an outcome measure. This threshold is based on the Care Considerations publication in which an FVC < 50% predicted was recommended as an indication to start nocturnally assisted ventilation. 2 The idea is that the longer a patient’s FVC can be maintained at a level of 50% or higher (for example, due to a therapy such as ataluren or eteplirsen), the longer a patient will not require assisted ventilation, benefiting the patient’s quality of life. However, it is important to recognize that an FVC level below 50% predicted was chosen as a threshold for initiating nocturnal ventilation from the consensus opinion of a group of experts using a structured rating method and is not an evidenced based recommendation derived from prospective, randomized trials.3 Very low levels of pulmonary function (VC < 1 liter) are associated with poor survival in patients who are not treated with assisted ventilation. 10 However, there are no controlled studies associating better outcomes, like improved survival, to initiating assisted ventilation at levels of FVC < 50%. While treatment with assisted ventilation is a primary reason for increased survival in patients with DMD,11 recent studies have shown poor adherence by patients with DMD to assisted ventilation,12 suboptimal adherence by clinicians to expert guidelines recommending initiation of assisted ventilation at specific levels of pulmonary function,13and in one new study, unexpected acceleration of the rate of pulmonary function decline in steroid treated DMD patients who had assisted ventilation initiated when their respiratory impairment was very mild, as assessed by polysomnography. 14 Thus, the use of FVC < 50% predicted as a clinically meaningful pulmonary outcome measure has marked limitations. Another limitation of the ataluren and eteplirsen studies is their assumption that a therapy which improves pulmonary function will necessarily improve patient survival. The last sentence of the eteplirsen study is, “The treatment effect of eteplirsen was clinically meaningful and may translate to prolonged time to required mechanical airway clearance, non invasive ventilation, reduced risk of hospitalization due to respiratory illnesses, improved quality of life, and improved survival.” To the contrary, in the study of prolonged DMD survivors and DMD patients who experienced early death referenced previously, cardiac function determined patient survival when patients with respiratory failure were treated with assisted ventilation. 7 Specifically, prolonged DMD survivors had an unmeasurably low FVC, but they were treated with 24 hour/day assisted ventilation. The key to their survival was their unexpectedly good cardiac function. The patients who experienced early death had surprisingly poor heart function and died at a young age, despite their relatively good pulmonary function. Again, cardiac function determined their survival. More studies are needed evaluating the cardiac effect of new DMD therapies, since cardiac function determines survival in DMD patients treated with contemporary respiratory management.
In this context, a recent cardiology study merits attention from pulmonologists and others interested in neuromuscular cardiopulmonary medicine. Standards of care for the management of cardiac complications of DMD consist of treatment with angiotensin- converting enzymes inhibitors (ACE) or angiotensin receptor blockers (ARB). A study from 2015 15 showed that when epleronone, a mineralcorticoid receptor antagonist (MRA), was added to ACE/ARB therapy, the outcome was superior to ACE/ARB alone. The outcome measure was left ventricular strain assessed by cardiac magnetic resonance imaging, a highly sensitive test that assesses early myocardial damage, and that precedes a fall in ejection fraction or evidence of myocardial damage assessed by late gadolinium enhancement (LGE). The latest study from the same group (Raman et al. 16) was undertaken to determine if a less costly, more widely available MRA, namely spironolactone, could be used instead of eplerenone. The study was a multicenter, double-blind, randomized trial. Subjects were boys aged ≥ 7 years with a clinical and genetic diagnosis of DMD. Twenty-six subjects were randomized to eplerenone, 26 to spironolactone and the primary outcome measure was the change in left ventricular strain (Ecc) after 12 months of therapy.
Baseline and 12 month follow-up cardiac magnetic resonance (CMR) with LGE were performed. LVEF was assessed by echocardiography, cardiac nuclear scan or CMR at baseline and at 12 months. All participants had a baseline LVEF ≥ 45 +/- 5%.Of the 52 subjects, 12 had no evident myocardial damage at baseline, the rest had a median of 2 LV segments that were abnormal. Most subjects were treated with corticosteroids (p=0.7); two-thirds were treated with an ACE or an ARB. The dose of both eplerenone and spironolactone was 50 mg administered once daily. Serum potassium levels, glomerular filtration rate, and spirometry were followed throughout the study to assess for adverse effects. Patients self-reported health was assessed throughout the study period via questionnaire.
LV strain (Ecc) by CMR remained stable in both groups between baseline and the 12- month follow-up (p=0.542). Myocardial damage, as measured by LGE, was similarly stable in both eplerenone and spironolactone groups. The authors also noted that the 50 mg dose achieved greater stabilization of LV strain compared to their previous study with a 25 mg dose of eplerenone alone. Pulmonary function, GFR, serum potassium and health status remained stable or normal in both groups. The authors concluded that spironolactone, a lower cost MRA with greater availability worldwide, was non-inferior to eplerenone, and may lessen progressive myocardial damage and deterioration if used early in the course of DMD cardiomyopathy, when ejection fraction is preserved.
This study has important implications for changing the standard of care for DMD cardiomyopathy, with the aim of preventing or delaying deterioration of cardiac function. However, it is important to keep in mind that the outcome measure (namely, left ventricular strain) is a very early finding in DMD cardiac involvement. An underlying assumption of the study is that patients with this early finding will progress inevitably and at a similar rate to clinically significant left ventricular dysfunction, and subsequently to congestive heart failure and death. However, it is unlikely that all DMD patients share a uniform cardiac natural history.9,17 Cardiac function over time appears to vary between patients, with some DMD patients experiencing normal or near-normal cardiac function into their early 30’s.8,9,17 As discussed above, this favorable cardiac phenotype has been associated with prolonged patient survival, irrespective of therapeutic pharmacologic interventions.
Two areas require more study in order to improve the clinical validity of studies of cardiac therapies: first, an understanding of the relationship between early cardiac changes identified by CMR and the subsequent natural history of DMD cardiomyopathy; and second, a more complete understanding of the prevalence and nature of cardiac phenotypic variability, by which favorable phenotypes can prolong survival, and detrimental phenotypes can cause early death.7
In summary, neuromuscular respiratory medicine is a rapidly evolving field. With the emergence of new therapies for DMD, respiratory outcome measures are being studied, and two of the articles in this year’s review describe the possible respiratory benefits of new genetic therapies. However, these studies have potential methodological flaws, which we have discussed in this review. For example, cardiopulmonary phenotypic variability may be common in patients with DMD. A patient may have unexpectedly good pulmonary function over time that is ascribed to a new therapy when, in fact, the patient is simply expressing a beneficial pulmonary phenotype. As a result, “outlier” patients with unexpectedly good or poor cardiopulmonary function could confound studies that are based on aggregate data. Positive results from studies of new DMD therapies conducted in young patients with good baseline pulmonary function over relatively short time periods should not be over-generalized, as they do not assure that those therapies can protect patients from pulmonary morbidity and mortality over the long-term. Moreover, in DMD patients who are treated with assisted ventilation for their respiratory failure, it is cardiac function, and not pulmonary function, that is the main determinant of their survival. Thus, it is incorrect to equate pulmonary function with prolongation of survival. The last article reviewed in this paper discusses the potential benefits of spironolactone for DMD cardiomyopathy. Future studies of new DMD therapies need to include cardiac outcomes in order to assess the effect of those therapies on patient survival.