August 2012: Volume 38, Number 8
Newborn Screening for Duchenne Muscular Dystrophy
New Therapies, Screening Strategy Poised to Change Practice
By Genna Rollins
Discovery in 1986 of the DMD gene, mutations of which cause Duchenne muscular dystrophy (DMD), raised hopes that treatment for DMD, the most common and severe form of muscular dystrophy, would be right around the corner. However, the lengthy and laborious process of fully understanding and devising responses to the disease pathology set in, and now, 26 years later, those long-awaited therapies finally are on the verge of clinical reality. At the same time, a recently published landmark study has established the feasibility and accuracy of a two-tiered newborn screening paradigm using dried blood spots. The convergence of these two pathways marks the start of a particularly promising era in the muscular dystrophy field, and it could lead in the near future to implementation of newborn screening for DMD.
“For Duchenne, we predict that within the next five years there will be disease-modifying therapies in the marketplace, so if we have a way to diagnose the condition at birth or before symptoms actually start we might be able to initiate therapies before the children begin to lose muscle,” said Sanjay Bidichandani, MBBS, PhD, vice president of research at the Muscular Dystrophy Association (MDA). “The recent newborn screening study demonstrates that it is feasible at the state level using a two-step strategy in the U.S. healthcare system. The point is not to have screening for Duchenne implemented today; obviously, we’re not ready for that at many levels. But the goal is to be ready by the time these disease-modifying treatments become available in the marketplace. We want to time it so that affected families don’t loose years of beneficial treatment and services.”
The Absence of Dystrophin
An X-linked condition estimated to affect 1 in 3,500 males, DMD causes progressive loss of muscle tone, contractibility, and strength, leading inevitably to death by the early 20s. Affected boys typically start showing symptoms by about age 2½, but are diagnosed on average at age 5, a statistic that hasn’t changed in at least 2 decades. This hard news usually comes only after families have undergone a long diagnostic odyssey involving multiple specialists and batteries of tests, and unfortunately for many, after they have had other affected children.
Discovery of the DMD gene led in turn to discovery of dystrophin, a protein that helps maintain muscle integrity by linking the muscle cytoskeleton and extracellular matrix. DMD mutations affect levels of dystrophin; without it, muscle membranes develop abnormalities that induce necrosis, enabling fat and connective tissue to replace muscle. Nearly two-thirds of DMD cases arise from out-of-frame deletions of one or more exons that code for dystrophin, truncating the protein and leaving essentially none available to support muscle structure. Up to 10% of cases result from duplications of one or more exons, and the remainder are due to point mutations, small insertions or deletions, and splice site changes. Milder, less common Becker MD arises from in-frame deletions or duplications, leaving shorter-than-normal but partially functional dystrophin. Although most cases of DMD and Becker MD are inherited, about one-third involve de novo mutations.
Strides on the Therapy Front
Sorting through the intricacies of mutations that cause DMD has led to several promising treatments for the disease that are edging closer to clinical practice. The most advanced approach involves exon-skipping, in which antisense oligomer drugs targeted at pre-messenger RNA induce the cellular machinery into bypassing mutated exons. This restores dystrophin, albeit in a shorter, partially functional, form (See Box, below). Phase II clinical trials evaluating two such agents showed promising results (N Engl J Med 2011;364:1513–22; Lancet 2011;378:595–605). In one, participants had new dystrophin expression and improved 6-minute walk tests. The other drug produced increased dystrophin and dystrophin-associated proteins in the sarcolemma, as well as decreased inflammatory infiltrate.
Another strategy about as far along as exon skipping involves suppressing stop codon mutations of the DMD gene. Drugs under development using this approach have shown preclinical safety and efficacy; however, one evaluated in a Phase IIa proof-of-concept study produced some conflicting clinical efficacy data. Considerable effort also is being placed on gene therapy involving administration of synthetic dystrophin genes. This strategy is further back in the development pipeline, though, in part because of difficulties finding appropriate vectors to deliver the DMD gene, the largest of all human genes, at about 2.5 Mbp.
“The dystrophin gene is huge and most gene therapy vectors are not able to hold the entire dystrophin coding sequence. Another challenge is how do you get it into the muscle? If we just needed the gene products to circulate in the blood, that would be a lot easier to do than get them into all the muscles,” explained Bidichandani.
Corticosteroids to reduce inflammation and slow muscle deterioration already have become a staple of the DMD treatment armamentarium. However, their well-known side-effects make them less-than-ideal for the DMD population. Therefore, another focus of drug development has been steroids with better side-effect profiles, as well as numerous other candidate therapies that would promote muscle regeneration, target oxidative stress, repair or protect membranes, alter protein turnover, or suppress inflammation or muscle atrophy. The state of therapy development in the field is so active that the journal Current Gene Therapy recently devoted an entire issue to the topic.
The Rise and Fall of DMD Newborn Screening
Newborn screening programs for DMD were implemented over the course of 2 decades in at least eight countries, but with the exception of Belgium, all have been discontinued. These models, including those of some states in the U.S., generally all involved a three-step process, with initial creatine kinase (CK) testing via bloodspot samples as a marker of muscle breakdown. CK testing from venous blood samples would be repeated 6–8 weeks later for babies with high baseline CK levels. DNA confirmation would then follow for those with persistently high CK concentrations. These programs ended due both to economics and the lack of any effective treatments for DMD, according to experts.
“There were multiple programs, but the issue is why screen if there’s no potential for therapeutic intervention? It becomes important to have some sort of therapeutic avenue when you’re telling someone their child has a devastating genetic disease and there’s nothing you can do about it. What sort of information are you providing in this context? That’s probably not enough,” explained Robert Weiss, PhD, professor of human genetics at the University of Utah in Salt Lake City. “That’s part of what squashed earlier screening programs. Economics might have been a driver too. But that component will be essential going forward, to have some other avenue available to these parents, other than prognostic information.”
A New Screening Strategy
The promise of and progress towards disease-modifying therapies during the past 5 years led a team including Weiss and career DMD investigator, Jerry Mendell, MD, to explore the feasibility of a novel newborn screening strategy (Ann Neurol 2012;71:304–313). “The potential for new therapeutic interventions for Duchenne is beginning to accelerate. Our thought was, do we need to reassess screening at this point if there are going to be novel therapies clinically available? That way, if there was diagnosis at birth, providing therapies to affected patients would have a uniform path,” Weiss explained.
He and Mendell were convinced they needed to completely rethink screening paradigms of the past in favor of a model better suited to the U.S. healthcare system. “In many European countries where they have a universal healthcare system, babies come back for a mandated six-week check-up. In this country with our private practice model, that’s not possible to do, so the systems in other countries won’t work in the U.S.,” explained Mendell. “Here, people go back to their primary care physician who is disconnected from the screening process and whether this individual orders any further testing would be discretionary. Further, there are some infants who don’t have the privilege of having a primary care physician so they might never get back for further testing—there are all kinds of variables and problems. We decided from the very beginning to identify a CK level that was elevated within 24–48 hours of birth, as virtually all newborn screening in the U.S. is done within that timeframe before the mother and baby leave the hospital.” Mendell is Curran-Peters chair in pediatric research and professor of pediatrics and neurology at Nationwide Children’s Hospital and The Ohio State University in Columbus.
Researchers at the Centers for Disease Control and Prevention, Georgia Department of Public Health, and Emory University recently studied the desirability, effectiveness of the consent process, and feasibility of conducting infant-based screening in the U.S. (PLoS Currents Muscular Dystrophy. 2012 doi: 10.1371/4f99c5654147a). While they found parental and provider support for and feasibility of such an approach, the authors emphasized that larger studies and more public discussion would be needed before DMD infant screening in pediatric offices can become a reality.
The model the Mendell team envisioned involves just two steps: initial CK testing from dried blood spots obtained as part of normal newborn screening, followed by DNA testing on those specimens above a pre-established cutoff. The researchers took a series of steps to test this approach, starting with a population-based analysis of more than 30,000 consecutive blood spot samples to determine CK levels in newborns. This analysis found the overall mean CK level to be 247.92 U/I, ranging from 253.37 U/I in samples taken from babies <48 hours in age to 201.64 U/I in those more than >120 hours old, and 251.53 U/I in males versus 246.39 U/I in females. Based upon this data, initially the researchers went with a cutoff of 600 U/I, which is three standard deviations above the mean. The goal was to set the level high enough to exclude any babies with elevated CK incidental to birth trauma, while not missing any boys with DMD.
Next the authors conducted pilot newborn screening in three sequentially larger populations, including 6,928 samples from major birthing hospitals in Columbus and Cincinnati, followed by an additional 10,937 samples from hospitals state-wide. The final phase of testing involved 19,884 de-identified newborn samples obtained in partnership with the Ohio Department of Health, for a grand total of 37,749 samples tested.
On a parallel track, Weiss developed the DNA sequencing and mutation analysis protocols for whole genome amplification followed by mutational analysis using multiplex ligation-dependent probe amplification (MLPA). “We had established the techniques we used on DNA purified from venous blood and had a fairly large testing database so that we knew the mutation spectrum of the DMD gene very well,” recalled Weiss. “We knew what class of mutations to look for, so the analysis went through a first stage of copy number analysis to detect whole exon deletions and that was done using MLPA, a tried-and-true assay. Those that were mutation-negative for deletion or duplication were then sequenced, and we did it with fairly conventional techniques looking for point mutations. The key thing was getting a uniform yield of DNA off the blood spots. I think that’s routine enough that the more recent developments in DNA sequencing technology can be applied to this same material, probably with the same results.” Weiss validated his method using blinded bloodspot samples from boys with diagnosed DMD.
What’s the Best CK Cutoff?
After the pilot study using a CK cutoff of 600 U/I identified 110 samples requiring DNA analysis, of which only two were found to have known DMD mutations, the authors looked again at the cutoff and used ≥750 U/I for the subsequent two phases, which cut by 68% the samples needing DNA analysis (See Table, below). Significantly, all six newborns found to have DMD genetic mutations had CK levels >2,000 U/I, leading the researchers to suggest an even higher cutoff—≥1,000 U/I—should DMD newborn screening be implemented in practice.
Duchenne Muscular Dystrophy Newborn Screening Results
|Statewide Dept of Public Health
|Adapted from Neurology Today, March 1, 2012|
“The real surprise of the study was that all the mutations we found were in the extreme tail of the distribution, above 2,000 U/I. That is a novel observation, because if the mutations were scattered throughout the range, it means one would have to do the DNA analysis on a much larger number of samples which means the cost might become prohibitive,” said Weiss. “However, this suggests that for the DNA analysis, one might fairly confidently concentrate on the tail of the distribution and still find a large majority of affected individuals.”
Nine other samples, from seven males and two females, had CK levels >2,000 U/I but no known DMD mutations. The researchers analyzed these samples further for common limb girdle MD genes and found one mutant allele each in three individuals, including one female and two males. These findings would not surprise most DMD specialists, according to James Dowling, MD, assistant professor of pediatric neurology at CS Mott Children’s Hospital in Ann Arbor, Mich.
“CKs are elevated in Duchenne from birth and the number of cases the authors found are thought to approximate the prevalence of the condition. The finding of a few individuals with high CK who don’t have Duchenne makes perfect sense to me because there are other forms of muscular dystrophy that can present with high CK in infancy,” he observed. “In my experience, a cutoff of 1,000 U/I should pickup very few false positives. It’s also probably unlikely to have false negatives because the authors’ data supports what’s been known about older children and diagnosis, which is they all have very high CKs. In fact, if a two year old boy with muscle weakness was referred to my clinic and his CK was only 600 U/I, Duchenne would be extremely low in my thoughts in terms of further diagnostic work-up.”
The process of both CK and DNA testing used three blood spot samples, out of the five normally collected in Ohio as part of the state’s newborn screening program. The researchers calculated that the CK assay added $1 in raw materials to the costs already associated with screening for 35 conditions in Ohio. The cost of raw materials for DNA testing was $150.
On the Road to Newborn Screening
Mendell’s findings have captured the attention of researchers worldwide and raised hopes that newborn screening for DMD once again might be introduced. Historically one of the criteria used to consider a condition for newborn screening included that it have an acceptable treatment protocol in place that changes outcomes for babies, and a recent CDC report on newborn screening outcomes emphasized that screening for a condition alone is not enough (MMWR 2012;61:390–393). However, given the progress on the therapy front, the muscular dystrophy community believes it has a good chance in the near future of convincing the influential Secretary’s Advisory Committee on Heritable Disorders in Newborns and Children to endorse newborn screening for DMD. States rely on the committee’s findings in making their own screening decisions.
With that in mind, the MDA is sponsoring an invitation-only symposium on September 12 to hash-out the state of science in the field. “We’re hosting a meeting of all the leading opinion leaders in the field to look at whether the technology used is state-of-the-art or whether there is something else available that’s even better. We also want to evaluate the current status of disease-modifying therapies—how real are they, and is there data to show that early intervention will help. We’ll have all these view points to assess whether we’re ready to nominate DMD to the Secretary’s panel and how the panel in turn might view the nomination package,” said Bidichandani.
The MDA expects to submit proceedings of this meeting for publication. A similar conference is planned in Europe in December. “Our overall aim is to review the current practices and procedures surrounding this complex issue, including the new techniques for screening, such as Dr. Mendell’s two-tier approach, and to hear views of parents, so as to draw up an action plan to develop a standard operating procedure for DMD newborn screening,” wrote conference co-organizer, Juliet Ellis, PhD, clinical study research coordinator at the Institute of Child Health at University College, London.
Australian researchers also are interested in implementing newborn screening for DMD, and have started a pilot project in New South Wales to determine normal CK levels from bloodspot samples collected in Australian screening programs.
Mendell, perhaps more than anyone else, welcomes these developments. “I think we have a good shot at getting newborn screening implemented, even in the context of hopeful treatments. That’s one of the reasons I’m still around. One thing I want to do before I leave this profession is to see newborn screening implemented for Duchenne. I’ve seen too many families wiped out by this disease.”