College of American Pathologists
Printable Version

  Untangling the knot of mitochondrial disease


CAP Today




September 2011
Feature Story

Karen Titus

Experts in mitochondrial disease are a persevering lot. Like Tristan and Isolde filling their lungs for Act 2, they’re in it for the long haul.

“It’s a field I’ve been in for 35 years,” says Michael J. Bennett, PhD, professor of pathology and laboratory medicine, Perelman School of Medicine at the University of Pennsylvania, and director, Metabolic Disease Laboratory, Children’s Hospital of Philadelphia. “And we’re all still learning, basically.”

“It’s a difficult field,” agrees Brian Robinson, PhD, director of the mitochondrial research laboratory, Hospital for Sick Children (SickKids), Toronto. “Even though I’ve been in it for 30 years, it’s hard for me to figure out where it’s headed.”

“I’ve been at this now for 40 years,” says Douglas C. Wallace, PhD, the Michael and Charles Barnett Chair of Pediatric Mitochondrial Medicine and Metabolic Disease, and director, Center for Mitochondrial and Epigenomic Medicine, CHOP. “When I started, nobody thought that studying mitochondrial DNA would be of any interest clinically.” Only in recent years has medicine done an about-face on the matter, says Dr. Wallace, who likes to talk about the need to upend 150 years of Mendelian genetics in order to arrive at a proper diagnosis of mitochondrial disease.

If those are the experts talking, what are the prospects for physicians whose outlooks are less mitochondrial-centric?

The pace may be about to pick up, actually. While there’s nothing easy about understanding and diagnosing mitochondrial disease—especially in children—breakthroughs appear imminent and could put laboratories in the thick of mitochondrial diagnostics. “These diseases are extraordinarily common, not rare,” says Dr. Wallace, who is also professor, Department of Pathology and Laboratory Medicine, Perelman School of Medicine at the University of Pennsylvania. (A common estimate is one in 4,000 or 5,000 lifetime risk, though others, including those interviewed for this article, suggest the risk may be one in 2,000, or even less, for children.) “So the need for this testing is going to be great in almost every path lab.”

Mitochondrial disorders are the Willy Lomans of disease—difficult to fathom, easy to overlook, dropping hints that don’t quite add up to a full portrait. But now, attention must be paid.

Dr. Robinson hadn’t planned on making mitochondrial disease his professional raison d’être. He fell into it almost by chance three decades ago when he was hired by the Hospital of Sick Children to look at why babies were developing hypoglycemia in the neonatal period.

Dr. Robinson, who trained as a biochemist, got busy with specific projects involving animal models. But early in his tenure, babies bumped the animals aside. “The clinicians kept coming to me, saying, ‘We’ve got these babies with lactic acidosis, and we don’t know what to do with them,’” recalls Dr. Robinson. He found himself trying to help his colleagues sort through a maze of individual metabolic defects and soon set up testing in his own laboratory.

In many ways, he’s not much further than where he started. Such is the nature of diagnosing mitochondrial disease. “One of my colleagues here says, ‘Mitochondrial disease can present at any age, any time, in any organ, with any set of symptoms,’” Dr. Robinson says. “And it’s true.”

There’s been progress, of course. Dr. Robinson and other leaders in the field initially identified biochemical defects related to mitochondrial disease, then began identifying the underlying genetic defects. Next-generation sequencing has sped up that process. (See “Next-gen sequencing in clinical debuts,” CAP TODAY, April 2011.) But while genetics has stormed all of medicine, mitochondrial disease has its own cruel twist: Diagnoses lie tangled within mutations both in mitochondrial DNA and in the nuclear genes. If genetics is a revolution, mitochondrial disease is an uprising within. Pathologists, it would appear, need to topple the king and set fire to the palace.

They’ve got their work cut out for them. “Mitochondrial disorder may be the most difficult disease to diagnose,” says Rong Mao, MD, medical director, molecular genetics and genomics, ARUP Laboratories, and assistant professor of pathology, University of Utah School of Medicine, Salt Lake City.

Why is it so difficult? Pull up a chair.

For starters, mitochondrial disease is given short shrift in medical training—understandably so. It’s hard to teach what even the experts are only now starting to learn. Most physicians probably know that mitochondrial disease can be caused by mutations in the mitochondrial DNA, says Dr. Mao. “But the mutations in the nuclear genes are something new,” she says, with the mitochondrial genome encompassing not only the mitochondrial DNA but more than a thousand nuclear DNA genes dispersed throughout the chromosomes.

On the biochemical end, the tests most physicians know about come up short. Says Dr. Bennett: “We have no good biomarkers whatsoever.”

The lineup of traditional tests includes lactate and pyruvate analysis in blood and cerebrospinal fluid; plasma and CSF amino acids; urine organic acids; and plasma acylcarnitines. “All those provide clues without giving a diagnosis,” says Dr. Bennett. (See Haas RH, et al. The in-depth evaluation of suspected mitochondrial disease. Mol Genet Metab. 2008;94:16–37; and Haas RH, et al. Mitochondrial disease: A practical approach for primary care physicians. Pediatrics. 2007;120:1326–1333.)

Lactate acid elevation should arouse suspicions, but it’s not specific for mitochondrial disease. And some patients with mitochondrial disease may have normal lactate and pyruvate levels. A plasma amino assay might show an elevated alanine level, but again, this is a clue, not a diagnostic marker, Dr. Bennett says, since anyone with lactic acidemia will have an elevated alanine. “For the clinician, it should say, ‘At least we’re on the right track.’ But these are soft clues without being particularly diagnostic,” Dr. Bennett says.

Are clinicians good at spotting these clues? “Physicians are getting better,” Dr. Bennett says. “There are some outstanding physicians out there. But it’s a very, very difficult area.”

Muscle biopsies have their own difficulties, especially in pediatric patients. There are the occasional slam-dunk cases—for example, ragged red fibers are a hallmark of mitochondrial myopathy. But generally, aimless dribbling seems to be the order of the day, particularly in the pediatric age group, where clues are much more subtle. “Probably the most common thing to find on muscle histology in pediatrics is nothing,” says Richard H. Haas, MD, professor, neurosciences and pediatrics, and director, UCSD Mitochondrial Disease Laboratory, UCSD Medical Center, La Jolla, Calif.

“Then that blends into a few patients who do have ragged red fibers,” Dr. Haas continues. “And in between you have patients with some evidence of mitochondrial proliferation. That middle group can be missed by people who aren’t clued into this.”

It doesn’t help that effective treatments are scarce. Staple treatments include high doses of B vitamins and coenzyme Q10. EPI-743 (Edison Pharmaceuticals) looks promising in very early trials, says Dr. Haas, and MELAS patients are generally put on arginine or citrulline. Other approaches use antioxidants. Says Dr. Haas: “In the next few years we’re going to be in a much better position in terms of having effective treatments for mitochondrial disease. We’re not really there yet.”

Even the phrase “mitochondrial disease” is suspect—it’s a lumpy, almost useless definition, like saying someone has “cancer.” “They’re heterogeneous diseases clinically and biochemically,” says Dr. Haas. Plotting out mitochondrial disease, with all its possible destinations, would produce an image resembling the iconic London Underground map, with its simplistic, geographically misleading routes. “We think in terms of primary genetic disease, and also secondary mitochondrial disease, which is another major category where there’s evidence of mitochondrial dysfunction. And neurodegenerative, later-onset disease—Alzheimer’s and Parkinson’s disease—may have a combination of genetic predisposition and secondary mitochondrial failure. Type 2 diabetes has a major mitochondrial component to it,” Dr. Haas says. In fact, “Almost any disease you care to mention, because mitochondria is so critical in metabolism, has some mitochondrial component.”

Nonetheless, Dr. Haas says, physicians are getting better at considering mitochondrial disease. “They all know about it now. Ten years ago that wasn’t the case.”

Dr. Bennett suggests that physicians fail to detect cases because of the difficulties in attaching clinical clues to certain phenotypes. That will change as the underlying genetic defects come into sharper focus. “But it’s been a slow molecular slog,” he says, given the poor correlation between phenotype and genotype. “It’s been a matter of working through the genes we know are associated with mitochondrial function. It’s almost like a fishing expedition.”

Dr. Wallace offers up Kearns-Sayre syndrome by way of explanation. The disease is severe, with onset typically before age 20, and is due to a deletion in the mitochondrial DNA. Because the mitochondrial DNA is inherited independent of the nuclear genes, however, and can be gained or lost during mitosis as well as meiosis, different types of tissue have different mitochondrial DNA genotypes, Dr. Wallace explains. “The muscle, the brain, the kidney—all the postmitotic organs have very high levels of this mutant, and that causes the disease.”

The mutant, he continues, is selected against in the bone marrow stem cells. There is no deletion in blood.

This point has not always been understood, Dr. Wallace says. “Even today, pathologists ask me to rule out Kearns-Sayre and send me a blood sample. That simply can’t work.”

Dr. Wallace sees a bigger problem beyond this specific example. For him, the walls of Jericho need to tumble—and he’s not shy about blowing the trumpet. “There’s a real disconnect between this totally new genetics and biology, which is never taught to physicians, and the diagnostic tools that we need to address it,” he says. “We’re constantly asked to do things that might make sense from the traditional way of doing molecular pathology; for nuclear genes, it’s actually not applicable to this novel set of genes.”

That gap is only the first of several layers of complication that bedevil the diagnoses of mitochondrial disease, Dr. Wallace says. Understanding mitochondrial disease, as it turns out, quickly becomes a trip with as many twists as San Francisco’s Lombard Street.

The second difficulty, he says, is that each cell contains thousands of copies of this mitochondrial DNA, whereas nuclear DNA contains only two copies. Thus, people with a mitochondrial DNA mutation can have a mixture of mutant and normal mitochondrial DNAs, a state known as heteroplasmy.

Genetic testing for nuclear DNA gene mutations is relatively simple since there are only three states: plus-plus, plus-minus, and “dead” (minus-minus). The possibilities explode for mitochondrial genotypes. Mutant levels could be low (0.1, one, three, five percent) or high (90, 95, 100 percent) or anywhere in between. “And different percentages of the same mutant give you totally different clinical presentations,” Dr. Wallace says. For example, a mutation in the ATPase 6 gene of mitochondrial DNA at nucleotide position 8993, with a 75 percent mutation level in the body, will present as mild, so-called salt and pepper retinitis pigmentosa, Dr. Wallace says. Higher levels of mutation lead to macular degeneration and movement disorders. But at 95 percent, the result is Leigh syndrome.

In other words, he explains, the exact same mutation can result in very different clinical phenotypes depending simply on the percentage heteroplasmy: subclinical or with late adult onset, clinical with young adult onset, or severe with early pediatric death.

For another heteroplasmic mitochondrial DNA mutation, the MELAS tRNA mutation, high levels of mutant can cause stroke-like symptoms, cardiomyopathy, and/or myopathy. However, at low levels of heteroplasmy, this same mutant causes type 2 diabetes.

The heteroplasmy problem is particularly difficult for the mitochondrial medicine laboratory, Dr. Wallace says, “since someone can send me a blood sample to test for the MERRF mutation [which can cause myoclonic epilepsy], which is caused by a heteroplasmic mitochondrial DNA tRNA mutation. The mutation may be at 10 percent in the blood, but at 80 percent in the brain. So now you have this total disconnect between the numbers I give the physician, from the clinical lab, and the functional and genetic effect on the target organ.”

The third major diagnostic dilemma is that the same clinical phenotype can be caused by mutations in totally different genes in the mitochondrial genome. For example, the severe childhood disease Leigh syndrome is now known to be caused by both mitochondrial DNA or nuclear DNA mutations in genes of the mitochondrial genome.

This is a striking difference from what most physicians were taught about nuclear mutations, with its linear approach: one gene, one polypeptide, one phenotype, Dr. Wallace says. With mitochondrial mutations, the same mutation can produce multiple phenotypes, simply from this statistical percentage of different mitochondrial DNAs. “That has been extraordinarily difficult for Western medicine, which is based on the differential diagnosis, the idea that there are given clinical presentations for given disease entities.” In the case of mitochondrial disease, that approach is starting to look like a rusty one.

In short, every turn is different from the one physicians are used to taking. The road Dr. Wallace says needs to be taken is stochastic, not deterministic. “This has baffled the diagnostic systems of this country.”

Not in other countries? Physicians in other countries are wrestling with it, Dr. Wallace concedes, but the paradigm in Western medicine is particularly difficult to transcend, he says. “Our clinical specialties, with the exception of the pediatrician, the internist, and the geneticist, are primarily based on anatomy. You’re either a neurologist or a cardiologist or an ophthalmologist, etc.” In China, by contrast, traditional medicine is based on “chi,” the concept that the body’s energy state causes symptoms. “The diseases we’re talking about are diseases of energy flow, much more aligned with the concept of chi,” Dr. Wallace says. “Since energy flow is essential for every organ in the body, but different organs of the body rely on energy flow to different extents, depending on the activity level, then you get systemic effects, genetic effects, that have organ-specific phenotypes.

“This is problematic for the way many people are trained in medicine in the U.S.,” he continues. “If there’s a symptom in a particular organ, then you tend to think of it as an organ-specific disease. We think of Alzheimer’s disease as a brain disease. But in fact we and others have published papers showing that Alz-heimer’s disease is a systemic mito-chondrial defect with tissue-specific symptomatology.”

For all the complexities, Dr. Wallace does not see an insurmountable problem when he looks at the future of mitochondrial diagnostics. The concepts aren’t hard to understand, he says. “But they are different.”

Biochemistry, typically the entree for most mitochondrial disease workups, is its own obstacle course. Looking to biochemistry to provide answers about mitochondrial dysfunction is “even more problematic” than the aforementioned DNA work, says Dr. Wallace.

Current biochemisty tests might raise suspicions, but don’t expect that flag to snap smartly in the breeze, signaling mitochondrial disease. The testing tends to be reduced to a simplistic formula, says Dr. Wallace: that you can tell mitochondrial diseases by elevated lactate, or by pyruvate dehydrogenase deficiencies by the lactate-pyruvate ratio.

This is “mildly true,” says Dr. Wallace. “But those are very, very insensitive approaches.” Leber’s hereditary optic neuropathy, though a mitochondrial DNA disease, rarely if ever has elevated lactate or altered lactate-pyruvate, he says; moreover, a muscle biopsy will show no altered mitochondrial morphology.

So at one end of the range are clinically relevant mutations that are so mild they pass under the variance of traditional assays, he says; at the other are devastating diseases, such as a severe Kearns-Sayre syndrome, which will show high levels of lactate in the blood, cerebrospinal fluid, and urine. “But those are the rare forms,” Dr. Wallace says.

Dr. Bennett says he and his colleagues have developed an assay for measuring tissue levels of acyl-CoA. But better markers are not easy pickings.

“Some years ago,” says Dr. Bennett, “we thought we had some clues toward defects of complex I of the respiratory chain,” which essentially converts NADH into NAD for recycling. The NAD becomes limiting in complex I deficiency, “so what happened is it affected other metabolic pathways that actually require NAD. And so we’ve got some secondary clues based on metabolites that pointed to other diseases but in fact were primarily mitochondrial.” To Dr. Bennett, this suggests there may be individual biomarkers for some of the mitochondrial diseases, and that the respiratory chain complexes—which relate to different parts of metabolism—might point the way. “But we’re certainly not there yet.”

Clinicians may not be aware of the many subtleties that can affect test results, Dr. Haas says. Some examples:

  • Pyruvate is an unstable compound, a concern because it’s often done as a send-out test, he says.
  • If plasma amino acids are collected in a postprandial state, the elevated results can lead to false-positives.
  • Patients who take carnitine will have a general elevation in acylcarnitine derivatives.
  • Some drugs, including Valproate, will impair fatty acid oxidation, leading to a fatty acid oxidation defect in the acylcarnitine profile.
  • Patients who are fasting tend to have elevated dicarboxylic acids and beta-hydroxybutyrate in the organic acids, which will also be reflected in the acylcarnitine profile.

“I don’t think there’s anywhere near the awareness of any of this that there should be,” says Dr. Haas.

Most biochemical testing is done in academically based biochemical genetics labs, Dr. Haas says, as well as in commercial laboratories such as ARUP and Quest. But even with send-outs, labs can help their colleagues interpret confusing results. Inconclusive results aren’t unusual during a diagnostic workup, Dr. Mao notes, which clinicians find unsettling. “They want to know how to interpret ‘equivocal,’” she says.

“They wind up not knowing what to do,” agrees Dr. Haas. “Often that generates a referral to us, often for things that are pretty trivial. We see patients where there’s a question of whether they have severe disease—and in fact they were simply fasting when the sample was collected. Or eating Jell-O,” Dr. Haas says with a laugh, noting that gelatin can cause adipic acid elevation in organic acids.

Dr. Bennett says clinicians might not fully appreciate the complexities of mitochondrial workups, or that interpretations are heavily dependent on skills of the interpreter. Many of the tests are in-house assays, with all the variability that implies. “Something I learned a long time ago is, if a physician suspects something strongly enough, and the tests don’t give them the right answer, they should at least repeat it and maybe go somewhere else and see if they can get a different interpretation, particularly in this field, where interpretation is open to different skill sets,” Dr. Bennett says.

All this assumes that clinicians are even thinking to look for a mitochondrial disorder. Dr. Bennett suggests labs might be able to help raise the concern, but, mitochondrial disease being what it is, it resists facile advice, such as “start testing sooner.” That, he says, could cause clinicians to pile on unnecessary tests. He wants labs to raise the flag of suspicion—but not too high, nor too quickly.

Typically, he says, patients with a mitochondrial disease first work their way through other subspecialties. Only at the end of the line, when no clear answer is forthcoming, will someone suggest a mitochondrial culprit. “If it’s a GI presentation they would have been through the GI docs. Or, it’s a neurological patient who goes through all the right testing for, say, a seizure disorder, but they don’t get an answer. Then I think the thought begins to occur that it could be metabolic.”

That’s the sort of thinking Dr. Bennett would like to change. “People ought to be aware that it’s a possibility from a clinical perspective, not something you suddenly realize you haven’t done at the end.”

Perhaps the answer will lie in the happy, and apparently inevitable, confluence of next-generation sequencing and (to use Dr. Mao’s phrase) next-generation pathologists.

Next-generation sequencing is the topic of lively discussions in pathology departments and universities, she says, since everyone has recognized the need to train residents and fellows in interpreting next-gen sequencing results. Ditto, she says, for the Association for Molecular Pathology, which is discussing educational guidelines for teaching the new technology.

Any lab that’s currently embracing next-generation sequencing is on the right path, Dr. Wallace says. “That obviously is going to be a tremendous benefit for looking at these thousand-gene nuclear problem sets,” he says.

But the answers still won’t come easily—next-generation sequencing won’t be turnkey. For mitochondrial DNA, “It would be tempting to say, ‘Oh, we’ll just throw it onto next-generation sequencing,” Dr. Wallace says. But physicians need to be mindful that the tissue of interest has its own unique genotype. “So just throwing blood samples onto your automated sequencer to give you a result could well be misleading.”

Researchers are beginning to look at other, accessible tissues as surrogates, he says. Urinary cell sediment sometimes contains mitochondrial DNA mutations that are not found in blood. Ditto for buccal swab tissue, as well as hair follicles. “But in the end, if it’s a neuromuscular disease, you probably are going to many times be driven to actually get at least a small sample of muscle, say through a needle biopsy, if you’re actually going to do a molecular diagnosis.”

And there’s nothing easy about tissue biopsies for mitochondrial diagnoses, especially in pediatric patients. Those with mitochondrial disease can be hypersensitive to the adverse effects of anesthesia, Dr. Wallace says; moreover, the large volume of tissue required can be disfiguring in small patients.

At CHOP, Dr. Wallace and his colleagues are building microchamber analysis tools, which will permit them to take a tiny segment of muscle from a needle biopsy, then perform traditional biochemical assays on one-hundredth the sample previously required.

Dr. Haas and colleagues are exploring new methods to extract useful information from muscle biopsies, using a high-resolution respirometer to look at oxygen consumption with different substrates of live mitochondria, using either intact tissue or with isolated mitochondria.

They’re also looking at electron transport assays, again using either a frozen tissue sample or isolated mitochondria. These assays fall prey to lab-to-lab variability, Dr. Haas says. “They’re not developed in a way CAP would like it, I’m sure, but it’s kind of the state of the art at the moment.”

Despite the abundant hurdles, a build-it-and-they-will-come hopefulness seems to prevail when it comes to next-generation sequencing. It’s almost a foregone conclusion that the technology will eventually allow physicians to obtain a molecular diagnosis and confirmation of disease in the majority of patients. Some companies already offer next-generation sequencing of candidate mitochondrial genes. Dr. Haas says once this technology is brought to bear on mitochondrial disease, “The whole world’s going to change.”

But on whose dime? “It’s incredibly expensive,” Dr. Haas warns.

Dr. Robinson sees a silver lining in that dark cloud, predicting a push to start, rather than end, with a genetic diagnosis. “The administrative people who look at costs will say, ‘This is going to be cheaper than doing a muscle biopsy and getting the pathology and so on,’” he says. “That actually might be quite an efficacious way of doing things.

“But,” he says, “it’s going to take awhile to sort it all out.”

For mitochondrial disease—a disease that offers minimal diagnostic clues, few treatments, and an equally perplexing (and expensive) future—that’s just par for the course.

Karen Titus is CAP TODAY contributing editor and co-managing editor.