David S Millington. American Scientist. Volume 90, Issue 1. Jan/Feb 2002.
Metabolic diseases arise from inherited defects in enzymes involved in the production of energy, a process that is essential for the well-being of cells in every organ of a healthy person. Most of these “inborn errors of metabolism” occur when a child inherits two copies of a defective gene, one from each parent. The parents are normally unaware that they carry defective genes because these diseases are recessive: Functional copies of these genes mask the effects of defective copies. When both parents carry a defective copy of the same gene, each of the children they conceive carries a 1-in-4 risk of being affected with the metabolic disease. Individual metabolic diseases are rare-ranging in frequency from about 1 in 10,000 to more than 1 in 1,000,000. But hundreds if not thousands of enzymes are involved in normal metabolism, so, collectively, these diseases account for a significant fraction of chronic illness and death in infancy. The March of Dimes estimates that metabolic disorders occur in at least 1 in 3,500 newborns.
The symptoms of metabolic diseases are easily confused with much more common conditions. Consequently, their diagnosis is challenging even to specialists. But correct diagnosis is essential for appropriate treatment, without which tragic outcomes are all too common.
Public awareness of metabolic diseases was all but unknown in the United States until 1964, when widespread neonatal testing began for phenylketonuria, or PKU, a disease that causes profound and irreversible mental retardation when not treated. Over the decades since then, most states have expanded screening to cover but a handful of additional diseases. Newborn screening involves pricking a newborn’s heel and collecting blood from the small wound on cotton fiber paper. The paper carrying the dried blood spots-commonly known as a “PKU card” or a “Guthrie card,” after its inventor-then goes to a state laboratory, which typically screens for PKU and a handful of other diseases.
The case of PKU screening exemplifies the benefits of early diagnosis of a metabolic disease. The case is particularly strong because of the decades of experience showing its benefits all over the world. PKU arises through the deficiency of an enzyme that converts the amino add phenylalanine to the amino acid tyrosine. The original test was a simple and inexpensive assay that detected the presence of excess phenylalanine when it inhibits the growth of a species of bacterium; many newborn screening laboratories still use this method. When a baby found to have excess phenylalanine immediately goes on a diet that restricts phenylalanine intake, the child grows essentially normally.
The benefits of such testing are obvious for individuals and families who might otherwise suffer from the worst effects of PKU, but the benefits to society are also large and have driven the expansion of testing for other diseases. Profoundly mentally retarded children often spend their lives in institutions and hospitals. PKU itself affects only about 1 in 23,000 newborns annually in the United States, but the benefits of finding and treating these cases far outweigh the costs of screening the entire population.
Unfortunately, the scope of newborn screening has not increased rapidly since 1964, so for the vast majority of rare diseases, an affected child might visit hospital emergency rooms several times and consult repeatedly with specialist physicians before being diagnosed correctly. Effective treatment cannot begin without diagnosis. Without question, the best time to obtain a diagnosis is before the onset of symptoms. Even delaying diagnosis by a week could, in certain instances, lead to permanent mental retardation or even death. This is particularly heartbreaking because for many metabolic diseases, the treatment is straightforward: Make sure the child does not fast. Preventing fasting keeps the child from entering a catabolic state wherein he or she breaks down molecules stored in the body, such as proteins, triglycerides (fats) and glycogen, to compensate for the lack of energy sources from the diet. If a baby who is unable to process a metabolic intermediate goes long without food for whatever reason, catabolism causes the accumulation of toxic by-products. This, in turn, can lead to symptoms such as vomiting, diarrhea, lethargy and even coma. The most effective immediate treatment is intravenous glucose therapy.
Because of the episodic nature of metabolic crises and the similarity of these symptoms to those of intentional poisoning, parents have been falsely accused of child abuse and even murder when health care providers did not understand that these children suffered from metabolic diseases. The severity of metabolic crises tends to ameliorate with increasing age. In infancy, the risks are greatest because the brain is still developing rapidly. Also, older children can increasingly regulate their own diets.
At present most states restrict newborn screening to diseases that follow the “PKU paradigm” of inexpensive detection of diseases for which dietary therapy can prevent dire outcomes. The slow development of new tests partly reflects the fact that each new test differed from previous ones in fundamental design and required a separate specimen. Therefore, each new test required a fresh infusion of resources, mainly personnel.
However, a small set of states are testing for more than 20 metabolic disorders in newborns, using methods developed in our laboratory at Duke University Medical Center. The method uses tandem mass spectrometry (MS/MS) to screen for many disorders using a single blood specimen. MS/MS can identify and quantify related metabolites simultaneously. Over time, the method will likely be used to detect an increasing number of metabolic conditions. MS/MS represents the second wave of technology for newborn screening. In the following paragraphs, I shall review the development of this technology and its impact on newborn screening.
The Metabolic Context
The driving force for applying tandem mass spectrometry to biomedicine was the need to analyze a class of compounds called the acylcarnitines, which accumulate from the defective breakdown of fatty acids and certain amino acids. The acylcarnitines are a homologous group of metabolites that differ only in the size or structure of the acyl group that is attached to the hydroxyl group of carnitine (L-3-hydroxy-4-aminobutyric acid). This common structural feature of acylcarnitines allows their selective detection and analysis by MS/ MS.
Acylcarnitines derive from the catabolism-or breakdown-of fatty acids and branched-chain amino acids; these chemical reactions occur primarily in the mitochondrion, the organelle that generates energy for the cell. During fasting, after the body has used up reserves of glucose, fatty acids become the main energy source. Long-chain fatty adds can only enter the inner mitochondrion by attaching to carnitine. Several enzymes are required to accomplish this. Inside the mitochondrion’s inner matrix, the longchain fatty acyl groups transfer from carnitine to coenzyme A (coA) and degrade by a process called beta oxidation. Each cycle of beta oxidation requires at least four enzymes that shorten the carbon chains in fatty acids by two carbons at a time, starting typically from 16- or 18-carbon chains.
The catabolism of the branched-chain amino acids leucine, isoleucine and valine also takes place primarily in the mitochondrion. Again, numerous enzymes are involved in these metabolic pathways. Normally, the catabolism of branched-chain amino acids and of fatty acids proceeds smoothly to the end products. However, if an enzyme in one of the pathways is defective or missing, abnormal acyl-coA intermediates accumulate, possibly with life-threatening effects or leading to permanent neurological damage. Like PKU, these disorders of catabolism may require the restriction of proteins or fats in the diet. However, the toxic intermediates accumulate in largest amounts when the body’s stored proteins and fats are broken down to provide energy during a prolonged period of fasting.
When abnormally high levels of acyl-coA intermediates accumulate, they remain trapped inside the mitochondrion, but the coA groups can exchange with carnitine, and the resulting acylcarnitines can cross the mitochondrial membranes and leave the cell. It is, therefore, possible to analyze the acylcarnitines in blood to diagnose more than a dozen defects of fatty acid and branched-chain amino acid catabolism. The most common of these disorders is medium-chain acyl-coA dehydrogenase (MCAD) deficiency, a potentially life-threatening disease if undiagnosed, but one for which early diagnosis and straightforward dietary therapy can avert illness.
The disorder is more common than PKU. Thus, MCAD deficiency satisfies the conventional criteria for inclusion into newborn-screening programs, and being able to diagnose newborns susceptible to this particular illness has helped spur the introduction of MS/MS into screening programs. Without proper diagnosis, children with MCAD deficiency and related disorders can be erroneously diagnosed as having “Reye-like” syndrome, which occurs in some children who take aspirin, because like Reye syndrome, metabolic illness can lead to cerebral edema. Or, because children with metabolic diseases sometimes stop breathing and die in the crib, they are mislabeled as having SIDS, or sudden infant death syndrome.
At this point, MS/MS can effectively diagnose a range of metabolic disorders related to the breakdown of fatty adds and certain amino acids. Investigators at Duke are also investigating methods using MS /MS to identify diseases involving defective glycogen metabolism-commonly called glycogen-storage diseases. The recent development of a viable therapy for one of these diseases-providing children with the enzyme they lack-now drives the effort to find effective means of diagnosing these diseases.
Tandem Mass Spectrometry
The biomedical applications of MS/MS that are leading to expanded newborn screening were developed at Duke University between 1983 and 1990. Organic mass spectrometry dates from the 1940s, when it was first used in the petroleum industry, but research instruments did not become widely available until the 1960s. Mass spectrometry was a solution waiting for problems, and biologists seldom used the technique until well into the 1980s.
The simplest definition of a mass spectrometer is a device that ionizes molecules, some of which break up spontaneously into fragments, and then separates the charged products, or ions, according to their mass-to-charge (m/z) ratios. A plot of the m/z values against relative abundance or intensity is called a mass spectrum. The mass spectrum of a molecule carries important structural information as well as indicating the molecular weight. Mass spectrometers usually have one ion source and one mass analyzer. When linked to a separation device such as a gas chromatograph (GC) or liquid chromatograph (LC), mass spectrometers are very powerful analytical tools for the analysis of complex mixtures. However, a typical GC/MS or LC/MS analysis requires 30-60 minutes, which prevents their application on a large scale for programs like neonatal screening. Tandem mass spectrometry, on the other hand, can analyze a sample in a matter of seconds.
Tandem mass spectrometry was developed in the 1970s as an alternative means of analyzing complex mixtures, and became popular with the development of the triple quadrupole mass spectrometer. This instrument has two mass analyzers separated by a “reaction cell.” The reaction cell contains an inert gas, usually argon, whose molecules collide with ions entering the cell from the first analyzer. These collisions increase the internal energies of the ions, causing some of them to fragment. The second mass analyzer then separates and detects the charged products. The first mass analyzer can be tuned to select ions derived from a particular component of a mixture that can be identified from its fragments. The advantage of MS/MS for analyzing mixtures is that the entire analytical process takes a fleetingly short time. Using a “soft” ionization method simplifies the analysis, because it produces predominantly intact molecular ions from the mixture. The most popular of the soft ionization methods is electrospray ionization, because it enables specimens to be introduced in solution.
To diagnose disorders of fatty acid and branched-chain amino acid catabolism, the acylcarnitines are chemically converted to their butyl esters. Electrospray ionization produces mostly intact molecular ions that undergo fragmentation in the reaction cell to produce a common fragment of m/z 85. Fixing the second mass analyzer to transmit only ions of m/z 85 and scanning the output of the first analyzer over the mass range 200-500, which includes the molecular weights of all the acylcarnitines, allows a computer to generate a special type of mass spectrum that consists of the molecular masses of individual acylcarnitines and their relative abundance. We refer to this as a “metabolic profile” of acylcarnitines. Adding fixed amounts of stable isotope-labeled analogues of selected acylcarnitines to the sample before the analysis allows each of the components to be quantified accurately. Other components in blood do not share the unique structural characteristics of acylcarnitines, and therefore the special scan function of the MS/MS allows the other components to remain undetected.
Figures 4a and 4b show, respectively, acylcarnitine profiles from a normal newborn’s blood and from one with MCAD deficiency. The normal profile shows primarily acetylcarnitine (carnitine attached to two carbons, denoted C2), the major mitochondrial product of amino acid and fatty acid catabolism. The patient with MCAD deficiency shows marked elevation of medium-chain acylcarnitines, especially eight-carbon chains, reflecting the enzyme defect. This profile is highly characteristic of MCAD deficiency, independent of the genetic mutations that cause the defect. MS/MS generates different characteristic acylcarnitine profiles for other disorders depending on the specific location of the metabolic block.
As a practical matter, the Duke team determined that it would be difficult to introduce MS/MS into newborn screening laboratories unless it could at least replace the existing methods for PKU screening. The current methods quantify phenylalanine, one of the amino acids. Some newborn screening labs also screen for other amino acid disorders. One of the problems with older methods that screen for amino acids in newborns’ blood spots is the high rate of false positives, leading to a large number of repeated tests that increase costs as well as anxiety for families. But amino acids share a common structural element and are therefore amenable to MS/MS analysis as a group. As with the acylcarnitines, electrospray ionization of amino acids converted into their butyl esters produces predominantly molecular ions; the molecular ions undergo fragmentation in the reaction cell, losing a common neutral fragment of mass 102. To specifically detect and analyze amino acids, the spectrometer scans both mass analyzers simultaneously looking for a constant difference of 102 mass units. The machine’s computer generates a mass spectrum that consists of the molecular masses of the amino acids with their relative intensities.
As with the acylcarnitine profile, the amino acid profile is essentially devoid of signals from other blood components. Again, adding fixed amounts of isotope-labeled analogues that serve as internal standards allows the accurate quantification of the targeted compounds. Patients with PKU have markedly elevated signals for phenylalanine compared with normal. This method of analysis can detect other amino acidopathies besides PKU. Being able to calculate the ratios between amino acids is a useful bonus of MS/MS that dramatically reduces the number of false positives. In practice, MS/MS analyzes the amino acids and acylcarnitines simultaneously from a single specimen. The value of multiple-analyte testing with MS/MS cannot be overstated.
Second Generation Screening
Metabolic diseases are rare, but their rarity alone does not account for how little people, including physicians, know about them; because of the difficulty of correctly diagnosing these conditions, the incidence of the disorders is frequently underestimated. I became involved in developing methods to detect these ailments nearly 20 years ago. A Duke hospital physician, Charles Roe (currently at Baylor University Medical Center in Dallas), was caring for a baby with propionic acidemia, a disorder of amino acid metabolism found in only 1 in 300,000 babies. Roe reasoned that giving the child carnitine might help this baby eliminate the toxic compound propionyl coenzyme A as propionyl carnitine, which is harmless. Indeed, the baby recovered dramatically. Roe wanted to prove that the chemical transformation he envisioned had indeed occurred by demonstrating high amounts of excreted propionyl carnitine in the urine. He contacted me, an organic chemist specializing in mass spectrometry, and thus began the collaboration that led to the introduction of MS/MS as a diagnostic tool.
Toward the end of the 1980s, the technology reached the point of being able to detect metabolites in plasma and whole blood, not just in urine. By 1990, we demonstrated that the test could use blood spots as a sample source. A Guthrie card typically has five circles, each of which holds about 50 microliters of blood; a single circle can provide enough sample for as many as four different tests. Existing newborn blood-collection protocols provide enough spare blood for repeating tests. We obtained newborn screening cards from the North Carolina state lab, particularly for infants who were diagnosed with MCAD deficiency later in life. We showed that we could have identified these children at birth. After proving the feasibility of extending newborn screening using MS/MS, we spent the next five years developing the procedure to a point where it could be automated and used to screen large numbers of babies. The state of North Carolina provided valuable financial support.
Tandem mass spectrometry clearly offers advantages over newborn screening methods that rely on a range of specialized biochemical tests. Among the advantages are the need for little blood, the ability to test for many conditions simultaneously and a very low false-positive rate. In 1999, North Carolina and Massachusetts became the first states to institute universal newborn screening by MS/MS.
North Carolina currently tests between 400 and 500 newborns a day. Expanded screening by MS/MS began as a cooperative research venture between Duke University Medical Center and the State of North Carolina Public Health Laboratory. While this project was under way, groups from such diverse countries as Australia, Argentina, Saudi Arabia and the United Kingdom began similar projects. The Duke research group has published most of its methodology and claimed no patents. Eventually, commercial labs began taking an interest in the prospect of mass screening. Now, a small number of states have instituted widespread newborn screening with MS/MS.
Trends in State Practices
Nine U.S. states now have at least a year of experience with tandem mass spectrometry in newborn screening and are screening routinely. Several states are conducting pilot studies. Not all the states employing MS/MS screen for all the diseases the technique can detect. Some of them have restricted the test to MCAD deficiency alone, the ailment that fits closest with existing screening paradigms since it can be treated easily and, without treatment, can lead to death. Many of the states are considering expanding the list of tests if they can become confident that diagnosing newborns won’t put an untenable burden on the health care system. The states that are currently using MS/MS in an expanded newborn screening program are doing so with their public health laboratories. In the future, some states may contract the service to private laboratories. After screening, all states have to be prepared to follow up by recommending further biochemical tests to establish a diagnosis. These tests are best referred to established facilities that specialize in this type of testing, such as the Duke University Biochemical Genetics Laboratory.
North Carolina uses a two-tier system for reporting abnormal results. The guidelines require immediate reporting of results when the metabolic pattern clearly exceeds an “alert” threshold level, giving a presumptive diagnostic result. In those cases, the state refers parents to the nearest medical genetics center for counseling and evaluation, which includes follow-up testing of plasma amino acids or of acylcarnitines and of urine organic acids, depending on the abnormal result. Other tests, such as measuring enzyme activity and testing DNA, may be appropriate in some cases. If, on the other hand, the initial screening result is abnormal but does not definitively exceed the diagnostic threshold, the state labs request a second specimen from the child. Positive results from the second test trigger referral for more involved follow-up testing; otherwise, no further action is taken.
Ideally, the newborn screening program should appoint a person to coordinate follow-up and to document the results. Follow-up should include provision of educational material to the local heath care providers and feedback to the state advisory committee regarding true positive, false positive and false negative results. Over the course of two years, North Carolina screened 237,774 babies by MS/MS and confirmed 54 diagnoses (approximately 1 in 4,400): 21 fatty acid oxidation defects, including 17 cases of MCAD deficiency; 18 diseases of amino acid catabolism; and 15 amino acidopathies, mainly PKU. Although some of these babies and the others diagnosed in North Carolina since August 1997 have required emergency treatments and hospitalization, prior knowledge of the diseases and of the correct management has successfully prevented serious medical complications. Many of the patients have not had any illness since their treatment started, soon after birth.
Controversy and debate still exist about the perceived paradigm shift with the introduction of MS/MS in newborn screening. The new test equipment is expensive, and not all the detectable disorders have effective treatments. Previously, the PKU test had suggested a three-part paradigm for newborn screening: Effective treatments should be available for the tested diseases; the diagnostic tests should be inexpensive; and the health costs of failing to treat the diseases are high. After PKU, the three diseases most often added to newborn screening panels-congenital adrenal hyperplasia (CAH), galactosemia and hypothyroidism-fit this paradigm. Without treatment, these conditions lead to profound mental retardation. Each of these diseases required a separate test. Some states, including North Carolina, also routinely screen for hemoglobin disorders, the most common of which is sickle-cell anemia, with yet another unrelated procedure.
With MS/MS, we detect several diseases at once, including ones that lack effective treatments. This introduces not so much an ethical problem, perhaps, as a challenge to how we should think about screening. Some in the medical community agree that all diseases benefit from appropriate early treatment. Also, having a decisive diagnosis is very important for families, bringing closure to parents and often alerting them to diseases that have affected or may affect other family members. Sometimes siblings have already died, and it is important for parents to know why and to be able to consider their future decisions to have children in the light of full information. Furthermore, from a biomedical perspective, it may never be possible to understand a disease or to find cures if we cannot identify those who are affected. Finally, even for disorders with no existing treatments, there is good reason to believe that some of the most severe metabolic diseases will have effective therapies in the near future. Even children who face daunting prognoses now may benefit from newly developed therapies, including enzyme replacement and gene replacement.
Whatever the overall benefits of a new screening procedure, once a state decides to institute a new program on such a wide scale, the state must accept the responsibility of doing it properly and of treating those it diagnoses. For that, they need personnel to care for the children-dietitians that can recommend the proper diets, genetic counselors to advise the families, biochemical geneticists who specialize in treating metabolic disorders. For some states, these requirements can be daunting hurdles, but they must be overcome because the long-term benefits, to public health and in saved social costs, outweigh the burdens.