Newborn Screening by Tandem Mass Spectrometry

Abstract

After completing this article, readers should be able to:Tandem mass spectrometry (MS/MS) technology has shown tremendous promise in newborn screening pilot programs worldwide with its capacity to measure numerous metabolites from a dried blood spot virtually simultaneously with an extremely rapid throughput per sample (approximately 2 min). The expansion of newborn screening programs by inclusion of an additional 15 to 30 metabolic disorders has the potential to decrease significantly the morbidity and mortality associated with inborn errors of metabolism and could offer the clinician critical diagnostic information when caring for an acutely ill neonate.Newborn screening for metabolic disorders began in 1962 in the United States, with the introduction of the Guthrie bacterial inhibition assay for phenylketonuria (PKU) in a Massachusetts voluntary program. Child health advocates, including the National Association for Retarded Citizens and the March of Dimes Birth Defects Foundation, developed model legislation and lobbied for passage of newborn screening laws at the state level. As a result of these efforts, newborn screening for PKU was legally mandated in most states in the early 1960s. The success of newborn screening for PKU led to development of tests for other conditions, including metabolic disorders (eg, galactosemia, maple syrup urine disease [MSUD], homocystinuria, and biotinidase deficiency), endocrinopathies (eg, congenital hypothyroidism and congenital adrenal hyperplasia), hemoglobinopathies (eg, sickle cell disease and thalassemias), and cystic fibrosis.Different states screen for a variable number of these disorders, with most screening for three to six conditions (FigureF1 ). Such differences reflect state political and economic environments, technologic capabilities of public health departments, regional ethnic composition, and community expectations. All states screen for PKU and congenital hypothyroidism, and 48 states screen for galactosemia. Other inborn errors of metabolism are screened for less frequently: biotinidase deficiency is tested for by 22 states, MSUD by 20 states and the District of Columbia, and homocystinuria by 13 states and the District of Columbia. The reason for inclusion or exclusion of a diagnostic test in a specific state program is not always clear. For example, sickle cell disease has the highest prevalence of any disorder screened for in the United States, but only 41 states and the District of Columbia perform universal screening. An additional three states screen infants from high-risk ethnic groups, and six states do not conduct screening for sickle cell disease. These nine states have a low prevalence of sickle cell disease in their populations (approximately 1:40,000). Nevertheless, these same nine states screen for galactosemia, a disorder that has a prevalence estimated to be approximately 1:70,000.MS/MS is a relatively new technology that offers great promise for detecting neonates who have a number of inborn errors of metabolism (usually 15 to 30 metabolic diseases are screened for during a typical run). However, only three states currently use this technology for comprehensive newborn screening, although other states have pilot programs in place or in the planning stages. Therefore, a neonate may be screened for as few as three disorders or as many as about 30, depending on where he or she was born. This state-to-state variability in screening profiles inevitably results in inequities in delivery of health care.A mass spectrometer is an analytic instrument that separates and determines the quantity of ions in a sample on the basis of their mass-to-charge (m/z) ratios. An MS/MS typically consists of two quadrupole mass spectrometers separated by a collision cell containing an inert gas. A detection apparatus placed after the second mass spectrometer quantifies specific ions that the instrument has been programmed to identify.For newborn screening, a small disk from the neonatal heelstick filter paper is eluted, derivatized with butanolic hydrochloride, and redissolved in an appropriate solvent for injection into the MS/MS. Isotopically labeled internal standards are added to the sample during preparation. Samples are subjected to a soft ionization procedure (eg, electrospray) and pass through the first quadrupole, which separates them on the basis of their m/z ratios. The ions then enter the collision cell, where they undergo fragmentation. These smaller fragments are analyzed in the second quadrupole, and data are interpreted with the aid of a computer.MS/MS can quantify the presence of butyl esters of amino acids and acylcarnitines in the same sample virtually simultaneously using different scan functions. A typical run assays 15 to 30 analytes in approximately 2 minutes, making MS/MS ideal for handling the high volume required for newborn screening. Many disorders that can be detected by MS/MS, including organic acidemias, fatty acid oxidation defects, and urea cycle defects, typically have not been included in past screening programs. Examples of analytes detected and the corresponding disorders diagnosed are listed in Table 1 MS/MS screening is more sensitive and specific than current methods used to detect aminoacidemias. In a comparison of MS/MS with fluorometric analysis for PKU detection, MS/MS not only confirmed all previous cases detected by the older methodology, but it also reduced the number of false-positive results from 91 to 3.Recent studies suggest that carbohydrates and steroids also can be assayed by MS/MS, making screening for galactosemia and congenital adrenal hyperplasia theoretically possible. Additionally, researchers have reported some success in screening for sickle cell disease and other variant hemoglobins using MS/MS or electrospray mass spectrometry. However, different methodologies are necessary at present to detect galactosemia, congenital hypothyroidism, congenital adrenal hyperplasia, biotinidase deficiency, and hemoglobinopathies.MCAD deficiency is an inborn error of fatty acid oxidation that has an incidence of approximately 1:10,000. It is a significant cause of childhood morbidity and mortality and has been an important focus of MS/MS newborn screening pilot programs. Affected children appear normal, unless they are exposed to environmental stress capable of causing catabolism and subsequent metabolic decompensation (eg, an infection or period of prolonged fasting). Patients often present with hypoketotic hypoglycemia or a Reye syndrome-like illness in late infancy or early childhood. Many require admission to an intensive care unit, and mortality at the time of first presentation is estimated to be 25% to 40%. Neurodevelopmental problems are present in approximately 40% of survivors.Recent attention has been focused on the possibility of screening for MCAD deficiency and other fatty acid oxidation defects by detection of carnitine esters via MS/MS using Guthrie card dried blood spots. An MS/MS pilot program in North Carolina detected 24 asymptomatic infants who had MCAD deficiency among 327,031 neonates screened, resulting in a population incidence of 1:10,095. Confirmation of the diagnosis was made by repeating the acylcarnitine profile on venous blood, with or without further DNA mutation analysis. All children were treated by carnitine supplementation, but fat intake was not restricted. Parents were given careful instructions on the importance of avoiding fasting and trained to use glucometers. After 36.5 patient years of follow-up, two patients were admitted to the hospital a total of four times for intravenous dextrose and rehydration. In striking contrast to patients detected on the basis of clinical features, there were no deaths, significant hypoglycemia, or seizures reported, and all patients appeared developmentally normal. Successful detection of MCAD deficiency by MS/MS newborn screening has been reported in other states, including Louisiana, Ohio, Pennsylvania, and Wisconsin.In addition to the 24 MCAD deficiency patients identified by the North Carolina newborn screening program, 36 other patients were confirmed to have metabolic disorders, resulting in a population incidence of 1:5,500 for the specific conditions screened for by this state. Hyperphenylalaninemia was detected in 15 neonates; organic acidemias in 12; citrullinemia in 3; galactosemia in 2; and argininosuccinic aciduria, hypermethioninemia, short-chain acylCoA dehydrogenase (SCAD) deficiency, or a long-chain fatty acid oxidation defect in 1 each. MS/MS in the Wisconsin newborn screening program, although in an early stage, has identified two patients who had MCAD deficiency, one who had SCAD deficiency, and one who had propionic acidemia.Although Massachusetts, North Carolina, and Wisconsin are the only states that have comprehensive MS/MS newborn screening programs, some hospitals offer expanded newborn screening on a fee-for-service basis by contracting with private or university-based laboratories. Naylor and Chace from NeoGen Screening, Inc (Pittsburgh, PA) have reported their experience in screening more than 700,000 newborns in Pennsylvania, Ohio, North Carolina, and Louisiana. Because of the capacity to screen for a large number of disorders simultaneously, 163 neonates who had inborn errors of metabolism were detected, resulting in a total population frequency of 1:4,000. Frequently detected metabolic diseases included MCAD deficiency (39 patients), PKU (37), nonPKU hyperphenylalaninemia (31), MSUD (9), and glutaric acidemia type I (9). Additionally, propionic acidemia (5), citrullinemia (4), and hypermethioninemia (4) as well as 11 other organic acidemias and 6 fatty acid oxidation defects were found. Similar experiences have been reported in other countries, including Australia and Saudi Arabia.A tandem mass spectrometer is an expensive instrument, and MS/MS screening programs must have access to back-up machines in case of equipment failure. Highly trained laboratory technical staff are necessary for routine operation and maintenance. Nevertheless, the cost of adding MS/MS to existing newborn screening programs has been estimated at $10 to $20 per sample. As emphasized by Charrow et al, the cost would be the same regardless of the number of metabolic disorders screened for by a given program. The previously cited estimate does not take into account other factors, including additional tests to confirm the diagnosis, patient notification, and treatment. Ciske et al report the cost of screening, confirmatory testing, and treating patients who have inborn errors of metabolism in Wisconsin using MS/MS to be approximately $400,000 annually. On the other hand, if MS/MS screening were not used and patients were detected strictly on the basis of clinical features, annual cost is estimated to be approximately $900,000.If full advantage is taken of the capabilities of MS/MS, the number of disorders screened for could increase dramatically, leading to an increase (estimated by Charrow et al to be 50% to 100%) in the number of patients diagnosed with inborn errors of metabolism. More physicians, nutritionists, and genetic counselors trained in biochemical genetics would be needed to provide care to patients and their families. Access to medical foods is also an important topic for consideration. Many of these disorders are treated by special dietary formulas and medical foods, but insurance companies have been slow or unwilling to provide coverage for such therapies.A given metabolite may be abnormally elevated due to different underlying inborn errors of metabolism. An elevated methylmalonic acid (MMA) level, for example, may be caused by various biochemical defects (methylmalonyl-CoA mutase deficiency, defects of cobalamin metabolism), as well as vitamin B12 deficiency. In a Quebec newborn screening program for methylmalonic aciduria, many patients who had initially elevated MMA levels were found to have reduced levels over time, and patients who had low-to-moderate MMA excretion often remained asymptomatic. Therefore, elevated neonatal MMA was found to be a poor predictor of eventual outcome. Because different disorders, as well as a vitamin deficiency, cause elevations of MMA, the significance of a borderline or moderately high level detected on a newborn screen may be difficult to interpret in some instances. Although MS/MS screening for methylmalonic acidemia is based on the detection of propionylcarnitine (not MMA), experience regarding the sensitivity and specificity of such analysis is limited. Similar uncertainty has been reported with 3-hydroxyisovalerylcarnitine levels used for the detection of 3-methylcrotonyl-CoA carboxylase deficiency.Analytic cut-off values must be established for each condition to produce an acceptable balance between false-positive and false-negative results. If the cut-off level is set too high, there is a risk of detecting only severe or moderately severe disease. Conversely, if the cut-off is set too low, patients who have abnormally elevated metabolites but not true disease may be detected. The setting of cut-off values in newborn screening programs is not a trivial undertaking; it clearly is linked to the problem of disease heterogeneity. It will be essential to follow patients over significant periods of time to collect data regarding false-positive and false-negative cases, which will form the basis for refining cut-off levels.Given the very low false-positive rate and extensive experience with MS/MS analysis in older patients who have inborn errors, the likelihood of false-positive results in newborn screening would appear to be small. On the other hand, the incidence of false-negative results will become apparent only after widespread MS/MS screening programs have been in place for a period of time. A preliminary report from the North Carolina pilot program suggests a potential low false-negative rate. In screening of more than 300,000 neonates, only three diagnoses have been missed and were diagnosed in infancy based on clinical features: two cases of glutaric acidemia type I and one of a late-onset form of methylmalonic acidemia. In New South Wales, Australia, two children were missed in screening of more than 130,000 neonates: one instance of cobalamin C disease and one of glutaric acidemia type I (glutarylcarnitine was not being screened for by this program).MS/MS screening is only the first step toward improved diagnosis of inborn errors of metabolism; confirmatory testing is required for all of these disorders. However, there are no universally accepted procedures for diagnosing most of these conditions. Different laboratories and physicians use various combinations of analyte analysis, enzymology, and DNA methodologies to confirm a specific diagnosis. More uniform protocols and identification of specific follow-up testing laboratories could help streamline the diagnostic process.There is clear evidence that newborn screening by MS/MS is beneficial in MCAD deficiency, but the benefit of early detection and treatment has not yet been proven for many other conditions. Nevertheless, the goal of MS/MS screening is to identify affected children before they sustain irreversible damage. Current therapies may not be “curative,” but it is reasonable to expect that presymptomatic detection at least will give a child who has a metabolic disorder the best chance for a good outcome. Furthermore, early diagnosis can spare the family the anguish of a lengthy diagnostic process and allow accurate genetic counseling before another affected sibling is born. Research also has begun into the possibility of screening for lysosomal storage disorders in newborns using MS/MS, but this is more controversial because of the lack of adequate therapy for most of these diseases. On the other hand, early diagnosis of an infant who has a lysosomal storage disorder could allow for accurate genetic counseling before the birth of a second affected child.The optimal postnatal age for sample collection for MS/MS screening has not yet been determined. Clayton et al concluded that screening for MCAD deficiency at 7 to 10 days would result in negligible false-positive and -negative results. However, many inborn errors, including urea cycle defects, MSUD, and some organic acidemias and fatty acid oxidation defects, may present in the first few days of life. Relatively early collection and rapid analysis of the newborn sample would be necessary to diagnose these conditions before the onset of symptoms. Samples most likely will continue to be collected between 24 and 48 hours after birth, following current newborn screening practice. Gestational age (at least in the 37 to 42 wk range) does not appear to affect levels of amino acids or acylcarnitines. However, more severe prematurity (28 to 32 wk) may be associated with decreased levels of long-chain acylcarnitines (C14:1 to C16). Preterm infants receiving total parenteral nutrition (TPN) have alterations in their biochemical profile that reflect TPN composition, including an increased phenylalanine-to-tyrosine ratio and the presence of linoleyl carnitine (C18:2). Furthermore, most amino acids appear to decrease in concentration between postnatal days 1 and 5. The significance of these observations to the interpretation of MS/MS screening data should become more clear with experience.In general, parents support newborn screening programs, especially if given adequate information about the process. The goal, of course, is to benefit the child without producing undue anxiety in the family. However, it may be difficult to explain such an elaborate multiple disorder screening program. A number of approaches to public education, using different types of media with different levels of detail, may be required to transmit the desired message.The list of disorders that may be detected by MS/MS continues to grow rapidly. In addition to carbohydrates (galactosemia), steroids (congenital adrenal hyperplasia), lysosomal substrates (lysosomal storage diseases), and hemoglobins (sickle cell disease and thalassemias), MS/MS also may detect metabolites associated with purine and pyrimidine disorders, congenital disorders of glycosylation, and bile acid synthetic defects (Table 2 ). With time, more disorders undoubtedly will be added to the already impressive potential MS/MS screening menu.MS/MS newborn screening technology has shown tremendous promise in its ability to detect a large number of metabolic diseases rapidly at a reasonable cost. MS/MS programs could provide physicians with critical information necessary for delivering optimal care to neonates who have metabolic disorders, who often present with nonspecific signs and symptoms suggestive of sepsis or hypoxic-ischemic encephalopathy. However, it is imperative for children diagnosed by such programs to be followed closely over a number of years to provide data on the long-term efficacy of early detection and treatment of metabolic disorders. Collaboration between screening laboratories is essential to establish accurate universal cut-offs for the numerous analytes used in the screening process. Given the extreme differences in state newborn screening programs, inequities in health care delivery may be minimized by having standard techniques and menus, regionalizing testing when necessary, and performing regular interlaboratory proficiency testing.The author would like to thank Dr. Stephen I. Goodman for the helpful discussions and critical reading of the manuscript.