|Home | About | Journals | Submit | Contact Us | Français|
Combined methylmalonic aciduria and homocystinuria, cobalamin C (cblC) type, is an inherited disorder of vitamin B12 metabolism caused by mutations in MMACHC. CblC typically presents in the neonatal period with neurological deterioration, failure to thrive, cytopenias, and multisystem pathology including renal and hepatic dysfunction. Rarely, affected individuals present in adulthood with gait ataxia and cognitive decline. Treatment with hydroxycobalamin may ameliorate the clinical features of early-onset disease and prevent clinical late-onset disease. Propionic acidemia (PA), methylmalonic acidemia (MMA), and various disorders of cobalamin metabolism are characterized by elevated propionylcarnitine (C3) on Newborn Screening (NBS). Distinctions can be made between these disorders with secondary analyte testing. Elevated methionine is already routinely used as a NBS marker for cystathionine ß-synthase deficiency. We propose that low methionine may be useful as a secondary analyte for specific detection of cbl disorders among a larger pool of infants with elevated C3 on NBS.
Retrospective analysis of dried blood spot (DBS) data in patients with molecularly confirmed cblC disease.
9 out of 10 patients with confirmed cblC born in New York between 2005 and 2008 had methionine below 13.4 μmol/L on NBS. Elevated C3, elevated C3:C2 ratio, and low methionine were incorporated into a simple screening algorithm that can be used to improve the specificity of newborn screening programs and provide a specific and novel method of distinguishing cblC from other disorders of propionate metabolism prior to recall for confirmatory testing.
It is anticipated that this algorithm will aid in early and specific detection of cobalamin C, D, and F diseases, with no additional expense to NBS laboratories screening for organic acidemias and classical homocystinuria.
Combined methylmalonic aciduria and homocystinuria, cobalamin C (cblC) type, is the most common inherited disorder of vitamin B12 metabolism, and has been ascribed to mutations in the MMACHC gene, located at 1p34.1 . The biochemical defect in cblC (as well as that observed in cblD and cblF diseases) results in impaired conversion of dietary vitamin B12 (cobalamin, cbl) to methylcobalamin (MeCbl) and adenosylcobalamin (AdoCbl), essential cofactors necessary for normal functioning of the cytoplasmic enzyme methionine synthase and the mitochondrial enzyme methylmalonyl-CoA mutase, respectively. Methionine synthase catalyzes remethylation of homocysteine to methionine, while methylmalonyl-CoA mutase is responsible for conversion of methylmalonyl-CoA to succinyl-CoA. Untreated cobalamin C, D, and F diseases are therefore characterized biochemically by hyperhomocysteinemia, hypomethioninemia, methylmalonic acidemia, and methylmalonic aciduria. The key steps of intracellular cobalamin metabolism are illustrated in Figure 1.
Although variable, clinical manifestations of cblC are unlike those of either isolated homocystinuria or isolated methylmalonic aciduria . The disorder typically presents in the newborn period with progressive neurological deterioration, manifesting as lethargy, hypotonia, and poor feeding, with seizures and coma appearing later. Affected infants frequently exhibit failure to thrive, anemia and/or pancytopenia, and multisystem pathology including renal and hepatic dysfunction, and cardiomyopathy. The disorder also appears to predispose to microangiopathic disease [5,6]. The most common MMACHC mutation, 271dupA, is found in approximately 40% of mutant alleles, is especially common in persons of European ancestry, and presents almost universally with the early-onset phenotype . More rarely, symptomatic onset of cblC occurs later in childhood or in adulthood, in the form of acute or chronic neurological deterioration (neuropsychiatric disturbance, dementia, and gait ataxia) without systemic manifestations [7-9]. The moderate or severe hyperhomocysteinemia inherent to untreated cblC confers high risk of thromboembolic events in all age groups . Following the confirmation of diagnosis, typically by complementation studies on cultured fibroblasts and/or mutational analysis to localize the defect within the cbl pathway, treatment with pharmacologic doses of cobalamin and other adjunctive therapies can be employed.
Routine screening for disorders of propionate metabolism at birth has only been possible with the introduction of tandem mass spectrometry (MS/MS). An increase in propionylcarnitine (C3) on acylcarnitine profile (ACP) may indicate the presence of propionic acidemia (PA, caused by propionyl-CoA carboxylase deficiency), methylmalonic acidemia (MMA, caused by methylmalonyl-CoA mutase deficiency), various cobalamin defects, or dietary deficiency of vitamin B12. A C3 value of 1.65 μmol/L represents the 50th centile in our region. Extremely high concentrations of C3 (>10 μmol/L) generally indicate acute PA, while moderate increases (5–10 μmol/L) may indicate MMA or a cobalamin defect. It is not possible, however, to reliably differentiate between these disorders on the basis of C3 levels alone , and organic acid analysis and other confirmatory tests are required for accurate diagnosis . There has been some debate as to the C3 level which best represents the most appropriate cut-off for newborn screening: high cut-offs will miss a proportion of cases of cobalamin disorders (false negatives), while low cut-offs will inevitably result in high numbers of false positives and consequent unnecessary parental anxiety. Some MMA disorders and cbl disorders may not produce significant concentrations of C3 and will therefore defy detection when C3 is used in isolation . Many laboratories now use ratios of C3 to other acylcarnitine species (C2, C0, C16) as primary  or secondary [12,15] parameters for screening for disorders of propionate metabolism. When used in conjunction with C3, these metabolite ratios serve to improve the diagnostic capabilities of screening and reduce false-positive rates.
In 2002, a validation study was undertaken by the New York State screening program (Wadsworth Center) in part to establish cut-off values for the majority of analytes (amino acids and acylcarnitines) detectable by MS/MS. The estimate of an appropriate C3 cut-off was calculated as the mean value from a population of normal specimens plus eight standard deviations . The cut-off was set at 7 μmol/L, slightly below the lowest calculated value, in order to accommodate data from the instrument producing the lowest cut-off. An additional cut off was set at 5 μmol/L [13,16] such that ‘borderline’ cases could be subjected to additional algorithmic evaluation (C3:C2 ratio) as detailed above and a repeat specimen requested, if appropriate.
Newborn screening for PA/MMA based on C3 and C3:C2 was begun in New York State in November 2004. The screening program identifies those children that may be considered at risk for these disorders. Results are referred to specialty care centers throughout the state. The diagnosis is made by the specialty care centers based on confirmatory results and evaluation by a metabolic specialist. Until 2008, a protocol was used in New York State which referred patients with suspected disorders of cobalamin or propionate metabolism under three categories, for severe or equivocal elevations of C3 (category 1 and 2, respectively). Category 1 referrals required both a C3 value >7 μmol/L and a C3:C2 ratio >0.2. Category 2 and 3 referrals are made in the context of more modest elevations. Category 2 referrals are defined as cases where initial C3 >7 μmol/L with a C3:C2 value less than 0.2, and category 3 referrals are those where initial C3 values are in the 5-7 μmol/L range and a C3:C2 value is elevated to more than 0.2. In both the category 2 and 3 scenarios, a repeat specimen is requested, acylcarnitine profiling is repeated and cases are referred to the relevant specialty care center if repeat specimen is again ‘positive’ for C3 (C3 greater than 7 μmol/L or C3 in the range 5-7 μmol/L and the C3:C2 ratio greater than 0.2). The algorithm utilized until 2008 is shown in Figure 2a.
In May 2005, methylmalonylcarnitine (C4DC) was added to the Newborn Screening panel as a secondary marker for PA/MMA. In some instances, samples were referred to rule out disorders of propionate metabolism on the basis of persistently elevated C4DC representative. The C4DC value of the 50th centile at the Wadsworth Center has been calculated as 0.15 μmol/L At least two specimens of elevated C4DC (>1 μmol/L) are required for referral to a specialty care center.
Methionine (Met) levels are routinely measured on dried blood spots (DBS) as part of the NBS protocol in New York State. When elevated above the laboratory cut-off (83.7 μmol/L for the New York State NBS laboratory) they are a useful indicator of classical homocystinuria (cystathionine ß-synthase deficiency) when followed up with second-tier assays of plasma homocysteine [12,17,18]. Cobalamin C, D, and F diseases, by contrast, are typically characterized by low or low-normal Met, in keeping with impaired production of MeCbl. Met concentrations are low in many infants during the hours and days immediately after birth when dietary protein intake is minimal. For this reason, low Met, when used in isolation, is not a useful screening test for insufficient production of MeCbl as is seen in cblC, D, and F diseases. Low Met levels are also frequently seen in methylene tetrahydrofolate reductase (MTHFR) deficiency . Consistent with the known defect of methylcobalamin in cblC, several infants with cblC recently managed at our institution were noted to have low Met on NBS.
Since September 2008, the Wadsworth Center has implemented an updated algorithm incorporating methionine as a secondary parameter in samples found to have moderate elevations of C3 (5-7 μmol/L) and C3:C2 ratio >0.2. A subcategory was therefore added to category 3; under the new scheme, samples suspicious for disorders of propionate or cobalamin metabolism are referred under four categories (see Table 1), the fourth of which (category 3b) is designed to be specific for cblC, D, and F diseases. Methionine is only reviewed as a secondary analyte in cases where C3 is in the 5-7 μmol/L range and C3:C2 is elevated (category 3 samples, prior to recall for repeat C3). Prior to 2008, in an algorithm designed primarily to detect cases of PA/MMA, category 3b did not exist and all samples falling into categories 2 or 3 were subject to repeat acylcarnitine profile (ACP) without attention to Met. The cut-off selected for ‘low’ methionine (13.4 μmol/L, <0.2 mg/dL) under the new algorithm was chosen after a review of the initial NBS methionine data of confirmed cases of cblC disease in New York State. The modified algorithm is summarized in Figure 2b.
In this communication, we report retrospectively reviewed NBS data for ten patients with molecularly confirmed cblC born in New York State since 2005, and early biochemical data from eight of these patients whose metabolic management is based at Mount Sinai School of Medicine. The referral pattern for these patients differed significantly based on the inclusion or exclusion of methionine as a secondary marker, supporting the utility of the updated referral algorithm for the prompt identification of newborns afflicted with cblC.
The total number of infants screened in four years between 2005 and 2008 (inclusive) was 1,006,325. We retrospectively reviewed NBS data for 10 patients with molecularly confirmed cblC born in New York during the same time period. Molecular analysis and chart review of early biochemical data from 8 patients whose metabolic management is based at Mount Sinai School of Medicine was performed with informed consent under an IRB-approved research protocol. Genotyping was done by sequencing with PCR primers as described . A subset of patient diagnoses were also confirmed by clinical sequencing and complementation testing.
For children born between January 2005 and December 2008, 10 cases of cblC have been diagnosed on NBS in New York State from the elevated C3 algorithm, 8 of which (6 males, 2 females) have been managed Mount Sinai School of Medicine. In the same time period there have been 5 confirmed diagnoses of propionic acidemia and 12 of methylmalonic acidemia (mutase deficiency) in the state from the elevated C3 algorithm. 4 additional cases of MMA have been confirmed from the elevated C4DC analyte. Cases of cblC are of varying ethnicities and exhibit a variety of pathogenic MMACHC mutations. 3 of the 8 cases managed at Mount Sinai School of Medicine are homozygous for the common 271dupA mutation. A summary of demographic and biochemical follow-up data for patients managed at our institution is provided in Table 2. There have been no cases of cblD or cblF reported to the New York NBS Program since January 2005.
A retrospective review of newborn screening data from the 10 cblC patients born in New York State revealed that cases had a mean C3 of 7.99 μmol/L (range 5.77-10.42 μmol/L) and a mean C3:C2 ratio of 0.46 (range 0.35-0.63, normal <0.2) on NBS at a median 2 days of age (range 2-5 days). The average Met among cases at initial NBS was 9.31 μmol/L (range 5.36-14.74 μmol/L, normal 13.4-83.7 μmol/L). The only case with Met >13.4 μmol/L satisfied criteria for category 1 referral on the basis of extreme C3 elevation (8.44 μmol/L) and elevated C3:C2 (0.45). Six of the ten patients had repeat plasma ACP on reaching the referral center, at a median of 11.5 days of age (range 6-42 days), prior to initiation of treatment measures. Mean C3 at this time was 5.22 μmol/L and mean C3:C2 ratio 0.59. During the interval between NBS and reaching the referral center, C3 fell in 5 out of 6 cases, in 3 cases below the threshold of 5 μmol/L that represents the upper limit of ‘normal’ for C3 on NBS using the algorithm. At a median follow-up age of 11.5 days, 8 patients had plasma methionine measured on a standard plasma amino acid panel. Mean Met on reaching the referral center was 12.06 μmol/L (range 0.34-32.2 μmol/L). At that time, before initiation of treatment measures, 2 of the 8 patients (25%) had methionine above the lower limit of ‘normal’ (>13.4 μmol/L) according to the algorithm. During the interval between initial NBS and referral, 4 out of 8 cases had decreases in Met, 3 had increases, and one value was unchanged.
Table 3 demonstrates the spectrum of NBS findings in the 10 patients with cblC diagnosed in New York State since 2005, and the algorithmic category under which each case was referred.
From 2005 to 2008, inclusive, there were a total of 182 referrals in all categories combined based on C3, the C3:C2 ratio, or C4DC. A diagnosis has been returned or disease has been excluded for 155 of these cases. A total of 31 cases (17% of total) have been confirmed as PA, MMA, or cblC. A total of 124 (68.1%) of these cases have been diagnosed as having no disease. Of the remaining 27 cases (14.8%), two have expired, nine have been lost to follow-up, and sixteen remain open at the time of this writing.
For severe elevations of C3 (>7 μmol/L, categories 1 and 2), there were 577 requests for repeat specimens and 141 referrals to the specialty care centers between January 2005 and December 2008 inclusive (48 months). 7 cblC patients, 5 PA patients (accounting for 100% of PA patients) and 11 MMA patients were referred as category 1 or 2 referrals between 2005 and 2008. Of the 141 referrals from categories 1 or 2, 97 have been confirmed as having no disease, and 21 cases are deceased, lost to follow-up, or remain open. Based on C3 elevations 5-7 μmol/L and C3:C2 >0.2 on NBS (category 3), there were 310 requests for repeat specimens and 9 referrals. One of these referrals has been diagnosed as MMA and one as cblC. 5 of the category 3 referrals have been confirmed as false positive referrals with no disease. In addition, since January 2005, there were 5,317 samples with C3 in the 5-7 μmol/L range and C3:C2 less than 0.2. These cases were excluded from the algorithm as negative screens.
None of the confirmed cases of cblC were referred on the basis of the C4DC analyte. The mean value of C4DC for the 10 cases of cblC was 0.20 μmol/L (population average approximately 0.15 μmol/L), with all CD4C results for the group falling within 3 standard deviations of the mean.
Applying the new algorithm retrospectively, 3 of the 10 cblC cases in New York State would have been referred under category 3b from 2005 to 2008 if the algorithm had existed for the duration of that period. A retrospective review of samples from 2008 (244,531 initial valid specimens) showed that the use of a low cut-off of 13.4 μmol/L for Met in addition to C3 and C3:C2 would result in 3 additional referrals for the year (2 of these 3 additional referrals are in fact confirmed cases of cblC). A low methionine cut-off of 16.8 μmol/L (a 25% increase on the existing cut-off) would result in one additional referral for a total of 4 referrals following category 3b.
Looking at positive predictive values (PPV) of category 3 referrals before and after introduction of Met as a secondary analyte from our experience from 2004 to 2008, referrals under category 3b have to date shown a PPV of 100% (both cases referred have later been confirmed as cblC). Under the undifferentiated category 3, operational until late 2008, PPV was 1/7 (14.3%) for cblC, and 2/7 (28.6%) with the inclusion of a case of MMA.
In 2009 to date (and not included in our tabulated results), two additional patients have been referred under category 3b. One has been diagnosed as having no disease and one has been assigned a diagnosis of cblC (initial NBS values: C3 5.55 μmol/L, C3:C2 0.23, Met 9.38 μmol/L).
All 8 cblC patients managed at Mount Sinai School of Medicine had markedly elevated plasma homocysteine (mean 177.4 μmol/L; range 103.6-276.9 μmol/L; normal <15 μmol/L) at initial follow-up at median 12 days of age, before initiation of treatment measures. At 3 months of age, all patients had started treatment with dietary protein restriction and daily parenteral hydroxycobalamin injections; one patient had already received daily betaine for management of refractory hyperhomocysteinemia. At 3 months, mean homocysteine was 45.7 μmol/L (range 32.8-56.9 μmol/L). No patients suffered from clinical thromboembolic complications before three months of age, although one patient born at term had necrotizing enterocolitis, while another had acute onset on generalized hypotonia and appearances consistent with cerebral microangiopathy of magnetic resonance imaging and venography (MRI, MRV) of the brain. Mean urine MMA on treatment with protein restriction and hydroxycobalamin was 27.0 mmol/mol creat (range 16.6-45.6 mmol/mol creat; normal range 0.8-5.5 mmol/mol creat). We have not measured plasma MMA levels routinely for existing patients, although we are currently exploring the utility of such an approach.
Patients were started on a protein-restricted diet at a median 12.5 days (range 6-59 days), hydroxycobalamin at a median 14.5 days of age (range 6-59 days), and in the 7 cases (88%) eventually requiring betaine for refractory hyperhomocysteinemia, therapy was started at a median 143 days of age (range 32-409 days).
The commonest MMACHC genotype among our 8 patients was homozygosity for the common 271dupA mutation, accounting for 3 patients (38%). All 3 of these patients have variable clinical evidence of developmental delay at most recent follow-up and all 3 had already developed disease manifestations by one month of age despite early treatment. In our small dataset, there were no statistically significant differences between these three patients and the other infants with cblC in terms of levels of NBS marker analytes (C3 p>0.13; Met p>0.33) or early biochemical follow-up data (initial and three month homocysteine, p>0.9 and p>0.7, respectively; urine MMA on treatment p>0.7). Another male patient, whose parents are Guatemalan, was homozygous for the 568insT mutation, which has not been reported previously and appears to be novel.
CblC and its sister diseases (cblD, cblF) are inborn errors of vitamin B12 metabolism characterized biochemically by methylmalonic aciduria and hyperhomocysteinemia. Approaches to NBS for cblC and other disorders of cobalamin and propionate metabolism vary considerably between countries. The advent of MS/MS has, in the United States, allowed state NBS programs to significantly increase the number of mandated disorders screened. From 1995 to 2005, the typical state nearly quadrupled the number of disorders screened. Using commonly cited specificities for MS/MS, it has been estimated that several thousand infants receive false-positive results for inborn errors of metabolism each year . Many European countries do not generally advocate widespread screening for diseases of propionate metabolism on the basis that long-term outcomes may not be significantly influenced by early detection . In the United States, screening is advocated: it has been argued that even in the absence of clear clinical benefit, early diagnosis enables families to receive genetic counseling more promptly and avoids expensive and time-consuming radiologic and biochemical investigations in affected children. Furthermore, it has been suggested that patients with the late-onset cblC phenotype may be a position to avoid clinical manifestations altogether if aggressive treatment is started sufficiently early. In one series, expanded NBS decreased early mortality in children with ‘classic’ organic acidemias, and symptoms at the time of diagnosis were less severe. In addition, short-term neurodevelopmental outcomes were superior among children diagnosed on the basis of NBS compared to children diagnosed at clinical presentation . The same may be true of cblC, although the benefits of early diagnosis and treatment with intramuscular hydroxycobalamin and dietary protein restriction are less well defined.
According to our data, incidence of cblC in New York State between 2005 and 2008 approximated 1:100,000 live births, raising the possibility that the disease is more common than has previously been speculated but still within a range suggested by other authors . Consistent with the findings of other groups , cblC appears to a panethnic disorder. Also consistent with the findings of other groups , homozygosity for the 271dupA mutation, observed in 3 of our patients (38%) was universally associated with an early-onset disease phenotype. This genotype has been shown to consistently cause underexpression of the MMACHC mRNA transcript . We are not able to make any other significant conclusions regarding the biochemical or clinical phenotype of patients with this genotype due to the small patient numbers in our dataset.
All 10 patients diagnosed with cblC after positive NBS since 2005 in New York State had elevations of C3 >5 μmol/L on NBS, and a majority (7 cases) among these had more pronounced elevations (>7 μmol/L). In the 3 most recent cases of cblC in New York State (two in 2008; one in 2009), C3 was only moderately elevated, and immediate referral was possible due to successful implementation of a modified NBS algorithm incorporating methionine as a secondary analyte. For the New York State newborn screening program, the use of low methionine changes categories of an abnormal from borderline to referral and therefore does not affect the total number of screen positive results. Methionine is already routinely assayed as a primary analyte for early detection of cystathionine ß-synthase deficiency. Historically, newborn screening for cbl disorders depended on use of an algorithm designed to detect classical organic acidemias. Low methionine appears to characterize the subpopulation of cblC patients (and theoretically also cblD and cblF patients) within the larger group of those with elevated C3. It is possible that the New York State NBS program will consider implementing an adjustment to the low cut-off of Met (warranting a category 3b referral) to enhance the sensitivity of testing for disorders of cbl and propionate metabolism once a greater body of data is available.
Cut-offs for C3, C3:C2 ratio and methionine vary between NBS programs. We estimate that a NBS program with similar population statistics to New York, but a C3 cut-off of 7 μmol and no secondary markers, would have a total of 718 positive results and miss approximately 3 cases of cblC/D/F and in 1 million babies screened (results of samples analyzed between 2005 and 2008 shown in table 1). Nationally and internationally, cut-offs vary from lab-to-lab and may depend on the exact method of analysis. The subject of lab-to-lab cut-off comparison is beyond the scope of this paper. Based on a review of results in the New York screening program lowering the cut-off for C3 from 7 to 5 μmol/L, without the use of any secondary markers or ratios, would result in an additional 5,317 abnormal specimens. Lowering the cut-off for C3 from 7 to 5 μmol and using a cut-off of 0.2 for the C3:C2 ratio would result in 319 additional abnormal specimens. Lowering the cut-off for C3 from 7 to 5 μmol/L and using a cut-off of 0.2 for the C3:C2 ratio and using a lower limit of 13.4 μmol/L for methionine would result in 8 additional abnormal specimens in one million samples analyzed.
One of our cases (case 3) was subject to late referral since she had only a moderate elevation of C3 and elevated C3:C2 ratio on NBS. At the time she was born, methionine was not used as a secondary analyte in the algorithm (as it was from late 2008 onwards) and the child was subject to repeat C3 testing. This showed a more pronounced elevation of C3 (7.41) μmol/L, prompting referral. A retrospective review of NBS data in this case reveals a Met of 10.7 μmol/L, which, under the new algorithm, would have fulfilled criteria for immediate referral and institution of treatment 2-4 weeks earlier than would have otherwise been likely. This case serves to illustrate the benefits of looking at methionine as a secondary analyte on initial screening in cases of moderate C3 elevation.
It should be noted that NBS data among our patients and corresponding follow-up data at the time of referral and beyond were poorly correlated. Several of our patients were to have declines in their C3 levels and/or rises in their Met levels between NBS and reaching the referral center, making disease more difficult to screen for. This illustrates again the importance of performing the newborn screen expediently after babies are born, since delays may lead to biochemical masking of significant pathology.
Based on early experiences with this modified protocol and this retrospective review of 10 cases, it appears that use of methionine as a secondary screening analyte in cases of moderate C3 elevation significantly improves the specificity and PPV of expanded NBS for cblC, D and F diseases with no additional cost to laboratories. It is likely that expeditious referral, diagnosis, and initiation of treatment provides significant biochemical and clinical benefit to patients, although the nature and extent of such benefit are difficult to demonstrate on the basis of this limited dataset. Many early-onset patients exhibit disease manifestations in the first few days of life, and it is this group who may see the greatest benefits of early and aggressive treatment measures. This area will require further investigation in the form of large multi-center trials. Detailed statistical analysis of larger groups of patients and samples may assist in further adjustment of cut-offs to further optimize sensitivity and specificity of NBS for disorders of cobalamin and propionate metabolism.
Competing and/or Conflicting Interests: None Declared.