- Email: [email protected]
MS analysis of C5-acylcarnitines in dried blood spots
Clinica Chimica Acta 421 (2013) 41–45
Contents lists available at SciVerse ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
UPLC–MS/MS analysis of C5-acylcarnitines in dried blood spots Nils Janzen a, b, c,⁎, Ulrike Steuerwald a, d, Stefanie Sander a, Michael Terhardt a, Michael Peter a, Johannes Sander a a
Screening-Labor Hannover, Hannover, Germany Clinic for Pediatric Kidney, Liver and Metabolic Diseases, Hannover Medical School, Hannover, Germany Department of Neuropediatrics, Children’s Hospital, Ruhr University of Bochum, Bochum, Germany d Department of Occupational and Public Health, National Hospital System, Tórshavn, Faroe Islands b c
a r t i c l e
i n f o
Article history: Received 9 February 2013 Received in revised form 28 February 2013 Accepted 1 March 2013 Available online 13 March 2013 Keywords: Acylcarnitine screening C5-acylcarnitine isoforms Pivalic acid Second tier method
a b s t r a c t Background: Metabolic screening including newborn screening requires further differentiation of C5acylcarnitines in order to separate different metabolic disorders and to detect interferents like pivalic acid originating from antibiotics. Methods: For individual quantification of C5-acylcarnitine isoforms in dried blood spots we combined UPLC using a C18 column and gradient elution with tandem mass spectrometry in ESI+ mode. Results: Results were linear, coefficients of determination (R2) > 0.9977, intra- and inter-assay coefficients of variations b 5.2%, recovery 96.8–105.2%, limits of detection and quantitation b 0.2 μmol/L. Out of 29.309 blood samples of the isolated population of the Faroe Islands 56 exceeded the cut-off of 0.5 μmol/L for C5-acylcarnitine; 45 of which could be retested using the method described. Pivaloylcarnitine was identified in 43 samples, isovalerylcarnitine was found in two samples. Conclusions: The method was developed to allow direct re-analysis of samples showing elevated concentrations of C5-acylcarnitines in a metabolic screening program based on quantification of acylcarnitines after butylation. The technique should be especially useful in newborn screening for exclusion of false positives and for differentiation between isovaleric acidemia and 2-methylbutyryl-CoA dehydrogenase deficiency. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In the isolated population of the Faroe Islands primary carnitine deficiency (PCD, OMIM #212140), a rare metabolic defect in most other industrial countries, is rather frequent and has led to several deaths related to use of pivalic acid containing antibiotics [1]. Formation of pivaloylcarnitine has been found to cause severe carnitine deficiency not only in homozygous PCD but also in heterozygotes. In cooperation with the Faroese National Health System our laboratory tested 29.309 dried blood samples of individuals living on the Faroe Islands. The aim was to identify persons with yet undetected PCD as well as cases of secondary carnitine depletion in the Faroese population. The possibility to use dried blood spots was considered advantageous under the local conditions given for blood collection and for postal transport. In this material acylcarnitines and free carnitine were quantified by tandem mass spectrometry after butylation. Since acylcarnitines may undergo solvolysis during derivatization, the concentration of free carnitine may be overestimated [2]. However, for detection of carnitine transporter deficiency it is most important to determine levels of free
⁎ Corresponding author at: Screening-Labor Hannover, Postbox 91 10 09, 30430 Hannover, Germany. Tel.: + 49 5108 921630. E-mail address: [email protected] (N. Janzen). 0009-8981/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cca.2013.03.001
carnitine accurately. Therefore, free carnitine was also measured without the butylation step. The analytical methods applied were identical to those used for newborn metabolic screening [3]. The full spectrum of acylcarnitines was quantified in order to be able to distinguish primary carnitine deficiency (PCD) from secondary carnitine deficiency due to inborn errors of metabolism e.g. 3-methylcrotonyl-CoA carboxylase deficiency (OMIM #210200), glutaric aciduria type 1 (OMIM #231670) or due to medication. Being aware of the frequent use of pivalic acid containing antibiotics in the Faroese population we expected several samples with a high signal for C5-acylcarnitine caused by pivaloylcarnitine. In tandem mass spectrometry (MS/MS) analysis pivaloylcarnitine cannot be differentiated from other isobaric C5-acylcarnitines e.g. isovalerylcarnitine quantified for detection of isovaleric acidemia (IVA, OMIM #243500), an inborn error of metabolism which may lead to severe acidotic episodes and neurocognitive defects [4]. Another isobaric acylcarnitine is 2-methylbutyrylcarnitine which accumulates in 2methylbutyryl-CoA dehydrogenase deficiency (OMIM #610006) [5]. In addition, valerylcarnitine, a further isoform, might be elevated as a result of odd chain fatty acid degradation [6]. This substance is also a pharmaceutical intermediate and a technical product. In order to identify pivaloylcarnitine correctly and to avoid false positive results suggesting IVA or 2-methylbutyryl-CoA dehydrogenase deficiency we aimed at finding a rapid second tier test immediately
42
N. Janzen et al. / Clinica Chimica Acta 421 (2013) 41–45
applicable on the original blood spot. For this application we developed a method separating C5-acylcarnitines by ultra-high performance liquid chromatography (UPLC) prior to quantification by tandem mass spectrometry (MS/MS).
with four aliquots of water. Subsequently 90 μL was transferred into a 384 well microtiter plate.
2. Materials and methods
A XEVO UPLC–MS/MS system (Waters, Eschborn, Germany) consisting of a gradient pump, integrated solvent degasser, temperature controlled autosampler, column oven and triple quadrupole mass spectrometer was used. In the autosampler the microtiter plates were cooled down to 10 °C. 5 μL of the sample solution was injected onto an Acquity UPLC BEH C18 1.7 μm, 2.1 × 50 mm column (Waters). The chromatography was performed at a flow rate of 0.45 mL/min using a gradient between solvent A (water with 0.1% formic acid and HFBA each) and solvent B (methanol with the same modifiers as solvent A). The gradient started with 5% B and was held for 0.5 min, raised to 25% B within 0.5 min and held to 2.35 min, raised to 59% in 4.75 min and to 100% in 5.25 min. The gradient was held to 7.75 min and went back to the initial conditions within 0.25 min. The MS/MS system was equipped with an ESI interface. The nebulizing gas was nitrogen, argon was used for collision at a pressure of 4.0 × 10−3 mbar. The source temperature was set at 150 °C. Capillary voltage was 0.5 kV. The C5-acylcarnitines were detected in ESI positive mode (ESI+) by multiple reaction monitoring (MRM) mode. Collision energy was 20 eV, cone voltage was 28 V. The analyzed transitions were m/z 246.1 → 85 for pivaloyl-, 2-methylbutyryl-, isovaleryl- and valerylcarnitine and 255.2 → 85 for d9 isovalerylcarnitine. All results were calculated using the software TargetLynx 4.1 (Waters) and MS Excel 2010.
2.1. Materials UPLC/MS grade methanol, acetonitrile, formic acid and water were from Biosolve BV (Valkenswaard, The Netherlands). The lyophilized deuterated acylcarnitine standard mixture and extraction buffer contained in the neonatal acylcarnitine kit MassChrom® were purchased from Chromsystems (Munich, Germany). Heptafluorobutyric acid (HFBA) was from Fluka (Deisenhofen, Germany). The acylcarnitines pivaloylcarnitine (2,2-dimethylpropionic carnitine), valerylcarnitine, 2-methylbutyrylcarnitine and isovalerylcarnitine were purchased from ten Brink (VU University Medical Center Amsterdam, The Netherlands). 2.2. Standards and solutions Lyophilized internal standards from the kit containing deuterated d9-isovalerylcarnitine were diluted with 50 mL of the extraction buffer. One milligram of pivaloyl-, valeryl-, 2-methylbutyryl- and isovalerylcarnitine each were prepared in 1 mL water. The stock was diluted 2000 fold to reach the concentration of 0.5 μg/mL. 10 μL was added to 190 μL internal standard solution and further diluted with water to get a final concentration of 5 ng/mL acylcarnitine each. Calibrators for C5-acylcarnitines were made with final concentrations of 0 (blank), 1, 5, 15 and 25 μmol/L. Controls were prepared by adding pivaloyl-, valeryl-, 2methylbutyryl- and isovalerylcarnitine (0, 2 and 20 μmol/L each) to whole blood (containing EDTA as anticoagulant) which was then spotted onto the filter card. 2.3. Samples Blood from children and adults of all ages of the Faroese population was collected and spotted onto filter paper cards (Munktell & Filtrak, Bärenstein, Germany). The cards were dried at ambient temperature overnight before they were shipped to the laboratory. Here they were stored at 4–6 °C in a refrigerator until use. After completing the analyses all cards were sent back to the laboratory on the Faroe Islands. 2.4. Sample preparation Acylcarnitines were extracted from dried blood spots of 4.7 mm diameter in 96 well microtiter plates. 200 μL of extraction buffer containing internal standards was added and the plate was shaken for 30 min at ambient temperature. The plate was centrifuged (2800 U/min) for 10 min to separate the filter card material from the supernatant before it was removed with a pipette and diluted
2.5. Chromatography and mass spectrometry
3. Results 3.1. Linearity, precision, recovery, LOD, LOQ For the linear regression data were plotted against the concentration of the spiked samples. Peak heights were determined with TargetLynx® software, concentrations were calculated by use of MS Excel®. The parameter slope and intercept of the linear calibration functions are listed in Table 1. Intra-day and inter-day precision data were assessed over a period of 6 days by measuring control samples at two levels. At concentrations of 2 and 20 μmol/L, respectively, the imprecision data showed that the intra- and inter-day CVs for all the C5-acylcarnitines were not greater than 5.2%. Recovery from dried blood spot samples spiked with unlabeled acylcarnitines was determined six fold. As shown in Table 1 the range of recovery was between 96.8% and 105.2%. In order to determine the limit of detection (LOD) and limit of quantification (LOQ), the entire range of the linear calibration curve of the respective C5-acylcarnitines was used [7]. The standard error sy/x of the measured concentration (y-estimate in the regression equation) was used in place of the standard deviation of the blank sample [8]. The slope was calculated from the corresponding calibration curves. The LOD was then calculated as LOD = 3sy/x/slope, and
Table 1 Validation data of C5-acylcarnitines from dried blood spots. Slope
Intercept
Acylcarnitine
[a]
[b]
Pivaloylcarnitine
1.7401
0.3565
2-Methylbutyrylcarnitine
0.9209
Isovalerylcarnitine Valerylcarnitine
R2
LOD
LOQ
Level
Recovery
Intra-assay
Inter-assay
[μmol/L]
[μmol/L]
[μmol/L]
[%]
CV [%]
CV [%]
0.9978
0.03
0.11
0.1733
0.9982
0.02
0.08
1.8242
0.4085
0.9980
0.05
0.16
1.9983
0.4203
0.9977
0.03
0.09
2 20 2 20 2 20 2 20
98.4 103.2 96.8 105.1 98.3 105.2 100.1 105.1
3.9 1.0 3.9 1.8 3.8 1.0 3.5 1.4
5.0 1.9 5.2 2.3 5.2 1.4 4.5 2.4
N. Janzen et al. / Clinica Chimica Acta 421 (2013) 41–45
43
isovalerylcarnitine peak and a sample with a high pivaloylcarnitine peak. Out of 29.309 samples 56 blood spots tested positive for C5-acylcarnitines (cut-off 0.5 μmol/L). Positives were from adult persons as well as from children. Concentrations ranged from 0.51 to 9.5 μmol/L. The C5-acylcarnitine/acetylcarnitine ratio (cut-off b 0.01) was also elevated showing values from 0.017 up to 0.464. 11 samples had already been sent back and were not available for retesting, of the remaining 45 samples chromatographic separation of C5 isobares revealed pivaloylcarnitine only in 43 cases. There were two cases presenting isovalerylcarnitine (2.6 and 2.7 μmol/L respectively) in persons with a mild asymptomatic form of IVA. Unfortunately, since both persons (50 and 68 years old, respectively) did not suffer from specific symptoms they refused to have further biochemical and molecular genetic examinations. Ingestion of pivampicillin (Pondocillin®) or pivmecillinam (Selexid®) was confirmed in 35 individuals testing positive for pivaloylcarnitine, eight persons were not sure about their medication at the time of blood collection.
LOQ = 10sy/x/slope. The LOD and LOQ for the C5-acylcarnitines were between 0.02 and 0.16 μmol/L. No matrix effects were observed within the retention time range relevant for measurement of C5 acylcarnitines. Quantitation of all C5-acylcarnitines was linear with determination coefficients (R 2) greater than 0.9977. For the quantitation of 2-methylbutyrylcarnitine the sum of both isoforms was calculated. In Fig. 1 the chromatographic separation of the relevant acylcarnitines is demonstrated. The C5-acylcarnitines were well separated. The isoforms of (S,R) 2-methylbutyrylcarnitine gave a double peak without a full baseline separation.
3.2. Results of population screening Fig. 2 shows a chromatogram of a healthy person compared with a sample of a patient with isovaleric acidemia with a prominent
3.52
100
MRM of 3 Channels ES+ 255.2 > 85.0 1.96e5
%
internal standard d9 isovalerylcarnitine
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.62
100
3.90
MRM of 3 Channels ES+ 246.1 > 85.0 7.55e5
%
valerylcarnitine
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.53
100
3.90
MRM of 3 Channels ES+ 246.1 > 85.0 1.65e6
%
isovalerylcarnitine
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.41
100
2-methylbuturylcarnitine
3.90
MRM of 3 Channels ES+ 246.1 > 85.0 2.09e5
%
3.37
3.52
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.29
100
3.80
3.90
MRM of 3 Channels ES+ 246.1 > 85.0 1.83e5
%
pivaloylcarnitine
3.53 3.43
Time
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.90
Fig. 1. Chromatographic separation of C5-acylcarnitines and internal deuterated standard. The peaks of the isoforms of (S,R) 2-methylbutyrylcarnitine are not baseline separated.
44
N. Janzen et al. / Clinica Chimica Acta 421 (2013) 41–45
100
MRM of 3 Channels ES+ 255.2 > 85.0 1.92e5
3.51
A
%
internal standard d9 isovalerylcarnitine
3.00 100
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
B
3.90 MRM of 3 Channels ES+ 246.1 > 85.0 2.14e4
%
sample of a healthy subject
3.52
3.00 100
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.53
C
3.90 MRM of 3 Channels ES+ 246.1 > 85.0 1.65e6
%
isovalerylcarnitine
3.00 100
3.10
3.20
D
3.30
3.40
3.50
3.60
3.70
3.80
3.29
3.90 MRM of 3 Channels ES+ 246.1 > 85.0 1.83e5
%
pivaloylcarnitine
3.53 3.43
Time
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.90
Fig. 2. Chromatogram of the internal standard d9-isovalerylcarnitine (A); sample of a healthy subject (B); sample of a patient with isovaleric acidemia (C) and from a patient treated with pivalic acid containing antibiotics (D).
4. Discussion Several methods for differentiation of C5 isomers have been published but these are not suitable for immediate reanalysis of dried blood spot samples. Maeda et al. [9] described a rather time consuming method including solid phase extraction for chromatographic separation of acylcarnitines using serum and urine samples meaning that a second sample is required. This is also true for a rapid method by Ferrer et al. [10] separating C4- and C5-acylcarnitine isomers in plasma samples. Forni et al. [6] who analyzed plasma and dried blood spots derivatized the acylcarnitines prior to chromatographic separation. During this chemical process, however, acylcarnitines are partially being fragmented into free acids and carnitine [2]. This particularly applies for short chain acylcarnitines including C5 derivatives. An analytical procedure applicable as a second tier test in dried blood spots was
developed by Shigematsu et al. to quantify isovalerylglycine in cases of elevated C5-carnitine in newborn screening samples [11]. This method is useful for confirming IVA but it does not identify other C5 isoforms that might have been the cause for elevated C5 levels in the original screening test. Formation and renal excretion of pivaloylcarnitine following ingestion of pivalate-bound drugs have been known for a long time [12]. The renal loss of carnitine is considerable, because more than 98% of pivalate is excreted as pivaloylcarnitine [13,14]. Nakajima et al. [15] showed the same effect of antibiotics esterified with pivalic acid and concluded that long-term administration of pivalate-containing antibiotics should be avoided particularly in children. While the risk of metabolic decompensation following short term pivalic acid application may be rather low in persons with full function of carnitine uptake [16] there is an important risk in a population with more widespread carnitine
N. Janzen et al. / Clinica Chimica Acta 421 (2013) 41–45
transporter deficiency like on the Faroese Islands as well as in patients with other metabolic disorders and high carnitine excretion. Several cases of hypoketotic hypoglycemia caused by impairment of fatty acid degradation as a consequence of hypocarnitinemia have been described [17–19]. In such cases our new method is very well suited to uncover pivalic acid as the cause of hypocarnitinemia. Our method was originally established for a specific application but it should be especially useful for newborn metabolic screening since the majority of screening results for IVA using standard MS/MS are false positive. In a screening program done on 146.000 newborns in Japan Shigematsu et al. [11] found only one true positive case but 1064 false positives using a cut-off of 0.5 μmol/L for C5. Betalactam antibiotics esterified with pivalic acid for better enteral absorption are used in several countries. This includes industrial countries with an extended metabolic newborn screening program like Japan. Even in countries where pivalic acid containing antibiotics has been taken off the market a high percentage of false positive results for isovaleric academia is observed in newborn screening. In a series of 1.6 million newborns in Germany Ensenauer et al. [20] detected 24 patients suffering from IVA. For identification of possible positive cases the authors applied several acylcarnitine ratios in addition to defined cut-off values for C5 resulting in a recall rate of 0.024%; this seems low but it corresponds to a true positive to false positive ratio of 1:16. A similar ratio has been observed in our laboratory. With the introduction of our new method into our own screening program we will be able to avoid false positives and also to differentiate between isovaleric acidemia and 2-methylbutyryl-CoA dehydrogenase deficiency. References [1] Rasmussen J, Nielsen OW, Lund AM, Kober L, Djurhuus H. Primary carnitine deficiency and pivalic acid exposure causing encephalopathy and fatal cardiac events. J Inherit Metab Dis 2013;36:35–41. [2] Johnson DW. An acid hydrolysis method for quantification of plasma free and total carnitine by flow injection tandem mass spectrometry. Clin Biochem 2010;43: 1362–7. [3] Chace DH, Kalas TA, Naylor EW. Use of tandem mass spectrometry for multianalyte screening of dried blood specimens from newborns. Clin Chem 2003;49:1797–817. [4] Grunert SC, Wendel U, Lindner M, et al. Clinical and neurocognitive outcome in symptomatic isovaleric acidemia. Orphanet J Rare Dis 2012;7:9.
45
[5] Sass JO, Ensenauer R, Roschinger W, et al. 2-Methylbutyryl-coenzyme A dehydrogenase deficiency: functional and molecular studies on a defect in isoleucine catabolism. Mol Genet Metab 2008;93:30–5. [6] Forni S, Fu X, Palmer SE, Sweetman L. Rapid determination of C4-acylcarnitine and C5-acylcarnitine isomers in plasma and dried blood spots by UPLC–MS/MS as a second tier test following flow-injection MS/MS acylcarnitine profile analysis. Mol Genet Metab 2010;101:25–32. [7] Mani DR, Abbatiello SE, Carr SA. Statistical characterization of multiple-reaction monitoring mass spectrometry (MRM-MS) assays for quantitative proteomics. BMC Bioinformatics 2012;13(Suppl. 16):S9. [8] Anderson DJ. Determination of the lower limit of detection. Clin Chem 1989;35: 2152–3. [9] Maeda Y, Ito T, Ohmi H, et al. Determination of 3-hydroxyisovalerylcarnitine and other acylcarnitine levels using liquid chromatography–tandem mass spectrometry in serum and urine of a patient with multiple carboxylase deficiency. J Chromatogr B Analyt Technol Biomed Life Sci 2008;870:154–9. [10] Ferrer I, Ruiz-Sala P, Vicente Y, Merinero B, Perez-Cerda C, Ugarte M. Separation and identification of plasma short-chain acylcarnitine isomers by HPLC/MS/MS for the differential diagnosis of fatty acid oxidation defects and organic acidemias. J Chromatogr B Analyt Technol Biomed Life Sci 2007;860:121–6. [11] Shigematsu Y, Hata I, Tajima G. Useful second-tier tests in expanded newborn screening of isovaleric acidemia and methylmalonic aciduria. J Inherit Metab Dis 2010;33:S283–8. [12] Holme E, Greter J, Jacobson CE, et al. Carnitine deficiency induced by pivampicillin and pivmecillinam therapy. Lancet 1989;2:469–73. [13] Konishi M, Hashimoto H. Determination of pivaloylcarnitine in human plasma and urine by high-performance liquid chromatography with fluorescence detection. J Pharm Sci 1992;81:1038–41. [14] Brass EP. Pivalate-generating prodrugs and carnitine homeostasis in man. Pharmacol Rev 2002;54:589–98. [15] Nakajima Y, Ito T, Maeda Y, et al. Detection of pivaloylcarnitine in pediatric patients with hypocarnitinemia after long-term administration of pivalate-containing antibiotics. Tohoku J Exp Med 2010;221:309–13. [16] Shimizu K, Saito A, Shimada J, et al. Carnitine status and safety after administration of S-1108, a new oral cephem, to patients. Antimicrob Agents Chemother 1993;37:1043–9. [17] Holme E, Jodal U, Linstedt S, Nordin I. Effects of pivalic acid-containing prodrugs on carnitine homeostasis and on response to fasting in children. Scand J Clin Lab Invest 1992;52:361–72. [18] Ito T, Sugiyama N, Kobayashi M, et al. Alteration of ammonia and carnitine levels in short-term treatment with pivalic acid-containing prodrug. Tohoku J Exp Med 1995;175:43–53. [19] Makino Y, Sugiura T, Ito T, Sugiyama N, Koyama N. Carnitine-associated encephalopathy caused by long-term treatment with an antibiotic containing pivalic acid. Pediatrics 2007;120:e739–41. [20] Ensenauer R, Fingerhut R, Maier EM, et al. Newborn screening for isovaleric acidemia using tandem mass spectrometry: data from 1.6 million newborns. Clin Chem 2011;57:623–6.
Contents lists available at SciVerse ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
UPLC–MS/MS analysis of C5-acylcarnitines in dried blood spots Nils Janzen a, b, c,⁎, Ulrike Steuerwald a, d, Stefanie Sander a, Michael Terhardt a, Michael Peter a, Johannes Sander a a
Screening-Labor Hannover, Hannover, Germany Clinic for Pediatric Kidney, Liver and Metabolic Diseases, Hannover Medical School, Hannover, Germany Department of Neuropediatrics, Children’s Hospital, Ruhr University of Bochum, Bochum, Germany d Department of Occupational and Public Health, National Hospital System, Tórshavn, Faroe Islands b c
a r t i c l e
i n f o
Article history: Received 9 February 2013 Received in revised form 28 February 2013 Accepted 1 March 2013 Available online 13 March 2013 Keywords: Acylcarnitine screening C5-acylcarnitine isoforms Pivalic acid Second tier method
a b s t r a c t Background: Metabolic screening including newborn screening requires further differentiation of C5acylcarnitines in order to separate different metabolic disorders and to detect interferents like pivalic acid originating from antibiotics. Methods: For individual quantification of C5-acylcarnitine isoforms in dried blood spots we combined UPLC using a C18 column and gradient elution with tandem mass spectrometry in ESI+ mode. Results: Results were linear, coefficients of determination (R2) > 0.9977, intra- and inter-assay coefficients of variations b 5.2%, recovery 96.8–105.2%, limits of detection and quantitation b 0.2 μmol/L. Out of 29.309 blood samples of the isolated population of the Faroe Islands 56 exceeded the cut-off of 0.5 μmol/L for C5-acylcarnitine; 45 of which could be retested using the method described. Pivaloylcarnitine was identified in 43 samples, isovalerylcarnitine was found in two samples. Conclusions: The method was developed to allow direct re-analysis of samples showing elevated concentrations of C5-acylcarnitines in a metabolic screening program based on quantification of acylcarnitines after butylation. The technique should be especially useful in newborn screening for exclusion of false positives and for differentiation between isovaleric acidemia and 2-methylbutyryl-CoA dehydrogenase deficiency. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In the isolated population of the Faroe Islands primary carnitine deficiency (PCD, OMIM #212140), a rare metabolic defect in most other industrial countries, is rather frequent and has led to several deaths related to use of pivalic acid containing antibiotics [1]. Formation of pivaloylcarnitine has been found to cause severe carnitine deficiency not only in homozygous PCD but also in heterozygotes. In cooperation with the Faroese National Health System our laboratory tested 29.309 dried blood samples of individuals living on the Faroe Islands. The aim was to identify persons with yet undetected PCD as well as cases of secondary carnitine depletion in the Faroese population. The possibility to use dried blood spots was considered advantageous under the local conditions given for blood collection and for postal transport. In this material acylcarnitines and free carnitine were quantified by tandem mass spectrometry after butylation. Since acylcarnitines may undergo solvolysis during derivatization, the concentration of free carnitine may be overestimated [2]. However, for detection of carnitine transporter deficiency it is most important to determine levels of free
⁎ Corresponding author at: Screening-Labor Hannover, Postbox 91 10 09, 30430 Hannover, Germany. Tel.: + 49 5108 921630. E-mail address: [email protected] (N. Janzen). 0009-8981/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cca.2013.03.001
carnitine accurately. Therefore, free carnitine was also measured without the butylation step. The analytical methods applied were identical to those used for newborn metabolic screening [3]. The full spectrum of acylcarnitines was quantified in order to be able to distinguish primary carnitine deficiency (PCD) from secondary carnitine deficiency due to inborn errors of metabolism e.g. 3-methylcrotonyl-CoA carboxylase deficiency (OMIM #210200), glutaric aciduria type 1 (OMIM #231670) or due to medication. Being aware of the frequent use of pivalic acid containing antibiotics in the Faroese population we expected several samples with a high signal for C5-acylcarnitine caused by pivaloylcarnitine. In tandem mass spectrometry (MS/MS) analysis pivaloylcarnitine cannot be differentiated from other isobaric C5-acylcarnitines e.g. isovalerylcarnitine quantified for detection of isovaleric acidemia (IVA, OMIM #243500), an inborn error of metabolism which may lead to severe acidotic episodes and neurocognitive defects [4]. Another isobaric acylcarnitine is 2-methylbutyrylcarnitine which accumulates in 2methylbutyryl-CoA dehydrogenase deficiency (OMIM #610006) [5]. In addition, valerylcarnitine, a further isoform, might be elevated as a result of odd chain fatty acid degradation [6]. This substance is also a pharmaceutical intermediate and a technical product. In order to identify pivaloylcarnitine correctly and to avoid false positive results suggesting IVA or 2-methylbutyryl-CoA dehydrogenase deficiency we aimed at finding a rapid second tier test immediately
42
N. Janzen et al. / Clinica Chimica Acta 421 (2013) 41–45
applicable on the original blood spot. For this application we developed a method separating C5-acylcarnitines by ultra-high performance liquid chromatography (UPLC) prior to quantification by tandem mass spectrometry (MS/MS).
with four aliquots of water. Subsequently 90 μL was transferred into a 384 well microtiter plate.
2. Materials and methods
A XEVO UPLC–MS/MS system (Waters, Eschborn, Germany) consisting of a gradient pump, integrated solvent degasser, temperature controlled autosampler, column oven and triple quadrupole mass spectrometer was used. In the autosampler the microtiter plates were cooled down to 10 °C. 5 μL of the sample solution was injected onto an Acquity UPLC BEH C18 1.7 μm, 2.1 × 50 mm column (Waters). The chromatography was performed at a flow rate of 0.45 mL/min using a gradient between solvent A (water with 0.1% formic acid and HFBA each) and solvent B (methanol with the same modifiers as solvent A). The gradient started with 5% B and was held for 0.5 min, raised to 25% B within 0.5 min and held to 2.35 min, raised to 59% in 4.75 min and to 100% in 5.25 min. The gradient was held to 7.75 min and went back to the initial conditions within 0.25 min. The MS/MS system was equipped with an ESI interface. The nebulizing gas was nitrogen, argon was used for collision at a pressure of 4.0 × 10−3 mbar. The source temperature was set at 150 °C. Capillary voltage was 0.5 kV. The C5-acylcarnitines were detected in ESI positive mode (ESI+) by multiple reaction monitoring (MRM) mode. Collision energy was 20 eV, cone voltage was 28 V. The analyzed transitions were m/z 246.1 → 85 for pivaloyl-, 2-methylbutyryl-, isovaleryl- and valerylcarnitine and 255.2 → 85 for d9 isovalerylcarnitine. All results were calculated using the software TargetLynx 4.1 (Waters) and MS Excel 2010.
2.1. Materials UPLC/MS grade methanol, acetonitrile, formic acid and water were from Biosolve BV (Valkenswaard, The Netherlands). The lyophilized deuterated acylcarnitine standard mixture and extraction buffer contained in the neonatal acylcarnitine kit MassChrom® were purchased from Chromsystems (Munich, Germany). Heptafluorobutyric acid (HFBA) was from Fluka (Deisenhofen, Germany). The acylcarnitines pivaloylcarnitine (2,2-dimethylpropionic carnitine), valerylcarnitine, 2-methylbutyrylcarnitine and isovalerylcarnitine were purchased from ten Brink (VU University Medical Center Amsterdam, The Netherlands). 2.2. Standards and solutions Lyophilized internal standards from the kit containing deuterated d9-isovalerylcarnitine were diluted with 50 mL of the extraction buffer. One milligram of pivaloyl-, valeryl-, 2-methylbutyryl- and isovalerylcarnitine each were prepared in 1 mL water. The stock was diluted 2000 fold to reach the concentration of 0.5 μg/mL. 10 μL was added to 190 μL internal standard solution and further diluted with water to get a final concentration of 5 ng/mL acylcarnitine each. Calibrators for C5-acylcarnitines were made with final concentrations of 0 (blank), 1, 5, 15 and 25 μmol/L. Controls were prepared by adding pivaloyl-, valeryl-, 2methylbutyryl- and isovalerylcarnitine (0, 2 and 20 μmol/L each) to whole blood (containing EDTA as anticoagulant) which was then spotted onto the filter card. 2.3. Samples Blood from children and adults of all ages of the Faroese population was collected and spotted onto filter paper cards (Munktell & Filtrak, Bärenstein, Germany). The cards were dried at ambient temperature overnight before they were shipped to the laboratory. Here they were stored at 4–6 °C in a refrigerator until use. After completing the analyses all cards were sent back to the laboratory on the Faroe Islands. 2.4. Sample preparation Acylcarnitines were extracted from dried blood spots of 4.7 mm diameter in 96 well microtiter plates. 200 μL of extraction buffer containing internal standards was added and the plate was shaken for 30 min at ambient temperature. The plate was centrifuged (2800 U/min) for 10 min to separate the filter card material from the supernatant before it was removed with a pipette and diluted
2.5. Chromatography and mass spectrometry
3. Results 3.1. Linearity, precision, recovery, LOD, LOQ For the linear regression data were plotted against the concentration of the spiked samples. Peak heights were determined with TargetLynx® software, concentrations were calculated by use of MS Excel®. The parameter slope and intercept of the linear calibration functions are listed in Table 1. Intra-day and inter-day precision data were assessed over a period of 6 days by measuring control samples at two levels. At concentrations of 2 and 20 μmol/L, respectively, the imprecision data showed that the intra- and inter-day CVs for all the C5-acylcarnitines were not greater than 5.2%. Recovery from dried blood spot samples spiked with unlabeled acylcarnitines was determined six fold. As shown in Table 1 the range of recovery was between 96.8% and 105.2%. In order to determine the limit of detection (LOD) and limit of quantification (LOQ), the entire range of the linear calibration curve of the respective C5-acylcarnitines was used [7]. The standard error sy/x of the measured concentration (y-estimate in the regression equation) was used in place of the standard deviation of the blank sample [8]. The slope was calculated from the corresponding calibration curves. The LOD was then calculated as LOD = 3sy/x/slope, and
Table 1 Validation data of C5-acylcarnitines from dried blood spots. Slope
Intercept
Acylcarnitine
[a]
[b]
Pivaloylcarnitine
1.7401
0.3565
2-Methylbutyrylcarnitine
0.9209
Isovalerylcarnitine Valerylcarnitine
R2
LOD
LOQ
Level
Recovery
Intra-assay
Inter-assay
[μmol/L]
[μmol/L]
[μmol/L]
[%]
CV [%]
CV [%]
0.9978
0.03
0.11
0.1733
0.9982
0.02
0.08
1.8242
0.4085
0.9980
0.05
0.16
1.9983
0.4203
0.9977
0.03
0.09
2 20 2 20 2 20 2 20
98.4 103.2 96.8 105.1 98.3 105.2 100.1 105.1
3.9 1.0 3.9 1.8 3.8 1.0 3.5 1.4
5.0 1.9 5.2 2.3 5.2 1.4 4.5 2.4
N. Janzen et al. / Clinica Chimica Acta 421 (2013) 41–45
43
isovalerylcarnitine peak and a sample with a high pivaloylcarnitine peak. Out of 29.309 samples 56 blood spots tested positive for C5-acylcarnitines (cut-off 0.5 μmol/L). Positives were from adult persons as well as from children. Concentrations ranged from 0.51 to 9.5 μmol/L. The C5-acylcarnitine/acetylcarnitine ratio (cut-off b 0.01) was also elevated showing values from 0.017 up to 0.464. 11 samples had already been sent back and were not available for retesting, of the remaining 45 samples chromatographic separation of C5 isobares revealed pivaloylcarnitine only in 43 cases. There were two cases presenting isovalerylcarnitine (2.6 and 2.7 μmol/L respectively) in persons with a mild asymptomatic form of IVA. Unfortunately, since both persons (50 and 68 years old, respectively) did not suffer from specific symptoms they refused to have further biochemical and molecular genetic examinations. Ingestion of pivampicillin (Pondocillin®) or pivmecillinam (Selexid®) was confirmed in 35 individuals testing positive for pivaloylcarnitine, eight persons were not sure about their medication at the time of blood collection.
LOQ = 10sy/x/slope. The LOD and LOQ for the C5-acylcarnitines were between 0.02 and 0.16 μmol/L. No matrix effects were observed within the retention time range relevant for measurement of C5 acylcarnitines. Quantitation of all C5-acylcarnitines was linear with determination coefficients (R 2) greater than 0.9977. For the quantitation of 2-methylbutyrylcarnitine the sum of both isoforms was calculated. In Fig. 1 the chromatographic separation of the relevant acylcarnitines is demonstrated. The C5-acylcarnitines were well separated. The isoforms of (S,R) 2-methylbutyrylcarnitine gave a double peak without a full baseline separation.
3.2. Results of population screening Fig. 2 shows a chromatogram of a healthy person compared with a sample of a patient with isovaleric acidemia with a prominent
3.52
100
MRM of 3 Channels ES+ 255.2 > 85.0 1.96e5
%
internal standard d9 isovalerylcarnitine
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.62
100
3.90
MRM of 3 Channels ES+ 246.1 > 85.0 7.55e5
%
valerylcarnitine
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.53
100
3.90
MRM of 3 Channels ES+ 246.1 > 85.0 1.65e6
%
isovalerylcarnitine
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.41
100
2-methylbuturylcarnitine
3.90
MRM of 3 Channels ES+ 246.1 > 85.0 2.09e5
%
3.37
3.52
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.29
100
3.80
3.90
MRM of 3 Channels ES+ 246.1 > 85.0 1.83e5
%
pivaloylcarnitine
3.53 3.43
Time
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.90
Fig. 1. Chromatographic separation of C5-acylcarnitines and internal deuterated standard. The peaks of the isoforms of (S,R) 2-methylbutyrylcarnitine are not baseline separated.
44
N. Janzen et al. / Clinica Chimica Acta 421 (2013) 41–45
100
MRM of 3 Channels ES+ 255.2 > 85.0 1.92e5
3.51
A
%
internal standard d9 isovalerylcarnitine
3.00 100
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
B
3.90 MRM of 3 Channels ES+ 246.1 > 85.0 2.14e4
%
sample of a healthy subject
3.52
3.00 100
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.53
C
3.90 MRM of 3 Channels ES+ 246.1 > 85.0 1.65e6
%
isovalerylcarnitine
3.00 100
3.10
3.20
D
3.30
3.40
3.50
3.60
3.70
3.80
3.29
3.90 MRM of 3 Channels ES+ 246.1 > 85.0 1.83e5
%
pivaloylcarnitine
3.53 3.43
Time
3.00
3.10
3.20
3.30
3.40
3.50
3.60
3.70
3.80
3.90
Fig. 2. Chromatogram of the internal standard d9-isovalerylcarnitine (A); sample of a healthy subject (B); sample of a patient with isovaleric acidemia (C) and from a patient treated with pivalic acid containing antibiotics (D).
4. Discussion Several methods for differentiation of C5 isomers have been published but these are not suitable for immediate reanalysis of dried blood spot samples. Maeda et al. [9] described a rather time consuming method including solid phase extraction for chromatographic separation of acylcarnitines using serum and urine samples meaning that a second sample is required. This is also true for a rapid method by Ferrer et al. [10] separating C4- and C5-acylcarnitine isomers in plasma samples. Forni et al. [6] who analyzed plasma and dried blood spots derivatized the acylcarnitines prior to chromatographic separation. During this chemical process, however, acylcarnitines are partially being fragmented into free acids and carnitine [2]. This particularly applies for short chain acylcarnitines including C5 derivatives. An analytical procedure applicable as a second tier test in dried blood spots was
developed by Shigematsu et al. to quantify isovalerylglycine in cases of elevated C5-carnitine in newborn screening samples [11]. This method is useful for confirming IVA but it does not identify other C5 isoforms that might have been the cause for elevated C5 levels in the original screening test. Formation and renal excretion of pivaloylcarnitine following ingestion of pivalate-bound drugs have been known for a long time [12]. The renal loss of carnitine is considerable, because more than 98% of pivalate is excreted as pivaloylcarnitine [13,14]. Nakajima et al. [15] showed the same effect of antibiotics esterified with pivalic acid and concluded that long-term administration of pivalate-containing antibiotics should be avoided particularly in children. While the risk of metabolic decompensation following short term pivalic acid application may be rather low in persons with full function of carnitine uptake [16] there is an important risk in a population with more widespread carnitine
N. Janzen et al. / Clinica Chimica Acta 421 (2013) 41–45
transporter deficiency like on the Faroese Islands as well as in patients with other metabolic disorders and high carnitine excretion. Several cases of hypoketotic hypoglycemia caused by impairment of fatty acid degradation as a consequence of hypocarnitinemia have been described [17–19]. In such cases our new method is very well suited to uncover pivalic acid as the cause of hypocarnitinemia. Our method was originally established for a specific application but it should be especially useful for newborn metabolic screening since the majority of screening results for IVA using standard MS/MS are false positive. In a screening program done on 146.000 newborns in Japan Shigematsu et al. [11] found only one true positive case but 1064 false positives using a cut-off of 0.5 μmol/L for C5. Betalactam antibiotics esterified with pivalic acid for better enteral absorption are used in several countries. This includes industrial countries with an extended metabolic newborn screening program like Japan. Even in countries where pivalic acid containing antibiotics has been taken off the market a high percentage of false positive results for isovaleric academia is observed in newborn screening. In a series of 1.6 million newborns in Germany Ensenauer et al. [20] detected 24 patients suffering from IVA. For identification of possible positive cases the authors applied several acylcarnitine ratios in addition to defined cut-off values for C5 resulting in a recall rate of 0.024%; this seems low but it corresponds to a true positive to false positive ratio of 1:16. A similar ratio has been observed in our laboratory. With the introduction of our new method into our own screening program we will be able to avoid false positives and also to differentiate between isovaleric acidemia and 2-methylbutyryl-CoA dehydrogenase deficiency. References [1] Rasmussen J, Nielsen OW, Lund AM, Kober L, Djurhuus H. Primary carnitine deficiency and pivalic acid exposure causing encephalopathy and fatal cardiac events. J Inherit Metab Dis 2013;36:35–41. [2] Johnson DW. An acid hydrolysis method for quantification of plasma free and total carnitine by flow injection tandem mass spectrometry. Clin Biochem 2010;43: 1362–7. [3] Chace DH, Kalas TA, Naylor EW. Use of tandem mass spectrometry for multianalyte screening of dried blood specimens from newborns. Clin Chem 2003;49:1797–817. [4] Grunert SC, Wendel U, Lindner M, et al. Clinical and neurocognitive outcome in symptomatic isovaleric acidemia. Orphanet J Rare Dis 2012;7:9.
45
[5] Sass JO, Ensenauer R, Roschinger W, et al. 2-Methylbutyryl-coenzyme A dehydrogenase deficiency: functional and molecular studies on a defect in isoleucine catabolism. Mol Genet Metab 2008;93:30–5. [6] Forni S, Fu X, Palmer SE, Sweetman L. Rapid determination of C4-acylcarnitine and C5-acylcarnitine isomers in plasma and dried blood spots by UPLC–MS/MS as a second tier test following flow-injection MS/MS acylcarnitine profile analysis. Mol Genet Metab 2010;101:25–32. [7] Mani DR, Abbatiello SE, Carr SA. Statistical characterization of multiple-reaction monitoring mass spectrometry (MRM-MS) assays for quantitative proteomics. BMC Bioinformatics 2012;13(Suppl. 16):S9. [8] Anderson DJ. Determination of the lower limit of detection. Clin Chem 1989;35: 2152–3. [9] Maeda Y, Ito T, Ohmi H, et al. Determination of 3-hydroxyisovalerylcarnitine and other acylcarnitine levels using liquid chromatography–tandem mass spectrometry in serum and urine of a patient with multiple carboxylase deficiency. J Chromatogr B Analyt Technol Biomed Life Sci 2008;870:154–9. [10] Ferrer I, Ruiz-Sala P, Vicente Y, Merinero B, Perez-Cerda C, Ugarte M. Separation and identification of plasma short-chain acylcarnitine isomers by HPLC/MS/MS for the differential diagnosis of fatty acid oxidation defects and organic acidemias. J Chromatogr B Analyt Technol Biomed Life Sci 2007;860:121–6. [11] Shigematsu Y, Hata I, Tajima G. Useful second-tier tests in expanded newborn screening of isovaleric acidemia and methylmalonic aciduria. J Inherit Metab Dis 2010;33:S283–8. [12] Holme E, Greter J, Jacobson CE, et al. Carnitine deficiency induced by pivampicillin and pivmecillinam therapy. Lancet 1989;2:469–73. [13] Konishi M, Hashimoto H. Determination of pivaloylcarnitine in human plasma and urine by high-performance liquid chromatography with fluorescence detection. J Pharm Sci 1992;81:1038–41. [14] Brass EP. Pivalate-generating prodrugs and carnitine homeostasis in man. Pharmacol Rev 2002;54:589–98. [15] Nakajima Y, Ito T, Maeda Y, et al. Detection of pivaloylcarnitine in pediatric patients with hypocarnitinemia after long-term administration of pivalate-containing antibiotics. Tohoku J Exp Med 2010;221:309–13. [16] Shimizu K, Saito A, Shimada J, et al. Carnitine status and safety after administration of S-1108, a new oral cephem, to patients. Antimicrob Agents Chemother 1993;37:1043–9. [17] Holme E, Jodal U, Linstedt S, Nordin I. Effects of pivalic acid-containing prodrugs on carnitine homeostasis and on response to fasting in children. Scand J Clin Lab Invest 1992;52:361–72. [18] Ito T, Sugiyama N, Kobayashi M, et al. Alteration of ammonia and carnitine levels in short-term treatment with pivalic acid-containing prodrug. Tohoku J Exp Med 1995;175:43–53. [19] Makino Y, Sugiura T, Ito T, Sugiyama N, Koyama N. Carnitine-associated encephalopathy caused by long-term treatment with an antibiotic containing pivalic acid. Pediatrics 2007;120:e739–41. [20] Ensenauer R, Fingerhut R, Maier EM, et al. Newborn screening for isovaleric acidemia using tandem mass spectrometry: data from 1.6 million newborns. Clin Chem 2011;57:623–6.