- Email: [email protected]
Comparison of four different phenylalanine determination methods
Clinica Chimica Acta 264 (1997) 65–73
Comparison of four different phenylalanine determination methods ¨ Ralph Fingerhut*, Maren Stehn, Alfried Kohlschutter Department of Paediatrics, University of Hamburg, Hamburg, Germany Received 5 December 1996; received in revised form 2 April 1997; accepted 11 April 1997
Abstract The ISOLAB NCS phenylalanine determination kit has not been widely applied for neonatal screening and patient follow up in Europe until now. This method, based on fluorescence enhancement of a phenylalanine-ninhydrin reaction product by the dipeptide L-leucyl-L-alanine, was compared with three other procedures: (1) The Quantase kit (Shield Diagnostics) for enzymatic determination of phenylalanine, (2) the standard amino acid analysis by means of ion exchange chromatography, and (3) the Guthrie Test as a bacterial inhibition assay (BIA). Only authentic samples from PKU patients were evaluated: once with the NCS kit and at least once with one of the three other methods. There was good agreement between the results obtained by the NCS kit using dried blood specimens and either of the other three methods, as well as between the NCS kit using plasma samples and the Quantase kit and ion exchange chromatography. Plasma sample measurement by NCS proved advantageous because of the option of measuring each microtiter plate twice by resetting the calibrators, i.e. special standards for plasma samples could be used on the same plate. We conclude that this method should prove time saving and cost effective when both neonatal screening and patient follow up are carried out in the same laboratory. 1997 Elsevier Science B.V. Keywords: Phenylketonuria; New-born screening
*Corresponding author. Address for correspondence: University Children’s Hospital, Martinistr. 52, D20246 Hamburg, Germany. Tel.: 1 49 40 47173735; fax: 1 49 40 47176717; email: [email protected] 0009-8981 / 97 / $17.00 1997 Elsevier Science B.V. All rights reserved. PII S0009-8981( 97 )00074-0
66
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
1. Introduction Methods currently in use for detecting phenylketonuria by mass screening of new-borns include measuring phenylalanine in dried blood samples by bacterial inhibition assay [1], by fluorimetry [2] or by an enzymatic method [3]. For managing recognised phenylketonuric patients by dietary intervention, repeated phenylalanine determinations are required at intervals depending on age of patient and status of metabolic control. The bacterial inhibition assay has also been recommended for this purpose [4] but has obvious disadvantages due to the long time lapses before results are obtained and the necessity to process many samples at the same time. In our practice, it is highly desirable to have the analytical results very quickly after taking the blood from an individual patient, since in this way interpretation of recent dietary history and current metabolic status can be derived from the same session, including meaningful management adjustments. Within the population of our phenylketonuric patients, different types of blood samples are used. Between clinical visits, most patients or their parents take blood by finger or heel prick, aspirate it into heparinised capillary tubes and mail it to the laboratory. A few patients prefer sending Guthrie cards and some send venous blood obtained by a local physician. In order to cope with the demands of physicians, dietitians and the practice of sampling by the patients, we routinely determine plasma phenylalanine levels twice a week with the Quantase Kit. Guthrie cards are incorporated into the new-born screening program and can be measured daily. Total amino acid profiles are only obtained on special request of the physician, normally during clinical visits.
2. Materials and methods The blood samples for this study were taken from specimens from PKU patients that were sent to our laboratory for estimation of phenylalanine as part of routine management. When excess material was available, dried blood samples were prepared by spotting 20 ml of whole blood onto filter cards prior to the preparation of plasma. Bacterial inhibition assay for phenylalanine was performed as described by Guthrie and Susi [1]. Heparinised plasma from venous or capillary blood was obtained by centrifugation and deproteinised with trichloroacetic acid. Phenylalanine was then determined with the Quantase Kit from Shield Diagnostics, Dundee, UK. Amino acid analysis was performed by cation exchange chromatography on a Biotronic LC 5001 automated amino acid analyser, according to a modified method of Spackman [5]. Fluorimetric determinations of phenylalanine were carried out with the NCS phenylalanine determination kit from ISOLAB,
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
67
Mechelen, Belgium. For determination of plasma phenylalanine levels with the ISOLAB NCS kit, 20 ml of plasma was deproteinised with 20 ml of 0.36 mol / l trichloroacetic acid and, centrifuged for 5 min at 10 000 rpm in an Eppendorf centrifuge. Then 200 ml of 0.3 mol / l succinate buffer pH 5.8 and 20 ml of 0.36 mol / l sodium hydroxide were added, and the sample was centrifuged a second time. Standards and controls were treated in the same manner, and 25 ml of the supernatants was transferred to white microtiter plates. The liquid phenylalanine calibrators required for this determination were prepared in our laboratory. Enzymopathics test agar base was purchased from Merck, Darmstadt, Germany. b-2-thienylalanine was from Sigma, Deisenhofen, Germany, and Bacillus subtilis spore suspension from Difco, Detroit, USA. All other chemicals were of analytical grade from Merck, Darmstadt, Germany. Filter paper cards ¨ (Guthrie cards) MN 818 from Macherey–Nagel, Duren, Germany, are used in our new-born screening program.
3. Results Determination of phenylalanine levels from dried blood spots with the NCS kit was compared with the results of the BIA (dried blood specimen) and with the results of plasma phenylalanine determinations with the Quantase kit (enzymatic method) and the automated amino acid analysis (ion exchange chromatography) (Table 1). For the NCS kit and BIA, good agreement was found for phenylalanine concentrations up to 800 mol / l only. At higher concentrations, the values obtained by BIA were systematically higher than those obtained by NCS (Fig. 1a). For the comparison of the NCS kit with the Quantase kit, only 10 PKU samples were available, due to the fact that most PKU patients send their blood for the routine phenylalanine determination in capillary tubes too small to prepare additional dried blood specimens. The mean difference of these two methods (NCS vs. Quantase) lies at 2 172mol / l. This means that phenylalanine concentrations determined with the Quantase kit were lower throughout the whole range determined (400–1700 mmol / l) but without a recognisable trend. The values were normally distributed and all lay well within the 62 3 SD range (Fig. 1b). For the comparison of the NCS kit with automated amino acid analysis (AA) 15 subjects could be examined. The mean difference (NCS 2 AA) was 2 25 mmol / l with an SD of 91 mmol / l. Most values lie within the 62 3 SD range, and more than 50% lie within the 61 3 SD range (Fig. 1c). When liquid samples were used instead of dried blood, the fluorescence increased linearly with increasing phenylalanine concentrations up to at least 3000 mmol / l (data not shown). With this method, determination of plasma
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
68
Table 1 Phenylalanine concentrations in dried blood samples (NCS kit and BIA) and plasma samples (Quantase kit and amino acid analysis by ion exchange chromatography, AA) Subject
NCS kit
BIA
Subject
NCS kit
Quantase kit
Subject
NCS kit
AA
(n 5 34)
[mmol/l]
[mmol/l]
(n 5 10)
[mmol/l]
[mmol/l]
(n 5 15)
[mmol/l]
[mmol/l] 55
1
70
120
35
340
503
45
60
2
149
61
36
382
521
46
252
55
3
104
121
37
636
958
47
226
212
4
140
121
38
758
952
48
282
291
5
181
120
39
875
933
49
248
333
6
188
120
40
1202
1303
50
290
364
7
194
120
41
1155
1430
51
382
455
8
226
120
42
1408
1594
52
405
509
9
228
121
43
1450
1582
53
448
545
10
232
120
44
1591
1745
54
713
855
11
230
151
55
1150
1109
12
202
182
56
1114
1430
13
228
181
57
1175
1430
14
289
120
58
1264
1430
15
195
242
59
1491
1503
16
285
182
17
194
303
18
256
242
19
293
242
20
238
303
21
326
242
22
315
303
23
339
303
24
356
303
25
379
303
26
451
303
27
479
303
28
593
485
29
537
727
30
718
727
31
755
909
32
664
1212
33
730
1212
34
801
1220
phenylalanine using the NCS kit was also compared with the Quantase kit and ion exchange chromatography (Table 2). For the comparison of the NCS kit with AA, 38 subjects were examined, with a concentration range from about 20 to 1300 mmol / l. The mean difference (NCS 2 AA) was at 2 10 mmol / l with a SD of 50 mmol / l. Here again, the
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
69
Fig. 1. Comparison of phenylalanine determinations by NCS kit (with dried blood specimens) with plasma phenylalanine determinations using the Quantase kit, ion exchange chromatography (AA), and classical Guthrie test (BIA). The difference in phenylalanine concentration is plotted against the average phenylalanine concentration determined by the two methods (NCS and BIA, Fig. la; NCS and Quantase, Fig. lb; NCS and AA, Fig. lc).
70
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
Table 2 Phenylalanine concentrations in plasma samples (NCS kit; amino acid analysis by ion exchange chromatography, AA; Quantase kit) Subject (n 5 38) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
NCS kit [mmol / l] 24 27 42 48 36 68 45 36 61 33 64 52 65 67 76 79 64 108 127 127 118 142 182 227 314 326 355 355 439 445 438 488 781 856 1005 1113 1202 1208
AA [mmol / l] 36 55 48 48 61 30 55 67 55 85 61 73 73 73 73 73 91 48 67 85 109 103 230 261 218 212 224 406 376 430 455 552 836 970 1255 1242 1188 1309
Subject (n 5 48) 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
NCS kit [mmol / l] 55 24 42 95 48 61 127 103 172 153 103 142 172 170 272 224 244 178 326 314 272 236 285 372 355 291 449 439 478 480 438 573 488 484 496 642 646 679 711 781 852 789 856 1005 970 1169 1161 1202
Quantase kit [mmol / l] 55 91 103 85 133 121 67 133 78 109 182 170 145 176 127 176 218 321 255 285 327 375 345 297 358 442 448 509 539 594 636 564 667 794 788 703 733 812 930 994 1018 1085 1176 1297 1333 1424 1491 1558
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
71
values were normally distributed and are mostly within the 62 3 SD range (Fig. 2a). Comparison of the NCS kit with the Quantase kit revealed a mean difference (NCS 2 Quantase) of 2 97 mmol / l with a SD of 107 mmol / l (Fig. 2b). This is about the same finding as for the comparison with dried blood specimens (see Fig. 1b). Here, however, a clear trend can be seen. At higher phenylalanine concentrations, the difference (NCS 2 Quantase) increases and becomes more negative.
Fig. 2. Comparison of phenylalanine determinations by NCS kit (with plasma specimens) with plasma phenylalanine determinations using the Quantase kit, and ion exchange chromatography (AA). The difference in phenylalanine concentration is plotted against the average phenylalanine concentration determined by the two methods (NCS and AA, Fig. 2a; NCS and Quantase, Fig. 2b).
72
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
4. Discussion The fast and accurate determination of blood phenylalanine levels is essential for an effective treatment of PKU patients, especially during the first years of life. Rohr et al. [4] have recently described 25 years of experience with the BIA in their neonatal screening program, and in patient follow up. Although we have managed only a handful of patients in a similar manner in our department, we feel that a more quantitative method for the determination of blood phenylalanine levels is important to achieve optimal metabolic control, at least in infants and pregnant women with PKU. The fluorimetric determination of phenylalanine with the NCS kit proved to be a reliable method for dried blood and plasma samples, compared to plasma phenylalanine determinations by Quantase kit or ion exchange chromatography. Nevertheless, the agreement was best between NCS and automated amino acid analysis, with mean differences for both dried blood samples and plasma samples of 2 25 mmol / l and 2 10 mmol / l, respectively, and standard deviations of 91 mmol / l and 50 mmol / l, respectively. When neonatal screening and patient follow up is carried out in the same laboratory, such as in our clinic, the NCS method seems to be advantageous. Patient monitoring can be included in the screening procedure with minimal additional costs and workload. It can also be achieved without losing accuracy by using dried blood samples instead of plasma. Concerning costs, the BIA is certainly the least expensive in terms of material cost but at the same time the most labour intensive. Because the BIA cannot distinguish between normal and decreased phenylalanine levels, we think this test is not adequate for patient follow up. The costs for the NCS Kit and Quantase Kit would be about the same if the NCS test were to be performed as stand-alone assay. However, when it is included in the neonatal screening procedure, there are essentially no additional reagent costs (apart from TCA, sodium hydroxide and buffer) because surplus reagent from the screening procedure can be used for the determination. Automated amino acid analysis is by far the most expensive of the four methods and therefore it is not recommended for routine phenylalanine determination. For patient follow up, however, it is from time to time necessary to determine a whole plasma amino acid spectrum in order to detect deficiencies of essential amino acids or other amino acid disequilibria.
Acknowledgments We thank ISOLAB NV for financial support of this study and Dr. J. Knispel, Hamburg, for his language advice.
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
73
References [1] Guthrie R, Susi A. A simple phenylalanine method for detecting phenylketonuria in large populations of new-born infants. Pediatrics 1963;32:338–43. [2] Hill JB, Summer GK, Pender MW, Roszell NO. An automated procedure for blood phenylalanine. Clin Chem 1965;11:541–6. [3] Keffler S, Denmeade R, Green A. Neonatal screening for phenylalanine: Evaluation of an automated enzymatic method. Ann Clin Biochem 1994;31:134–9. [4] Rohr FJ, Allred EN, Turner M, Simmons J, Levy BL. Use of the Guthrie bacterial inhibition assay to monitor blood phenylalanine for dietary treatment of phenylketonuria. Screening 1996;2:205–11. [5] Spackman DH, Stein VH, Moore S. Automated recording apparatus for use in the chromatography of amino acids. Anal Chem 1958;30:1190–206.
Comparison of four different phenylalanine determination methods ¨ Ralph Fingerhut*, Maren Stehn, Alfried Kohlschutter Department of Paediatrics, University of Hamburg, Hamburg, Germany Received 5 December 1996; received in revised form 2 April 1997; accepted 11 April 1997
Abstract The ISOLAB NCS phenylalanine determination kit has not been widely applied for neonatal screening and patient follow up in Europe until now. This method, based on fluorescence enhancement of a phenylalanine-ninhydrin reaction product by the dipeptide L-leucyl-L-alanine, was compared with three other procedures: (1) The Quantase kit (Shield Diagnostics) for enzymatic determination of phenylalanine, (2) the standard amino acid analysis by means of ion exchange chromatography, and (3) the Guthrie Test as a bacterial inhibition assay (BIA). Only authentic samples from PKU patients were evaluated: once with the NCS kit and at least once with one of the three other methods. There was good agreement between the results obtained by the NCS kit using dried blood specimens and either of the other three methods, as well as between the NCS kit using plasma samples and the Quantase kit and ion exchange chromatography. Plasma sample measurement by NCS proved advantageous because of the option of measuring each microtiter plate twice by resetting the calibrators, i.e. special standards for plasma samples could be used on the same plate. We conclude that this method should prove time saving and cost effective when both neonatal screening and patient follow up are carried out in the same laboratory. 1997 Elsevier Science B.V. Keywords: Phenylketonuria; New-born screening
*Corresponding author. Address for correspondence: University Children’s Hospital, Martinistr. 52, D20246 Hamburg, Germany. Tel.: 1 49 40 47173735; fax: 1 49 40 47176717; email: [email protected] 0009-8981 / 97 / $17.00 1997 Elsevier Science B.V. All rights reserved. PII S0009-8981( 97 )00074-0
66
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
1. Introduction Methods currently in use for detecting phenylketonuria by mass screening of new-borns include measuring phenylalanine in dried blood samples by bacterial inhibition assay [1], by fluorimetry [2] or by an enzymatic method [3]. For managing recognised phenylketonuric patients by dietary intervention, repeated phenylalanine determinations are required at intervals depending on age of patient and status of metabolic control. The bacterial inhibition assay has also been recommended for this purpose [4] but has obvious disadvantages due to the long time lapses before results are obtained and the necessity to process many samples at the same time. In our practice, it is highly desirable to have the analytical results very quickly after taking the blood from an individual patient, since in this way interpretation of recent dietary history and current metabolic status can be derived from the same session, including meaningful management adjustments. Within the population of our phenylketonuric patients, different types of blood samples are used. Between clinical visits, most patients or their parents take blood by finger or heel prick, aspirate it into heparinised capillary tubes and mail it to the laboratory. A few patients prefer sending Guthrie cards and some send venous blood obtained by a local physician. In order to cope with the demands of physicians, dietitians and the practice of sampling by the patients, we routinely determine plasma phenylalanine levels twice a week with the Quantase Kit. Guthrie cards are incorporated into the new-born screening program and can be measured daily. Total amino acid profiles are only obtained on special request of the physician, normally during clinical visits.
2. Materials and methods The blood samples for this study were taken from specimens from PKU patients that were sent to our laboratory for estimation of phenylalanine as part of routine management. When excess material was available, dried blood samples were prepared by spotting 20 ml of whole blood onto filter cards prior to the preparation of plasma. Bacterial inhibition assay for phenylalanine was performed as described by Guthrie and Susi [1]. Heparinised plasma from venous or capillary blood was obtained by centrifugation and deproteinised with trichloroacetic acid. Phenylalanine was then determined with the Quantase Kit from Shield Diagnostics, Dundee, UK. Amino acid analysis was performed by cation exchange chromatography on a Biotronic LC 5001 automated amino acid analyser, according to a modified method of Spackman [5]. Fluorimetric determinations of phenylalanine were carried out with the NCS phenylalanine determination kit from ISOLAB,
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
67
Mechelen, Belgium. For determination of plasma phenylalanine levels with the ISOLAB NCS kit, 20 ml of plasma was deproteinised with 20 ml of 0.36 mol / l trichloroacetic acid and, centrifuged for 5 min at 10 000 rpm in an Eppendorf centrifuge. Then 200 ml of 0.3 mol / l succinate buffer pH 5.8 and 20 ml of 0.36 mol / l sodium hydroxide were added, and the sample was centrifuged a second time. Standards and controls were treated in the same manner, and 25 ml of the supernatants was transferred to white microtiter plates. The liquid phenylalanine calibrators required for this determination were prepared in our laboratory. Enzymopathics test agar base was purchased from Merck, Darmstadt, Germany. b-2-thienylalanine was from Sigma, Deisenhofen, Germany, and Bacillus subtilis spore suspension from Difco, Detroit, USA. All other chemicals were of analytical grade from Merck, Darmstadt, Germany. Filter paper cards ¨ (Guthrie cards) MN 818 from Macherey–Nagel, Duren, Germany, are used in our new-born screening program.
3. Results Determination of phenylalanine levels from dried blood spots with the NCS kit was compared with the results of the BIA (dried blood specimen) and with the results of plasma phenylalanine determinations with the Quantase kit (enzymatic method) and the automated amino acid analysis (ion exchange chromatography) (Table 1). For the NCS kit and BIA, good agreement was found for phenylalanine concentrations up to 800 mol / l only. At higher concentrations, the values obtained by BIA were systematically higher than those obtained by NCS (Fig. 1a). For the comparison of the NCS kit with the Quantase kit, only 10 PKU samples were available, due to the fact that most PKU patients send their blood for the routine phenylalanine determination in capillary tubes too small to prepare additional dried blood specimens. The mean difference of these two methods (NCS vs. Quantase) lies at 2 172mol / l. This means that phenylalanine concentrations determined with the Quantase kit were lower throughout the whole range determined (400–1700 mmol / l) but without a recognisable trend. The values were normally distributed and all lay well within the 62 3 SD range (Fig. 1b). For the comparison of the NCS kit with automated amino acid analysis (AA) 15 subjects could be examined. The mean difference (NCS 2 AA) was 2 25 mmol / l with an SD of 91 mmol / l. Most values lie within the 62 3 SD range, and more than 50% lie within the 61 3 SD range (Fig. 1c). When liquid samples were used instead of dried blood, the fluorescence increased linearly with increasing phenylalanine concentrations up to at least 3000 mmol / l (data not shown). With this method, determination of plasma
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
68
Table 1 Phenylalanine concentrations in dried blood samples (NCS kit and BIA) and plasma samples (Quantase kit and amino acid analysis by ion exchange chromatography, AA) Subject
NCS kit
BIA
Subject
NCS kit
Quantase kit
Subject
NCS kit
AA
(n 5 34)
[mmol/l]
[mmol/l]
(n 5 10)
[mmol/l]
[mmol/l]
(n 5 15)
[mmol/l]
[mmol/l] 55
1
70
120
35
340
503
45
60
2
149
61
36
382
521
46
252
55
3
104
121
37
636
958
47
226
212
4
140
121
38
758
952
48
282
291
5
181
120
39
875
933
49
248
333
6
188
120
40
1202
1303
50
290
364
7
194
120
41
1155
1430
51
382
455
8
226
120
42
1408
1594
52
405
509
9
228
121
43
1450
1582
53
448
545
10
232
120
44
1591
1745
54
713
855
11
230
151
55
1150
1109
12
202
182
56
1114
1430
13
228
181
57
1175
1430
14
289
120
58
1264
1430
15
195
242
59
1491
1503
16
285
182
17
194
303
18
256
242
19
293
242
20
238
303
21
326
242
22
315
303
23
339
303
24
356
303
25
379
303
26
451
303
27
479
303
28
593
485
29
537
727
30
718
727
31
755
909
32
664
1212
33
730
1212
34
801
1220
phenylalanine using the NCS kit was also compared with the Quantase kit and ion exchange chromatography (Table 2). For the comparison of the NCS kit with AA, 38 subjects were examined, with a concentration range from about 20 to 1300 mmol / l. The mean difference (NCS 2 AA) was at 2 10 mmol / l with a SD of 50 mmol / l. Here again, the
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
69
Fig. 1. Comparison of phenylalanine determinations by NCS kit (with dried blood specimens) with plasma phenylalanine determinations using the Quantase kit, ion exchange chromatography (AA), and classical Guthrie test (BIA). The difference in phenylalanine concentration is plotted against the average phenylalanine concentration determined by the two methods (NCS and BIA, Fig. la; NCS and Quantase, Fig. lb; NCS and AA, Fig. lc).
70
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
Table 2 Phenylalanine concentrations in plasma samples (NCS kit; amino acid analysis by ion exchange chromatography, AA; Quantase kit) Subject (n 5 38) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
NCS kit [mmol / l] 24 27 42 48 36 68 45 36 61 33 64 52 65 67 76 79 64 108 127 127 118 142 182 227 314 326 355 355 439 445 438 488 781 856 1005 1113 1202 1208
AA [mmol / l] 36 55 48 48 61 30 55 67 55 85 61 73 73 73 73 73 91 48 67 85 109 103 230 261 218 212 224 406 376 430 455 552 836 970 1255 1242 1188 1309
Subject (n 5 48) 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
NCS kit [mmol / l] 55 24 42 95 48 61 127 103 172 153 103 142 172 170 272 224 244 178 326 314 272 236 285 372 355 291 449 439 478 480 438 573 488 484 496 642 646 679 711 781 852 789 856 1005 970 1169 1161 1202
Quantase kit [mmol / l] 55 91 103 85 133 121 67 133 78 109 182 170 145 176 127 176 218 321 255 285 327 375 345 297 358 442 448 509 539 594 636 564 667 794 788 703 733 812 930 994 1018 1085 1176 1297 1333 1424 1491 1558
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
71
values were normally distributed and are mostly within the 62 3 SD range (Fig. 2a). Comparison of the NCS kit with the Quantase kit revealed a mean difference (NCS 2 Quantase) of 2 97 mmol / l with a SD of 107 mmol / l (Fig. 2b). This is about the same finding as for the comparison with dried blood specimens (see Fig. 1b). Here, however, a clear trend can be seen. At higher phenylalanine concentrations, the difference (NCS 2 Quantase) increases and becomes more negative.
Fig. 2. Comparison of phenylalanine determinations by NCS kit (with plasma specimens) with plasma phenylalanine determinations using the Quantase kit, and ion exchange chromatography (AA). The difference in phenylalanine concentration is plotted against the average phenylalanine concentration determined by the two methods (NCS and AA, Fig. 2a; NCS and Quantase, Fig. 2b).
72
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
4. Discussion The fast and accurate determination of blood phenylalanine levels is essential for an effective treatment of PKU patients, especially during the first years of life. Rohr et al. [4] have recently described 25 years of experience with the BIA in their neonatal screening program, and in patient follow up. Although we have managed only a handful of patients in a similar manner in our department, we feel that a more quantitative method for the determination of blood phenylalanine levels is important to achieve optimal metabolic control, at least in infants and pregnant women with PKU. The fluorimetric determination of phenylalanine with the NCS kit proved to be a reliable method for dried blood and plasma samples, compared to plasma phenylalanine determinations by Quantase kit or ion exchange chromatography. Nevertheless, the agreement was best between NCS and automated amino acid analysis, with mean differences for both dried blood samples and plasma samples of 2 25 mmol / l and 2 10 mmol / l, respectively, and standard deviations of 91 mmol / l and 50 mmol / l, respectively. When neonatal screening and patient follow up is carried out in the same laboratory, such as in our clinic, the NCS method seems to be advantageous. Patient monitoring can be included in the screening procedure with minimal additional costs and workload. It can also be achieved without losing accuracy by using dried blood samples instead of plasma. Concerning costs, the BIA is certainly the least expensive in terms of material cost but at the same time the most labour intensive. Because the BIA cannot distinguish between normal and decreased phenylalanine levels, we think this test is not adequate for patient follow up. The costs for the NCS Kit and Quantase Kit would be about the same if the NCS test were to be performed as stand-alone assay. However, when it is included in the neonatal screening procedure, there are essentially no additional reagent costs (apart from TCA, sodium hydroxide and buffer) because surplus reagent from the screening procedure can be used for the determination. Automated amino acid analysis is by far the most expensive of the four methods and therefore it is not recommended for routine phenylalanine determination. For patient follow up, however, it is from time to time necessary to determine a whole plasma amino acid spectrum in order to detect deficiencies of essential amino acids or other amino acid disequilibria.
Acknowledgments We thank ISOLAB NV for financial support of this study and Dr. J. Knispel, Hamburg, for his language advice.
R. Fingerhut et al. / Clinica Chimica Acta 264 (1997) 65 – 73
73
References [1] Guthrie R, Susi A. A simple phenylalanine method for detecting phenylketonuria in large populations of new-born infants. Pediatrics 1963;32:338–43. [2] Hill JB, Summer GK, Pender MW, Roszell NO. An automated procedure for blood phenylalanine. Clin Chem 1965;11:541–6. [3] Keffler S, Denmeade R, Green A. Neonatal screening for phenylalanine: Evaluation of an automated enzymatic method. Ann Clin Biochem 1994;31:134–9. [4] Rohr FJ, Allred EN, Turner M, Simmons J, Levy BL. Use of the Guthrie bacterial inhibition assay to monitor blood phenylalanine for dietary treatment of phenylketonuria. Screening 1996;2:205–11. [5] Spackman DH, Stein VH, Moore S. Automated recording apparatus for use in the chromatography of amino acids. Anal Chem 1958;30:1190–206.