Cohen AB, Geczy D, James HL. Interaction of a1-antitrypsin with porcine trypsin. Biochemis ..... Edouard Herriot. Place d'Arsonval. 69 437 Lyon Cedex 3, France.
1298
Letters
Monoclonal ter molecules also Table 1. Stabiilty of lyophilized 90:10 calIbrator at 25 #{176}C after reconstitution. Immunoglobulln greatly reduces the Group 1 Group 2 Group 3 Interferences In force of hydrogen Days stored tPSA fPSA tPSA IPSA tPSA fPSA Measurement of bonds. 0 56.39 8.46 28.85 4.10 12.05 1.66 Serum Inorganic No generally 30 56.60 8.61 30.81 4.74 12.81 1.62 Phosphate wIth a accepted mecha60 58.51 8.83 29.95 4.97 12.23 1.90 New Modified nism yet explains 90 56.36 8.99 29.82 4.51 11.71 1.81 how serpins are Reagent 135 54.48 6.88 28.81 4.32 12.20 1.55 complexed with The lyophilized 90:10 calibrators (90% PSA-ACTcomplex, 10% free PSA) were incubated at 25 #{176}C. At various To the Editor: proteinases, beintervals, one of the calibrators was reconstituted and diluted to three different concentration (.xg/L( groups. The interference of cause no detailed Each group was measured for total PSA ItPSA) by the Tosoh assay and free PSA (fPSA) by the Diannon assay. monocbonal immux-ray crystal strucnogbobulins (Mture of the El ase inhibitors. Annu Rev Biochem 1983:52: Igs) on inorganic complex is available. In contrast to the 655-709. phosphate measurement has been known above proposed mechanism, a recent 6. Carrell RW, Evans DL. Serpins: mobile conforfor several years [1-3]. In the phosphomations in a family of proteinase inhibitors. study supports the hypothesis that the Curn Opin Stnuct Biol 1992:2:438-46. molybdate method, monitored at 340 native El complex is a covalently linked 7. Cohen AB, Geczy D, James HL. Interaction of nm, the interference was rebated to the a1-antitrypsin with porcine trypsin. Biochemisserpin acyl-enzyme complex with a precipitation of M-Igs with strongly acid try 1978:17:392-400. cleaved inhibitor at the PI-Pi’ peptide 8. Longas MO, Finlay TH. The covalent nature of reagents [2, 3]. Such an interference is bond [17]. If it is true that the PSA-AGT the human antithrombin Ill-thrombin bond. Bioencountered with the Boehringer Mannchem 1 1980:189:481-9. complex is an acyl-enzyme complex, a heim (BM) phosphate test, which uses 9. Mahoney WC, Kurachi K, Henmodson MA. Forwater molecule is necessary for deacylathe phosphomolybdate method [4-6]. mation and dissociation of the covalent complexes between trypsin and two homologous tion to release the cleaved ACT. Without Faced, like others, with increased phosinhibitors, a1-antitrypsin and antithnombin Ill. water, dissociation of such a complex is phatemia in patients with M-Igs and also Eur J Biochem 1980:105:545-52. impossible, making lyophilization or with negative interference with lipemic 10. Monoi M, Aoki N. On the interaction of plasmin inhibitor and non-covalent weak bondsamples, we found and reported that freeze-drying a powerful technique for ings between the inhibitor and proteases. 810decreasing the ionic strength of both stabilizing the acyl-enzyme complex. In chin Biophys Acta 1977:482:412-20. reagents through dilution in purified waIi.. Tomono T, Sawada E. Preparation of anhydrolight of either of the above recent studies, thrombin and its interaction with plasma antiter prevented both precipitation of the the PSA-AGT complex should be much thrombin Ill. Acta Haematol Jpn 1986:49:969M-Igs and the interference from lipids more stable in a lyophilized form than in 79. [6]. Recently, the BM phosphate test has 12. Shieh B-H, Potempa I, Travis I. The use of aqueous solutions, which is precisely a2-antiplasmin as a model for the demonstrabeen modified and is reported to be M-Ig what we have found in our efforts to tion of complex reversibility in serpins. I Biol interference free [4, 5]. We therefore unChem 1989:264:13420-3. produce an acceptable international standertook a study to verify whether the new 13. Longstaff C, Gaffney PJ. Serpin-serine prodard for PSA immunoassays. reagent formulation was truly free from tease binding kinetics: a2-antiplasmin as a model inhibitor. Biochemistry 1991;3O:979interference either from M-Igs or from 86. lipids. 14. Matheson NR, van Halbeek H, Travis J. EviThis work was supported in part by the We selected 100 patients whose serum dence for a tetrahedral intermediate complex during serpin-proteinase interactions. I Biol Richard M. Lucas Foundation. exhibited an unequivocal monocbonab Chem 1991;266:13489-91. component (M-component) upon ebec15. Potempa I, Korzus E, Travis I. The serpin References trophoresis. Of these, 69 patients were superfamily of proteinase inhibitor: structure, 1. Pettersson K, Piinonen 1, Sepp#{225}l#{227} M, Liukfunction, and regulation. J Biol Chem 1994: followed up for B-cell malignancies: mu!konen L, Chnistensson A, Matikainen M-T, et al. 269:15957-60. tiple myeloma (n = 57), Waldenstrom Free and complexed prostate-specific antigen 16. Skniven K, Wikoff WR, Patston PA, Tausk F, macroglobulinemia (n = 7), plasmacy(PSA(: in vitro stability, epitope map, and develSchapira M, Kaplan AP, Bock SC. Substrate opment of immunofluorometric assays for speproperties of Cl inhibitor Ma lalanine 434toma (n = 2), chronic lymphocytic leucific and sensitive detection of free PSA and glutamic acid). J Biol Chem 1991:266:9216kemia (n = 4), and amyboidosis (n 1). PSAa1afltichymotiypsin complex. Clin Chem 21. The 41 remaining patients were followed 1995:41:1480-8. 17. Wilczynska M, Fa M, Ohlsson P-I. Ny T. The inhibition mechanism of serpins. I Biol Chem 2. Christensson A. Laurell C-B, Lilja H. Enzymatic up for a monoclonal gammopathy of 1995:270:29652-5. activity of prostate-specific antigen and its reundetermined significance (MGUS). action with extnacellular senine proteinase inThe study was approved by the Medical hibitors. EunI Biochem 1990:194:755-63. Zuxiong Chen Ethical Committee of the hospital. 3. Chen Z, Pnestigiacomo A, StameyT. Purification Anthony Prestigiacomo and characterization of prostate-specific antiPatients’ sera assayed for total protein gen (PSA) complexed to a1-antichymotrypsin: Thomas A. Stamey* concentration by either refractometry or potential reference material for international Dept. of Urol. S287 the biuret method were submitted to standardization of PSA immunoassays. Clin StanfordUniv.Med. Center Chem 1995:41:1273-82. ebectrophoresis with the REP#{174} analyzer 300 PasteurDr. 4. Huben R, Carell R. Implications of the three(Helena, Saint-Leu-la-For#{234}t, France). Stanford, CA 94305-5118 dimensional structure of a1-antitrypsin for Immunoelectrophoresis and immimofixstructure and function of serpins. Biochemistry ation were used to classify and type the 1989:28:8951-66. *corresponding author. 5. Travis I, Salvesen GS. Human plasma proteinM-components. The subclasses of M-
Clinical Chemistry
42, No. 8, 1996
1299
Table 1. Comparison of phosphate concentrations obtained wIth the BM methods and the Ektachem method. Total proteins, g/L
Phosphate, mmol/L*
Abnormal kinetic
BMu
BMm
Bmd
Ek
BMu
BMm
2.23 2.39 2.58 1.12 1.61 1.64 0.49 0.92 1.33
1.52 1.55 1.64 1.06 1.21 1.26 0.37 0.81 1.11
1.25 1.28 1.39 0.92 1.23 1.19 0.40 0.85 1.05
1.25 1.34 1.54 0.96 1.19 1.30 0.40 0.78 1.13
+
+
+
+
+
+
+
+
86
1.63
1.50
1.50
1.41
+
G2k + G2A
71
1.09
1.00
1.00
0.97
+
MGUS
MK
71
1.01
0.89
0.94
0.92
+
13
W
MK
82
1.20
1.13
1.14
1.14
+
14
MM
90
1.47
1.33
1.37
1.37
+
Patients
Pathoiogy
M-ig
1 2 3 4 5 6 7 8 9
MM MM MM MM MM MM MM P MM
G1K G1K G1A G1K G1K G1A MK G2A G1K
94 94 106 90 87 111 79 69 94
10
MM
G1K
11
MM
12
a MM
=
multiple myeloma, P
=
plasmacytoma, MGUS
=
monoclonal gammopathy of undetermined significance, W
=
BMd
+ + + + +
Waldenstr#{244}m macroglobulinemia.
“BMu, unmodified BM method; BMm, modified BM method; BMd, in-house modified BM method Idiluted reagents); Ek, Ektachem method.
IgGs were determined by immunoebectrophoresis with an IgG subclass typing set (The Binding Site; Immunotech, Marseilbe, France). IgG, -A, and -M were quantified with the Behring Nephebometer Automat (BNA Behring, Rueib Malmaison, France) according to the manufacturer’s protocols and reagents. Samples showing well-defined M-components were assayed immediately for inorganic phosphate concentrations on the BM (Meyban, France) Hitachi 747 and the Ektachem 700 XRC (Ek; Johnson & Johnson Clinical Diagnostics, Les Ulis, France). The Ek colorimetric method for phosphate is based on reduction of the ammonium phosphomobybdate complex to a blue chromophore measured by reflectance spectrophotometry at 680 nm. No interference has been described with the Ek method, which is explained by the multilayer configuration of the Ektachem slides, especially the presence of a spreading layer that retains proteins. Using the BM/Hitachi 747, we performed three kinds of assays with the Inorganic Phosphorus UV test kit (cat. nos. 1127933 and 1128019): BM unmodified kit (BMu), BM kit modified by the manufacturer (BMm), and “in-house” modified (diluted reagents, see below; BMd). In all cases, this test is a tworeagent technique (Ri and R2) that quantifies the formation of ammonium phosphomolybdate by measuring absorbance bichromatically (principal A = 340 nm, secondary A = 660 nm). In the BMm reagents, the manufacturer removed the
NaC1 and replaced the detergent Corexit (anionic pobyglycol ether containing 100 mL/L isopropanob) with Thesit (nonaethylene glycol monododecyl ether) in Rb. For the BMd method, we used the BMu reagents Ri and R2 diluted fivefold in purified water, as previously described [6]. The presence of interference was ascertained by associating two criteria: phosphate results obtained with BMu, BMm, or BMd methods differing by 5% from results by the Ek method (the comparison method), and abnormal aspects of the kinetic reaction on the BM/ Hitachi 747. To investigate the effect of lipemia, we assayed serial dilutions of Intralipid#{174} added to a clear pool of sera (final triglycerides concentration, 1.87-28 mmoV L), determining the phosphate concentration in each sample with the BMm method. Of the 100 sera tested in the preliminary study, 14 demonstrated an interference with the BMu method, 4 with the BMm method, and none with the BMd method. Concentrations of phosphate and total proteins, the presence or absence of an abnormal reaction kinetic, the patients’ pathology, and the M-Ig types of these 14 interferent sera are detailed in Table I. Multiple myeloma was diagnosed in the four BMm interferent samples, all M-Igs were IgG1, and total protein exceeded 90 g/L. Taking into consideration only the patients with total protein >90 g/L, the percentage interference with the BMrn method was 17%, compared with 4% for
the whole cohort of samples. In a further study, we booked at the possibility of interference with the BMm method in 27 patients selected for M-Ig (multiple myeboma 24, Waldenstr#{246}m macroglobulinemia 2, and MGUS 1) and a total protein concentration >90 g/L. (Some patients were tested more than once.) Four of the 27 patients (15%) showed a positive interference. In addition, for 2 of these 4 patients, the instrument did not indicate or flag an effect on phosphate concentration. In this group, the phosphate results determined for a patient before and after plasmapheresis showed that interference present before plasmapheresis disappeared after plasmapheresis. However, the degree of interference was apparently unrelatedto proteinconcentration, as suggested by the results from the 80-day follow-up of another patient. As for the effects of lipemia,we observed that negative interference, which was important with the BMu method [6], had totallydisappeared with the BMm method. In conclusion, we confirm that the modified phosphate kit recently released by BM greatly improves the reliability of phosphate determination by reducing the appearance of false hyperphosphatemia due to the presence of M-Ig. However, in contrast to recent reports [4, 5], positive interferences in such samples have not totally disappeared. As a consequence, laboratory staff must still be watchful; as Lamer said [7], the phenomenon of M-Ig interference may generally apply to other
Letters
1300
spectrophotometric assays. The kinetic phosphate reaction in patients’ samples from the hematology department, especially those whose protein concentrations are high, should be carefully observed.
that we experienced after one type of tube (LiHep-Gel no. 03.1631.001; Sarstedt,
introducing 4.7 mL, cat. Leicester,
UK).
A hematology outpatient presented with a plasma potassium of 7.2 mmob/L. References The other plasma electrolytes were un1. Landowne RA. Immunoglobulin interference remarkable, as had been the potassium with phosphorus and chloride determination previously. The specimen was not hemowith the Coulter Chemistry [Letterl. Clin Chem lyzed and had stood unrefrigerated for 1979:25:1189-90. 2. Adler SG, Laidlaw SA, Lubran MM, Kopple ID. several hours before separation. The paHypergiobulinemia may spuriously elevate meatient was a 74-year-old man with T-cebl sured serum inorganic phosphate levels. Am I chronic lymphocytic leukemia, whose Kidney Dis 1988:11:260-3. 3. Hawkins RC. Pseudohypenphosphatemia in features were consistent with those of the multiple myeloma. Ann Clin Biochem 1991:28: small-cell variant of T-prolymphocytic 226-8. leukemia. He was subsequently brought 4. Zaman Z, Sneyers L. Van Orshoven A, Blanckaert N, Mani#{234}n G. Elimination of paraprotein in as an emergency for a further blood interference in determination of plasma inortest, when the plasma potassium was 5.1 ganic phosphate by ammonium molybdate mmol/L, but the next day his plasma method. Clin Chem 1995;41:6O9-14. 5. Savory Di, Pearce Ci. Panapnotein interference potassium was 7.5 mmol/L. This latter causing pseudohyperphosphataemia: evaluablood specimen had stood unrefrigerated tion of an improved methodology. Ann Clin for at beast an hour before separation. Biochem 1995:32:498-501. 6. Fleuret C, Steghens IP. Collombel C. Interf#{233}r- When further blood was taken and spun ence des dysglobulinemies et de Ia lacteswithin -10 mm (as in the emergency cence sur le dosage des phosphates s#{233}niques: situation), the problem seemed to reune correction simple et efficace. Ann Biol Clin 1995:53:141-3. solve: plasma potassium was 4.2 mmol/L. 7. Lamer Al. Monoclonal immunoglobulins and As the patient’s white cell count was the measurement of inorganic phosphate in 425.0 X 109/L, we thought it possible serum [Letter]. Ann Clin Biochem 1995:32: 102. that these cells were bysing and causing the high potassium values. All the above blood specimens were Carole Poupon_Fbeuret* Colette Chapuis-Cellier collected into heparin tubes containing F#{233}d#{233}ration de BiochIm. the gel separator. When blood was colH#{244}pital Edouard Herriot lected from this patient into heparin Place d’Arsonval tubes without the gel, plasma potassium 69 437 Lyon Cedex 3, France values of 4.8 and 4.4 mmol/L were reported even after plasma separation had been delayed as described above. SubseAuthor for correspondence. quently when gel tubes were used and not spun directly, plasma potassium values were increased, e.g., 7.1 mmob/L. To check that plasma potassium values Variation in Plasma Potassium When were not normally erroneous when sepUsing Collection Tubes Containing Gel aration of blood collected into gel tubes Separators was delayed, we compared plasma potassium concentrations in blood collected in To the Editor: gel and non-gel tubes. Blood was collected from nine Hematology Clinic paBlood-collecting tubes containing gel tients at a remote site and was subjected separators have been introduced by sevto the normal transportation delay of at eral manufacturers, and evaluations have least an hour before separation. The shown test results, including plasma pomaximum difference in plasma potassium tassium, to give good agreement when results between any pair was 0.3 mmobfL, compared with non-gel tubes [1, 2]. One with a range of potassium values 3.6-4.5 report [3] did show changes in some mmol/L. None had extremely high white analytes when plasma from gel tubes, cell counts (range 3.7-29.5 X 109/L, which had been centrifuged and stored reference range 4.0-11.0 x b09/L). This overnight, was compared with plasma good agreement in plasma potassium valfrom non-gel tubes. The changes in ues from the two types of tubes suggests plasma potassium ranged from 0 to 0.5 that in our patient the problem was mmolfL. We think that laboratory personnel should be aware of a phenomenon caused by a combination of the gel tube,
a high white cell count, and the delay in separation of the blood. This patient was treated with cycles of chborambucib and predriisolone and made satisfactory progress. When his white cell count returned to normal, there was no recurrence of the increased potassium phenomenon. This experience serves as a warning that an unexpectedly high and erroneous plasma potassium may be obtained in a specimen with a very high white cell count and when blood is collected into this type of gel tube. References 1. Chan KM, Daft M, Koenig 1W, Ladenson iH.
Plasma separator tube of Becton Dickinson evaluated [Tech Brief]. Clin Chem 1988:34: 2158-9. 2. Doumas BT, Hause LL, Simuncak DM, Breitenfield D. Differences between values for plasma and serum in tests performed in the Ektachem 700 XR analyzers and evaluation of Plasma Separator Tubes (PST)’. Clin Chem 1989:35: 151-3. 3. Bailey IR. Effect on 16 analytes of overnight storage of specimens collected into hepaninised evacuated tubes with plasma separator. Ann Clin Biochem 1990:27:56-8
Simon Jowitt1 G. Longlands2* Don E. Probe& Depts. of ‘Hematol.and 2Chem. Pathol. North ManchesterGeneralHospital ManchesterM8 SRB, UK Michael
*Author for correspondence.
Discrimination Among Dyshemoglobins:Analytical Approach to a Toxicological Query To the Editor: Several instruments, i.e., CO-oximeters, are now available for the spectrophotometric determination of hemoglobin species, including oxyhemoglobin (O2Hb), reduced hemoglobin (HHb), methemogbobin (MetHb), and carboxyhemoglobin (COHb). Although all instruments share the same overall principle of measurement, they differ in the number of wavelengths and in the algorithms they use to obtain the final results [1]. The instruments that use more wavelengths can discriminate a greater number of hemoglobin compounds [2]. An example of such discrimination is reflected in three commercially available
