Letters - Clinical Chemistry

Download PDF
3 downloads 0 Views 141KB Size Report
Comment
EDTA or between lithium heparin and citrate tubes (P 0.05). P. 0.051 for comparison of concentrations obtained in the EDTA and citrate tubes. Kisspeptin-IR.
Letters

Clinical Chemistry 54:3 615–623 (2008)

Preanalytical Factors Affecting RIA Measurement of Plasma Kisspeptin

tors of the hypothalamo-pituitarygonadal axis. Mice and humans with defective kisspeptin signaling show hypogonadotrophic hypogonadism and impaired sexual development (1, 2 ). Plasma kisspeptin concentrations are ⬍2 pmol/L in men and nonpregnant women. KiSS-1 mRNA is highly expressed in the placenta, and plasma kisspeptin concentrations increase dramatically, to thousands of picomoles per liter, during pregnancy (3 ). In addition, plasma kisspeptin is increased in women

To the Editor: Kisspeptins are peptide products of the KiSS-1 metastasis-suppressor (KISS1) gene and the natural ligands of the G-protein– coupled receptor GPR54. KISS1 was initially investigated as an antimetastasis gene. More recent studies have demonstrated that the kisspeptins are potent stimula-

with gestational trophoblastic tumors, thus raising the possibility of measuring plasma kisspeptin as a novel tumor marker (4 ). Previous studies that have measured plasma kisspeptin in women during pregnancy have found significantly different concentrations of circulating kisspeptin (3, 4 ) that may be attributable to differences in preanalytical variables, such as collection tube type, processing times, and storage conditions. Use of a standardized sample collection method for the mea-

Table 1. Kisspeptin-IR values (pmol/L) measured in plasma and serum samples collected in 4 different tube types from 4 healthy pregnant women.a EDTA Time, h

Subject 1

Subject 2

Subject 3

Subject 4

Mean (SD)

Paired t-test (time vs t ⴝ 0)

0

8200

7700

4600

3900

6100 (2200)

—

1

6600

6100

4900

4200

5500 (1100)

⬎0.05

2

6600

6000

3800

3800

5100 (1400)

⬎0.05

4

4700

1900

4400

3100

3500 (1300)

⬎0.05

Lithium heparin 0

8800

5300

3600

3800

5400 (2400)

—

1

7200

3800

3500

2900

4400 (1900)

⬎0.05

2

8600

3700

1000

3100

4100 (3200)

⬎0.05

4

6500

1200

⬍2

2600

2700 (2800)

⬍0.05

Citrate 0

6000

3000

3100

2600

3700 (1500)

—

1

5400

3400

2300

3700

3700 (1300)

⬎0.05

2

4500

2900

1800

1900

2800 (2800)

⫽ 0.05

4800

1100

700

1100

1900 (1900)

⬍0.05

1500 (600)

—

—

⬍0.05

4

SST 0

2300

800

1500

1400

1

⬍2

⬍2

⬍2

⬍2

2

⬍2

⬍2

⬍2

⬍2

—

⬍0.05

4

⬍2

⬍2

⬍2

⬍2

—

⬍0.05

700

1500

2800

2800 (2200)

—

Serum 0

6500

1

⬍2

⬍2

⬍2

⬍2

—

⬎0.05

2

⬍2

⬍2

⬍2

⬍2

—

⬎0.05

4

⬍2

⬍2

⬍2

⬍2

—

⬎0.05

a

The paired t-test was used to look for significant differences for a given sample type at different times compared to time zero (P values shown in the final column). Paired t-test was used to compare the values at time zero in each tube type to those obtained with lithium heparin tubes (used in our laboratory as a calibrator). At time zero a statistically significant difference was observed between lithium heparin and serum and SST tubes (P ⬍0.05) but not between lithium heparin and EDTA or between lithium heparin and citrate tubes (P ⬎0.05). P ⫽ 0.051 for comparison of concentrations obtained in the EDTA and citrate tubes. Kisspeptin-IR concentrations reported to the nearest 100 pmol/L.

615

Letters surement of circulating kisspeptin immunoreactivity (IR) would facilitate comparisons between separate studies and is necessary if plasma kisspeptin concentrations are to be used as a disease marker. We thus assessed the effects of processing time, anticoagulant type, and repeated freeze-thaw cycles on kisspeptin-IR measurement in plasma samples. To evaluate the effects of collection tube type and processing time on plasma kisspeptin-IR measurement, we recruited 4 healthy pregnant women who gave informed consent [age range 25–37 years, mean (SD) 32.8 (5.3) years; body mass index (BMI) 19 –35.5, 25.4 (7.07); gestation 15–35 weeks, 27 (8.83) weeks]. Blood was collected, using a Vacutainer® system, directly into 4 tubes of each of the tube-types studied: lithium heparin with 2000 U/tube of trasylol (L), citrate (C), EDTA (E), serum clot-activator (S), and serum separator (SST) tubes (BD Vacutainer® Blood Collection Tubes). Both S and SST tubes are coated with silicone and micronized silicon particles to accelerate clotting. The SST tubes have, in addition, a barrier polymer at the bottom of the tube. The density of this polymer allows it to rise up to the clot–serum interface during centrifugation, thus forming a physical barrier separating the serum from the clot. Of the 4 samples in each tube type, the first was processed immediately, and the second, third, and fourth samples were maintained at room temperature (18 – 20 °C) for 1, 2, and 4 h, respectively, before processing. All samples were processed by centrifugation at 4 °C for 10 min at 855g. Serum and plasma were then aspirated, divided into aliquots, frozen, and stored at ⫺80 °C until assayed. Kisspeptin-IR was assayed using a specific RIA (5 ). The detection limit of the assay was 2 pmol/L 616 Clinical Chemistry 54:3 (2008)

of plasma kisspeptin-IR at a 95% confidence limit. The inter- and intraassay CVs were ⬍11% and ⬍9%, respectively. All samples were assayed in duplicate. Differences of means were assessed by paired t-test. Kisspeptin-IR degraded rapidly in serum tubes. Kisspeptin-IR was undetectable in serum samples that were processed at t ⫽ 1 h and later time points, and concentrations detected in serum samples processed immediately were significantly lower than those detected in plasma (P ⬍0.05). Kisspeptin-IR concentrations in plasma samples collected in C tubes were consistently lower than those obtained from E and L collected samples, but this difference was not statistically significant (Table 1). Kisspeptin-IR in plasma samples decreased with increased processing time, suggesting that kisspeptin-IR is best measured in plasma samples that are processed immediately after sampling. To evaluate the effects of repeated freezing and thawing on plasma kisspeptin-IR concentrations, 5 mL of blood from each of 4 volunteers [age range 23– 41 years, mean (SD) 30.5 (7.6) years; BMI 22–28, 25 (2.45); gestation 23–31 weeks, 30 (0.82) weeks] was collected directly into L and E collection tubes. L and E tubes were used, on the basis of results obtained in the first part of the study, which suggested that kisspeptin-IR was most stable in these 2 tube types. The samples were centrifuged and plasma was separated immediately after collection. An aliquot was taken from the plasma before the first cycle of freezing and after 1, 2, and 3 freeze-thaw cycles. Kisspeptin-IR concentrations in samples collected in L or E tubes did not significantly change even after 4 freeze-thaw cycles (data not shown). Our studies suggest that circulating kisspeptin-IR concentra-

tions should be measured in plasma samples processed immediately after collection. L or E tubes may preserve kisspeptin-IR better than C tubes. Repeated freezing and thawing does not significantly influence the measurement of kisspeptin-IR concentrations. Grant/funding Support: KGM is supported by a BBSRC New Investigator Award. MP is supported by the BBSRC. WSD is supported by a UK Department of Health Clinician Scientist Award. The Department of Metabolic Medicine receives funding from the Medical Research Council, Wellcome Trust, EU 6th Framework Programme (LSHM-CT-2003503041) and the NIHR Biomedical Research Centre Funding Scheme. Financial Disclosures: None declared. References 1. Seminara SB, Messager S, Chatzidaki EE, Thresher RR, Acierno JS Jr, Shagoury JK, et al. The GPR54 gene as a regulator of puberty. N Engl J Med 2003;349:1614 –27. 2. Kuohung W, Kaiser UB. GPR54 and KiSS-1: role in the regulation of puberty and reproduction. Rev Endocr Metab Disord 2006;7:257– 63. 3. Horikoshi Y, Matsumoto H, Takatsu Y, Ohtaki T, Kitada C, Usuki S, et al. Dramatic elevation of plasma metastin concentrations in human pregnancy: metastin as a novel placenta-derived hormone in humans. J Clin Endocrinol Metab 2003;88:914 –9. 4. Dhillo WS, Savage P, Murphy KG, Chaudhri OB, Patterson M, Nijher GM, et al. Plasma kisspeptin is raised in patients with gestational trophoblastic neoplasia and falls during treatment. Am J Physiol Endocrinol Metab 2006;291:E878 – 84. 5. Dhillo WS, Chaudhri OB, Patterson M, Thompson EL, Murphy KG, Badman MK, et al. Kisspeptin-54 stimulates the hypothalamic-pituitary gonadal axis in human males. J Clin Endocrinol Metab 2005;90:6609 –15.

Radha Ramachandran1,2 Michael Patterson1 Kevin G. Murphy1 Waljit S. Dhillo1 Sejal Patel1 Anna Kazarian1 Mohammad A. Ghatei1 Stephen R. Bloom1*

Letters 1

Department of Metabolic Medicine Imperial College London Hammersmith Hospital London, UK 2 Division of Clinical Chemistry Hammersmith Hospitals NHS Trust London, UK

* Address correspondence to this author at: Department of Metabolic Medicine Imperial College London 6th Floor Commonwealth Building Hammersmith Hospital Du Cane Road London, W12 0NN, UK Fax ⫹44 (0) 20 8383 3142 e-mail [email protected] DOI: 10.1373/clinchem.2007.093005

Relationship of MRIDetermined Infarct Size and cTnI Measurements in Patients with ST-Elevation Myocardial Infarction To the Editor: The extent of myocardial infarction (MI) is related to patient outcomes (1 ), and clinicians often wish to have a sense for this critical measure. Imaging methods, although accurate, have limited accessibility and high cost. Thus clinicians often use biomarkers to provide such an estimate. Measurement of cardiac troponin T (cTnT) at 96 h after onset of MI was observed to correlate with MRI-determined infarct size in both ST-elevation MI (STEMI) and non-STEMI (2 ). We tested the hypothesis that cardiac troponin I (cTnI), another myocardiumspecific biomarker, would provide an equivalent estimation in the subset of STEMI patients previously described (2 ).

The 28 patients with STEMI had sufficient sample remaining to allow for determination of cTnI (2 ). The characteristics of this group have previously been reported (2 ), but in brief the mean age (SD) was 56 (11) years, and 17.4% of patients were women. Mean (SD) body mass index was 26.46 (3.5) kg/m2, and 71.4% of patients were hypertensive, 64.3% were current smokers, 60.7% had hypercholesterolemia, and 14.3% had diabetes. Blood samples were available at admission and at 24, 48, 72, and 96 h after onset of symptoms. cTnI concentrations were measured at a laboratory in Heidelberg, Germany, with the AccuTnI assay (Beckman-Coulter). The assay has a limit of detection of 0.01 ␮g/L, with a 99th percentile as low as 0.02– 0.03 ␮g/L. In the laboratory performing the measurements, a cutoff value of 0.03 ␮g/L was used. Cardiac MRI was performed as described elsewhere (2 ). Plasma concentrations of cTnI are reported as median with the corresponding interquartile range (IQR). For all analyses, a value of P ⬍0.05 was considered statistically significant. Correlation coefficients were calculated by the Spearman test. Of the 28 study patients with STEMI, 7.1% of patients (2 of 28) received fibrinolytic agents before percutaneous coronary intervention; the remainder underwent primary percutaneous coronary intervention. Preinterventional thrombolysis in myocardial infarction flow grade 3 was present in 12 of 28 patients (42.3%) before PCI and in 27 of 28 patients (96.4%) after. The median time delay from onset of symptoms to balloon angioplasty was 6.25 h. MRI was performed at median of 4 days (range 3– 4 days). All patients manifested delayed hyperenhancement observed with MRI; in 60.71% of patients (17 of 28) hyperenhancement was transmural. Mean infarct size relative to heart weight was

18.2% (IQR 7%– 49%). Ventricular function evaluated by MRI revealed a mean ejection fraction of 54.4% (27.9%– 63.6%) and a mean stroke volume of 91.32 mL (42.4 – 109.3 mL). The median absolute value for infarct size was 29.3 g (IQR 16.6 –53.0 g). Spearman analysis demonstrated a strong correlation between cTnI values and infarct mass at 24 (n ⫽ 24), 48 (n ⫽ 26), 72 (n ⫽ 23), and 96 h (n ⫽ 28) after onset of symptoms (Fig. 1A and B). As with other studies (2, 3 ), correlations between the infarct size and cTnI were significant for all time-points except for admission (Spearman correlation coefficient Rho ⫽ 0.2; data not shown). These data document that cTnI values provide accurate estimates of infarct size in patients with STEMI. As with cTnT (2, 3 ), cTnI correlated with infarct size in reperfused STEMI patients at 24 –96 h as well (Fig. 1A). These data indicate that clinicians can rely on values on days 1– 4 to provide an approximation to MRIdetermined reperfused infarct size. Importantly, the slopes of the correlation curves are different for each day (Fig. 1B). The parallelism of these data with the data for cTnT suggest that similar relationships, as with cTnT, are also likely to occur with non-STEMI. It is also likely that there will be differences with and without reperfusion. The use of imaging techniques such as positron-emission tomography and contrast-enhanced MRI (4 ) to enable determination of infarct size are currently impeded owing to limited availability and high cost. Small infarcts may escape visualization because of inadequate resolution. We now show with MRI that single-point values of cTnI between 24 and 96h, as with sestamibi measurements (5 ), correlate well. Troponin measurements are not only cheaper and more available but provide an estiClinical Chemistry 54:3 (2008) 617

Letters

Fig. 1. Scatter plot of cTnI values (␮g/L) and infarct size (percent of the total heart mass) at 72 h after onset of symptoms (A) and correlation of the infarct size and cTnI values (B). Correlations of the infarct mass (g) and cTnI values (␮g/L) at 24, 48, 72, and 96 h and peak value after onset of symptoms are shown. The slope coefficients (Rho) are indicated for each time point (B). For each time point shown, P value ⬍0.001.

mate of the size of infarction not confounded by prior infarction. This investigation shows that for patients with reperfused STEMI, early as well as later measurements of cTnI are reliable estimates of infarct size. The fact that cTnI measurement at time-points earlier than 96 h correlate well with MRIdetermined infarct size should allow for an earlier evaluation of prognosis in of these patients. Larger prospective and controlled studies are needed to confirm our results. Grant/funding Support: None declared. Financial Disclosures: E.G. has received financial support for studies from Roche Diagnostics and is consultant to Roche. H.A.K. developed the assay for cTnT and holds a patent jointly 618 Clinical Chemistry 54:3 (2008)

with Roche Diagnostics. He has received research support from Roche Diagnostics. A.S.J. is a consultant and has received research support from DadeBehring and Beckman-Coulter. He is a consultant to Ortho Diagnostics and has consulted over time for most of the major diagnostic companies.

measurements of cardiac troponin T. J Am Coll Cardiol 2008;51:307–14. 4. Gibbons RJ, Valeti US, Araoz PA, Jaffe AS. The quantification of infarct size. J Am Coll Cardiol 2004;44:1533– 42. 5. Panteghini M, Cuccia C, Bonetti G, Giubbini R, Pagani F, Bonini E. Single-point cardiac troponin T at coronary care unit discharge after myocardial infarction correlates with infarct size and ejection fraction. Clin Chem 2002;48:1432– 6.

Vlad C. Vasile1 Luciano Babuin1 Evangelos Giannitsis2 Hugo A. Katus2 Allan S. Jaffe1*

References 1. Walsh MN, Bergmann SR, Steele RL, Kenzora JL, Ter-Pogossian MM, Sobel BE, Geltman EM. Delineation of impaired regional myocardial perfusion by positron emission tomography with H2(15)O. Circulation 1988;78:612–20. 2. Steen H, Giannitsis E, Futterer S, Merten C, Juenger C, Katus HA. Cardiac troponin T at 96 hours after acute myocardial infarction correlates with infarct size and cardiac function. J Am Coll Cardiol 2006;48:2192– 4. 3. Giannitsis E, Steen H, Kurz K, Ivandic B, Simon AC, Futterer S, et al. Cardiac magnetic resonance imaging study for quantification of infarct size comparing directly serial versus single time-point

1

Departments of Internal Medicine Division of Cardiovascular Diseases and Laboratory Medicine and Pathology Mayo Clinic and Mayo Medical School Rochester, MN 2 Abteilung Innere Medizin III Medizinische Klinik Universita¨tsklinikum Heidelberg Heidelberg, Germany

Letters

* Address correspondence to this author at: Mayo Clinic 200 First St SW Division of Cardiovascular Diseases Gonda 5 Rochester, MN 55905 Fax 507-266-0228 e-mail [email protected] DOI: 10.1373/clinchem.2007.095604

Cross-Reactivity of BNP, NT-proBNP, and proBNP in Commercial BNP and NT-proBNP Assays: Preliminary Observations from the IFCC Committee for Standardization of Markers of Cardiac Damage To the Editor: B-type natriuretic peptide (BNP) is a 32 amino acid cardiacsynthesized hormone that reduces blood pressure and increases sodium excretion (1 ). Following proteolytic cleavage of proBNP, a 108-amino acid precursor, an Nterminal fragment (NT-proBNP) and BNP are released (2 ). Increased concentrations of BNP and NT-proBNP can be used clinically to monitor heart failure, but a lack of alignment between commercial BNP and NT-proBNP assays (3 ) can lead to confusion when clinicians or laboratorians compare results measured for the same analyte on different instruments. Some of this confusion arises from variable assay specificity regarding what peptides are being measured. We studied whether (a) BNP assays demonstrated crossreactivity with NT-proBNP or proBNP, and (b) whether NTproBNP assays demonstrated crossreactivity with BNP or proBNP, by using 5 commercial BNP and 3 commercial NT-proBNP assays

with 2 BNP, 2 NT-proBNP, and 2 proBNP materials. The NPs studied were: Peptide Institute synthetic BNP (aa 77–108), Scios human recombinant BNP (aa 77–108), HyTest human recombinant NT-proBNP (aa 1–76), Roche modified (amidated for stabilization) synthetic NTproBNP, HyTest human recombinant proBNP (aa 1–108), and Scios glycosylated human recombinant proBNP. BNP assays evaluated were Abbott Architect, Abbott AxSYM, Bayer Centaur, Biosite Triage, and Beckman Access (Biosite assay packaged for use by Beckman). NT-proBNP assays (all based on Roche reagents) were Dade-Behring Dimension, OrthoClinical Diagnostics Vitros, and Roche Elecsys 2010. All assays, for which epitopes of the antibodies used have been previous described (3 ), were run according to the manufacturers’ guidelines. BNP, NT-proBNP, and proBNP materials were diluted with normal (low NP concentration) EDTAplasma pools and lithium-heparin plasma (Dade assay only) pools, collected from healthy donors after institutional review board approval was obtained, to achieve target concentrations of 250, 500, and 1000 ng/L. Baseline BNP and NT-proBNP were quantified first in the pools and then after the pools were spiked with NP peptides. All measurements were performed in duplicate. Baseline BNP or NT-proBNP concentrations were subtracted from each spiked pool measurement. Percent crossreactivity was calculated by dividing the measured concentration for the spiked pool into the expected peptide concentration, multiplying by 100, and then averaging across all 3 expected concentrations. Recoveries and cross-reactivity percentages between peptides and BNP and NT-proBNP assays are displayed in Table 1. The BNP assays were more specific for the

BNP peptides, with recovery ranging from 79% to 199%, compared to 5% to 38% cross-reactivity to the proBNP peptides and ⬍1% to 7% cross-reactivity to the NTproBNP peptides. Similarly, the NT-proBNP assays were more specific for the NT-proBNP peptides, showing 47% to 243% recovery, with substantial crossreactivity to proBNP peptides (⬍1% to 249%), and no crossreactivity to the BNP peptides (⬍1% across all assays). This study demonstrates that the BNP peptides used do not substantially cross-react with NTproBNP assays, and that the NTproBNP peptides do not substantially cross-react with BNP assays. We confirm that there is substantial cross-reactivity between proBNP peptides and commercially available BNP and NT-proBNP assays. Variations depended on the different sources and types of peptide tested in each assay. We observed minimal crossreactivity with the glycosylated Scios proBNP peptide, compared with substantial cross-reactivity to the nonglycosylated HyTest proBNP peptide with the NTproBNP assays. Glycosylation likely interfered with peptide antibody binding. The mechanisms responsible for different reactivities between the HyTest and Roche NTproBNP peptides using different NT-proBNP assays, which use the same antibodies but different assay architectures, cannot be explained presently. The modest differences in reactivities for the recombinant (Scios) and synthetic (Peptide Institute) BNP materials using different BNP assays also requires additional study; with different assay architectures for the same reagents (Biosite, Beckman) showing diverse recovery. Little is known about which NP forms are circulating physiologically. The clinical significance of measured cross-reactivities will Clinical Chemistry 54:3 (2008) 619

Letters

Table 1. Percentage recoveries and cross-reactivities by BNP and NT-proBNP assays for each peptide. Assay

Sa BNP

P BNP

S proBNP

H proBNP

H NT-proBNP

R NT-proBNP

Architect

151 (142, 160)

98 (85, 110)

38 (37, 38)

6 (2, 11)

⬍1 (0, 0.8)

4 (4, 5)

AxSYM

184 (164, 205)

124 (117, 130)

34 (28, 39)

9 (3, 15)

⬍1 (0, 0.5)

4 (3, 5)

Centaur

194 (189, 199)

137 (133, 141)

17 (17, 18)

14 (10, 17)

⬍1 (0, 0.3)

7 (5, 9)

Access

199 (192, 205)

130 (129, 130)

24 (24, 25)

13 (7, 19)

⬍1 (0, 0.4)

6 (5, 7)

Triage

135 (115, 156)

79 (78, 80)

19 (18, 20)

5 (0, 12)

⬍1 (0, 0.2)

3 (2, 4)

b

NT-proBNP

a b

Dimension

⬍1 (0, 0)

⬍1 (0, 0)

249 (230, 267)

243 (235, 251)

95 (91, 99)

Vitros

⬍1 (0.6, 0.8)

⬍1 (0.03, 0.4)

⬍1 (0, 0) 2 (1.6, 2.2)

55 (52, 59)

127 (124, 130)

71 (68, 74)

Elecsys

⬍1 (0.2, 0.8)

⬍1 (0, 0.04)

1 (0.7, 2)

29 (28, 30)

131 (126, 137)

47 (45, 49)

S indicates Scios; P, Peptide Institute; H, HyTest; R, Roche. 95th Percentile confidence intervals are in parentheses.

also depend on the relative quantities of different molecular forms found in individual patient blood. Mass spectrometry has shown that mature BNP1⫺32 is absent in severe heart failure patients, but a commercial assay still detected the presence of BNP1⫺32 (4 ). Glycosylated forms of proBNP in heart failure patients have also been found, along with uncleaved proBNP (1, 5 ). Expression of the multiple forms that circulate during acute and chronic heart failure must be investigated in future studies to ensure that commercial assays are measuring the appropriate diagnostic biomarkers for heart failure. Our preliminary observations challenge the analytical field to better characterize what immunoassays measure and challenge the clinical field to better understand what molecules are measured and how to best interpret both BNP and NT-proBNP findings in regard to patient care. Grant /funding Support: Support for this project was from the IFCC Committee for Standardization of Markers of Cardiac Damage through generous donations by the in vitro diagnostic manufacturers of BNP and NT-proBNP assays. 620 Clinical Chemistry 54:3 (2008)

Financial Disclosures: FSA has consulted for Abbott and Ortho Clinical Diagnostics; served on advisory boards for Ortho, Biosite, and Beckman Coulter; and received research grant funding from Abbott, Beckman, bioMeriuex, Biosite, Dade-Behring, MKI, Ortho Clinical Diagnostics, Response Biomedical, Siemens, and Roche. ASJ has received research support and consultation from Dade-Behring, Beckman Coulter, and Ortho Clinical Diagnostics and consulting from Critical Diagnostics, Singulex, Intermune, and Celladon. JO-L has received research funding from Roche. He has received honoraria for speaking from Roche and bioMerieux. RHC has received research funding from Biosite, Dade Behring, Response Biomedical, and Roche; received honoraria from Biosite, Inverness Medical Solutions, and Response Biomedical; and served on advisory boards for Biosite, Dade Behring, Inverness Innovations, and Response Biomedical. JM has received research funding and honoraria from Roche, Abbott, and Siemens. AHBW, KNL, FP, and JT have no declarations. Acknowledgments: We thank the Peptide Institute, Scios, HyTest,

and Roche for their generous donation of peptides. References 1. Schellenberger U, O’Rear J, Guzzetta A, Jue RA, Protter AA, Pollitt NS. The precursor to B-type natriuretic peptide is an O-linked glycoprotein. Arch Biochem Biophys 2006;451:160 – 6. 2. Liang F, O’Rear J, Schellenberger U, Tai L, Lasecki M, Schreiner GF, et al. Evidence for functional heterogeneity of circulating BNP. J Am Coll Cardiol 2007;49:1071– 8. 3. Apple F, Panteghini M, Ravkilde J, Mair J, Wu AH, Tate J, et al. Quality specifications for B-type natriuretic peptide assays. Clin Chem 2005;51: 486 –93. 4. Hawkridge AM, Heublein DM, Bergen HR 3rd, Cataliotti A, Burnett JC Jr, Muddiman DC. Quantitative mass spectral evidence for the absence of circulating brain natriuretic peptide (BNP-32) in severe human heart failure. Proc Natl Acad Sci 2005;102:17442–7. 5. Giuliani I, Rieunier F, Larue C, Delagneau JF, Granier C, Pau B, et al. Assay for measurement of intact B-type natriuretic peptide prohormone in blood. Clin Chem 2006;52:1054 – 61.

Kristin N. Luckenbill1 Robert H. Christenson2 Allan S. Jaffe3 Johannes Mair4 Jordi Ordonez-Llanos5 Franca Pagani6 Jillian Tate7 Alan H. B. Wu8 Ranka Ler1 Fred S. Apple1*

Letters 1

Hennepin County Medical Center University of Minnesota School of Medicine Minneapolis, MN 2 University of Maryland School of Medicine Baltimore, MD 3 Mayo Clinic and Mayo College of Medicine Rochester, MN 4 Department of Internal Medicine Clinical Division of Cardiology Innsbruck Medical University Austria 5 Department of Biochemistry Hospital de la Santa Creu i Sant Pau and Universitat Autonoma Barcelona, Spain 6 Laboratorio Analisi Chimico Cliniche Dipartimento di Medicina di Laboratorio Azienda Ospedaliera ‘Spedali Civili’ Brescia, Italy 7 Chemical Pathology Department Pathology Queensland Royal Brisbane and Women’s Hospital Brisbane, Australia 8 University of California at San Francisco San Francisco General Hospital San Francisco, CA * Address correspondence to this author at: Hennepin County Medical Center Clinical Laboratories P4 701 Park Ave Minneapolis, MN 55415 Fax 612-904-4229 e-mail [email protected] DOI: 10.1373/clinchem.2007.097998

The ⴚ590CC Genotype in the IL4 Gene as a Strong Predictive Factor for the Development of Hypothyroidism in Hashimoto Disease To the Editor: The severity of Hashimoto disease (HD) varies among patients and is difficult to predict when the dis-

ease is in the subclinical state and diagnosed by the presence of thyroid-specific autoantibody. Likewise, the intractability of Graves disease (GD) is difficult to predict. Autoimmune thyroid destruction that underlies both diseases is strongly determined by T-cell cytotoxicity, which is activated by interferon (IFN)-␥ (1 ), and the T allele in ⫹874A/T polymorphism of the interferon gamma (IFNG) gene, which promotes increased IFN-␥ production, has been noted more frequently among patients with severe HD (2 ). Cytokine balance between T-helper 1 (Th1) cytokines, such as IFN-␥, and Th2 cytokines is important in immune regulation (3 ). Therefore, it is possible that Th2 cytokines may also affect the severity of HD and the intractability of GD. Interleukin (IL)-4, one of the key Th2 cytokines, stimulates humoral immunity and suppresses the production of inflammatory Th1 cytokines, including IFN-␥ (3 ). Individuals who carry the T allele in ⫺590C/T polymorphism of the interleukin 4 (IL4) gene (rs2243250) have a higher proportion of IL-4 –producing T-helper cells (4 ). Individuals with the Ile/Ile genotype in the Ile50Val polymorphism (rs1805010) of the interleukin 4 receptor, alpha (IL4RA) gene also have higher IL-4R␣ activity compared to individuals with the Ile/Val or Val/Val genotype (5 ). To clarify the association of these polymorphisms of the IL4 and IL4RA genes with HD severity and GD intractability, we performed genotyping using a direct sequencing reaction for ⫺590C/T and TaqMan威 single-nucleotide polymorphism genotyping assays for Ile50Val in the following individuals: 35 patients ⬍50 years old with HD who developed permanent hypothyroidism and were treated with L-thyroxine (rapid destruction type; severe HD), 39 euthyroid patients ⬎50 years old with HD who were untreated (slow destruction type; mild HD),

50 euthyroid patients with GD who remained positive for antithyrotropin receptor antibody (TRAb) despite being on antithyroid drug treatment for ⬎5 years (intractable GD), 26 patients with GD in remission who had been euthyroid and negative for TRAb for ⬎2 years after cessation of antithyroid drug treatment, and 47 controls. All patients with HD were positive for antithyroid microsomal antibodies. All patients with GD showed increased TRAb concentrations and thyrotoxicosis when first diagnosed. All study participants were Japanese and unrelated. Written informed consent was obtained from all patients and controls, and the study protocol was approved by the Ethics Committee of Osaka University. The frequency of the ⫺590CC genotype in the IL4 gene among the patients with severe HD was 20.0%, whereas the frequency among the patients with mild HD was 0.0% (P ⫽ 0.0037) (Table 1). If this finding is confirmed prospectively, a high percentage of HD patients with the IL-4 ⫺590CC genotype would be expected to develop severe disease (hypothyroidism) before 50 years of age. A possible mechanism to explain this phenomenon is that in HD patients with the ⫺590CC genotype, the proportion of IL-4 –producing T-helper cells is lower than that in HD patients with the ⫺590TT or ⫺590CT genotype (4 ), which results in higher activity of inflammatory Th1 cytokines and more rapid progression of thyroid destruction, followed by early development of hypothyroidism. Conversely, we found no association between the Ile50Val polymorphism in the IL4RA gene and HD severity (Table 1) despite the receptor activity being different among individuals with this polymorphism. Furthermore, we found no association between these IL4 and IL4RA gene polymorphisms and GD intractability (Table 1); however, GD is caused by TRAb and tends to be more intractable in paClinical Chemistry 54:3 (2008) 621

Letters Table 1. Genotype frequencies of IL-4 ⴚ590C/T and IL-4R␣ Ile50Val polymorphisms in patients with autoimmune thyroid disease and in controls. Hashimoto disease Severea (n ⴝ 35)

Mildb (n ⴝ 39)

Graves disease Intractablec (n ⴝ 50)

P

In remissiond (n ⴝ 26)

P

Controls (n ⴝ 47)

IL-4 ⫺590C/T CC

7 (20.0%)

0 (0.0%)

CT

13 (37.1%)

17 (43.6%)

TT

15 (42.9%)

22 (56.4%)

7 (20.0%)

0 (0.0%)

28 (80.0%)

39 (100.0%)

IleIle

4 (11.4%)

4 (10.3%)

IleVal

21 (60.0%)

ValVal

4 (8.0%) e

0.0034

2 (7.7%)

3 (6.4%) e,f

22 (44.0%)

11 (42.3%)

24 (48.0%)

13 (50.0%)

NS

19 (40.4%)

4 (8.0%)

2 (7.7%)

46 (92.0%)

24 (92.3%)

7 (14.0%)

4 (15.4%)

24 (61.5%)

23 (46.0%)

13 (50.0%)

26 (55.3%)

10 (28.6%)

11 (28.2%)

20 (40.0%)

9 (34.6%)

15 (31.9%)

4 (11.4%)

3 (10.3%)

7 (14.0%)

4 (15.4%)

31 (88.6%)

32 (89.7%)

43 (86.0%)

22 (84.6%)

25 (53.2%)

CC vs T carriage CC CT ⫹ TT

0.0037g

NSg

3 (6.4%) 44 (93.6%)

IL-4R␣ Ile50Val NSe

NSe

6 (12.8%)

IleIle vs Val carriage IleIle IleVal ⫹ ValVal

NSg

NSg

6 (12.8%) 41 (87.2%)

Clinical characteristics at sample collectionh Age, y Female/male, n

52.1 (10.0) 30/5

56.2 (6.6) 30/9

50.2 (16.1) 41/9

49.1 (14.1)

47.6 (12.2)

24/2

40/7

Free T4 , ng/L

14.1 (3.5)

11.9 (1.4)

14.1 (9.8)

13.3 (4.6)

ND

Free T3, ng/L

2.43 (0.54)

2.90 (0.31)

3.14 (2.81)

2.96 (1.49)

ND

TSH, mU/L

3.18 (1.13)

2.21 (1.72)

2.79 (0.77)

1.97 (1.32)

ND

⬍1.0

⬍1.0

McAb, 2n ⫻ 100

5.9 (2.7)

2.0 (2.4)

4.6 (2.6)

5.6 (2.1)

TgAb, 2n ⫻ 100

7.7 (3.1)

1.7 (2.7)

2.7 (2.8)

1.6 (1.5)

Negative

Goiter size, cm

3.7 (1.7)

4.9 (1.5)

4.5 (1.0)

4.2 (0.5)

Impalpable

Treatment

L-thyroxine

TRAb, IU/L

None

5.63 (11.36)

Thiamazole or PTU

⬍1.0

ND Negative

None

None

Hashimoto disease patients ⬍50 years old who developed hypothyroidism and were treated with L-thyroxine. Untreated and euthyroid Hashimoto disease patients ⬎50 years old. c Graves disease patients who remained positive for TRAb despite being treated by antithyroid drug for more than 5 years. d Graves disease patients in remission who have been euthyroid and negative for TRAb for more than 2 years after cessation of anti-thyroid drug treatment. e Analyzed by ␹2 test. f NS, not significant; ND, not determined; T4, thyroxine; T3, triiodothyronine; TSH, thyrotropin; TRAb, antithyrotropin receptor antibody; McAb, antithyroid microsomal antibody; TgAb, antithyroglobulin antibody; PTU, Propylthiouracil. g Analyzed by Fisher exact test. h Results of clinical characteristics are expressed as mean (SD). a

b

tients with higher TRAb concentrations. In this context, we also found no relationship between these polymorphisms and serum TRAb concentrations (data not shown). In conclusion, the ⫺590CC genotype in the IL4 gene appears to be 622 Clinical Chemistry 54:3 (2008)

a strong predictive factor for the development of hypothyroidism in HD. Grant/funding Support: This work was supported by KAKENHI (17390164).

Financial declared.

Disclosures:

None

References 1. Watanabe M, Yamamoto N, Maruoka H, Tamai H, Matsuzuka F, Miyauch A, Iwatani Y. Inde-

Letters

2.

3.

4.

5.

pendent involvement of CD8⫹CD25⫹ cells and thyroid autoantibodies in disease severity of Hashimoto’s disease. Thyroid 2002;12:801– 8. Ito C, Watanabe M, Okuda N, Watanabe C, Iwatani Y. Association between the severity of Hashimoto’s disease and the functional ⫹874A/T polymorphism in the interferon-␥ gene. Endocr J 2006;53:473– 8. Iwatani Y, Watanabe M. Normal mechanisms for self-tolerance. In: Volpe R, editor. Autoimmune Endocrinopathies. Totawa: Humana Press; 1999. p 1–30. Nakashima H, Miyake K, Inoue Y, Shimizu S, Akahoshi M, Tanaka Y, et al. Association between IL-4 genotype and IL-4 production in the Japanese population. Genes Immun 2002;3: 107–9. Mitsuyasu H, Izuhara K, Mao XQ, Gao PS, Arinobu Y, Enomoto T, et al. Ile50Val variant of

IL4R␣ upregulates IgE synthesis and associates with atopic asthma. Nat Genet 1998;19:119 –20.

Kyoto University School of Medicine Kyoto, Japan

Takashi Nanba1 Mikio Watanabe1 Takashi Akamizu2 Yoshinori Iwatani1*

* Address correspondence to this author at: Department of Biomedical Informatics Division of Health Sciences Osaka University Graduate School of Medicine 1-7 Yamadaoka, Suita Osaka 565-0871, Japan Fax 81-6-6879-2592 e-mail [email protected]

1

Department of Biomedical Informatics Division of Health Sciences Osaka University Graduate School of Medicine Osaka, Japan 2 Translational Research Center Kyoto University Hospital

DOI: 10.1373/clinchem.2007.099739

Clinical Chemistry 54:3 (2008) 623