Exp Clin Endocrinol Diabetes 2015; 123 - LB_11
DOI: 10.1055/s-0035-1549077

Inter-laboratory comparison of IGF-I concentrations measured by an automated immunoassay: Results from a multicentre study across Europe

P Grimminger 1, J Frystyk 2, O Blankenstein 3, BP Hauffa 4, G Johansson 5, AC Muller Kobold 6, J Kratzsch 7, E Cavalier 8, A Piazza 9, C Wüster 10, P Monaghan 11, M Droste 12, YB de Rijke 13, M Bidlingmaier 1
  • 1Medizinische Klinik und Poliklinik IV, Klinikum der Universität München, Munich, Germany
  • 2Medical Research Laboratory, Department of Clinical Medicine, Faculty of Health, Aarhus University, Aarhus, Denmark and Department of Endocrinology and Internal Medicine, Aarhus University Hospital, Aarhus, Denmark
  • 3Oliver Blankenstein, Labor Berlin – Charité Vivantes GmbH, Fachbereich Endokrinologie & Stoffwechsel, Berlin, Germany
  • 4Abteilung für pädiatrische Endokrinologie und Diabetologie, Zentrum für Kinderheilkunde und Jugendmedizin, Klinik für Kinderheilkunde II – Universität Duisburg-Essen, Essen, Germany
  • 5Gudmundur Johannsson, Department of Endocrinology, Sahlgrenska University Hospital and Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
  • 6Dept. of Laboratory Medicine, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands
  • 7Institut für Laboratoriumsmedizin, Klinische Chemie und Molekulare Diagnostik, Universitätsklinikum Leipzig, Leipzig, Germany
  • 8Department of Clinical Chemistry, University of Liège, CHU de Liège, Liège, Belgium
  • 9Immunodiagnostic Systems S.A., Liège, Belgium
  • 10Hormon-&Stoffwechselzentrum, MED Facharztzentrum, Mainz, Germany
  • 11The Christie Pathology Partnership, The Christie Hospital, Manchester, UK
  • 12Praxis für Endokrinologie und Diabetologie – Oldenburg, Oldenburg, Germany
  • 13Depts. of Internal Medicine and Clinical Chemistry, Erasmus MC, Rotterdam, The Netherlands

Introduction: Measurement of Insulin-like Growth-factor I (IGF-I) is key in diagnosis and monitoring of Growth Hormone (GH) related diseases. Considerable variability between results from different assays, but also between different laboratories using the same assay was a major concern in the past. Recently, we reported age- and sex-specific reference intervals for IGF-I measured by a new automated assay (IDS-iSYS). We now evaluated the between laboratory agreement of measurements of IGF-I using this assay in laboratories across Europe.

Methods: 52 samples (serum; normal n = 20, GH deficiency/GH excess n = 20, QC samples n = 12) were aliquoted, distributed to participating laboratories (n = 13) and measured using one batch of the IDS-iSYS IGF-I assay. Results were reported in mass units (ng/mL). SD scores based on the reference intervals were calculated by the LMS formula in 6 labs.

Results: IGF-I concentrations ranged from 13 to 1056 ng/mL. Bland Altman plots show a bias between individual laboratories results (ILR) and the all participants mean (APM) between -3.44% and +4.74% (mean -0.10%) for serum samples and between -7.98% and +6.18% (mean -0.25%) for QC samples. 95% of all results (concentrations) were within +/-8.6% of the APM. The respective slopes for the correlation between ILR and APM in Passing Bablok analysis for concentrations range from 0.963 to 1.045 (serum samples) and 0.961 to 1.049 (QC samples). Corresponding SD scores ranged from -4.18 to +6.79 and were highly correlated between labs (0.974 to 1.037).

Conclusion: Our data demonstrate very good agreement of IGF-I concentrations measured by the IDS-iSYS assay in laboratories across Europe, especially in real patients' samples. SD scores calculated by the same mathematical method also exhibit strong agreement between laboratories. Harmonisation of assay reagents and statistical methods to calculate SD scores is key to reduce between laboratory differences in determination and interpretation of IGF-I data.