Órlaith Burke1, Samantha Benton2, Pawel Szafranski3, Peter von Dadelszen2, S Catalin Buhimschi4, Irene Cetin5, Lucy Chappell6, Francesc Figueras7, Alberto Galindo8, Ignacio Herraiz8, Claudia Holzman9, Carl Hubel10, Ulla Knudsen11, Camilla Kronborg12, Hannele Laivuori13, Olav Lapaire14, Thomas McElrath15, Manfred Moertl16, Jenny Myers17, Roberta B Ness18, Leandro Oliveira19, Gayle Olson20, Lucilla Poston6, Carrie Ris-Stalpers21, James M Roberts10, Sarah Schalekamp-Timmermans22, Dietmar Schlembach23, Eric Steegers22, Holger Stepan24, Vassilis Tsatsaris25, Joris A van der Post21, Stefan Verlohren26, Pia M Villa27, David Williams28, Harald Zeisler29, Christopher W G Redman3, Anne Cathrine Staff30. 1. Nuffield Department of Population Health, University of Oxford, Oxford, UK. Electronic address: orlaith.burke@ndph.ox.ac.uk. 2. Department of Obstetrics and Gynaecology and Child and Family Research Institute, University of British Columbia, Vancouver, British Columbia, Canada. 3. Nuffield Department of Obstetrics and Gynaecology, University of Oxford, Oxford, UK. 4. Yale University, New Haven, CT, USA. 5. University of Milan, Italy. 6. Women's Health Academic Centre, King's College London, London, UK. 7. Hospital Clinic, University of Barcelona, Barcelona, Spain. 8. Fetal Medicine Unit, Dept of Obstetrics and Gynecology, Hospital Universitario 12 de Octubre and Universidad Complutense, Madrid, Spain. 9. Michigan State University, East Lansing, MI, USA. 10. Magee-Womens Research Institute, University of Pittsburgh, Pittsburgh, USA. 11. Dept. of Obstetrics and Gynecology, Aarhus University Hospital and Aarhus University, Aarhus, Denmark. 12. Dept. of Oncology, Aarhus University Hospital and Aarhus University, Aarhus, Denmark. 13. Medical Genetics, Obstetrics and Gynaecology and Institute for Molecular Medicine Finland, University of Helsinki and Helsinki University Hospital, Helsinki, Finland. 14. University of Basel, Basel, Switzerland. 15. Brigham & Women's Hospital, Harvard Medical School, Boston, MA, USA. 16. Interdisciplinary Centre for Gynecology, Gyneco-oncology and Feto-Maternal Medicine, Klagenfurt & Medical University, Graz, Austria. 17. Maternal and Fetal Health Research Centre, St. Mary's Hospital and University of Manchester, Manchester, UK. 18. University of Texas, School of Public Health, Houston, TX, USA. 19. Obstetrics Department, Universidade Estadual Paulista, Botucatu, Brazil. 20. University of Texas Medical Branch, Galveston, TX, USA. 21. Department of Obstetrics and Gynaecology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands. 22. Erasmus Medical Centre, Rotterdam, The Netherlands. 23. University of Jena, Germany. 24. University of Leipzig, Germany. 25. Université Paris Descartes, Paris, France. 26. Charité University Medicine, Berlin, Germany. 27. Obstetrics and Gynecology, University of Helsinki and Helsinki University Hospital, Helsinki, Finland. 28. Institute for Women's Health, NIHR University College London Hospitals Biomedical Research Centre, London, UK. 29. Department of Obstetrics and Gynecology, Medical University Vienna, Austria. 30. Department of Obstetrics and Gynecology, Oslo University Hospital and University of Oslo, Oslo, Norway.
Abstract
BACKGROUND: A common challenge in medicine, exemplified in the analysis of biomarker data, is that large studies are needed for sufficient statistical power. Often, this may only be achievable by aggregating multiple cohorts. However, different studies may use disparate platforms for laboratory analysis, which can hinder merging. METHODS: Using circulating placental growth factor (PlGF), a potential biomarker for hypertensive disorders of pregnancy (HDP) such as preeclampsia, as an example, we investigated how such issues can be overcome by inter-platform standardization and merging algorithms. We studied 16,462 pregnancies from 22 study cohorts. PlGF measurements (gestational age ⩾20 weeks) analyzed on one of four platforms: R&D Systems, AlereTriage, RocheElecsys or AbbottArchitect, were available for 13,429 women. Two merging algorithms, using Z-Score and Multiple of Median transformations, were applied. RESULTS: Best reference curves (BRC), based on merged, transformed PlGF measurements in uncomplicated pregnancy across six gestational age groups, were estimated. Identification of HDP by these PlGF-BRCs was compared to that of platform-specific curves. CONCLUSIONS: We demonstrate the feasibility of merging PlGF concentrations from different analytical platforms. Overall BRC identification of HDP performed at least as well as platform-specific curves. Our method can be extended to any set of biomarkers obtained from different laboratory platforms in any field. Merged biomarker data from multiple studies will improve statistical power and enlarge our understanding of the pathophysiology and management of medical syndromes.
BACKGROUND: A common challenge in medicine, exemplified in the analysis of biomarker data, is that large studies are needed for sufficient statistical power. Often, this may only be achievable by aggregating multiple cohorts. However, different studies may use disparate platforms for laboratory analysis, which can hinder merging. METHODS: Using circulating placental growth factor (PlGF), a potential biomarker for hypertensive disorders of pregnancy (HDP) such as preeclampsia, as an example, we investigated how such issues can be overcome by inter-platform standardization and merging algorithms. We studied 16,462 pregnancies from 22 study cohorts. PlGF measurements (gestational age ⩾20 weeks) analyzed on one of four platforms: R&D Systems, AlereTriage, RocheElecsys or AbbottArchitect, were available for 13,429 women. Two merging algorithms, using Z-Score and Multiple of Median transformations, were applied. RESULTS: Best reference curves (BRC), based on merged, transformed PlGF measurements in uncomplicated pregnancy across six gestational age groups, were estimated. Identification of HDP by these PlGF-BRCs was compared to that of platform-specific curves. CONCLUSIONS: We demonstrate the feasibility of merging PlGF concentrations from different analytical platforms. Overall BRC identification of HDP performed at least as well as platform-specific curves. Our method can be extended to any set of biomarkers obtained from different laboratory platforms in any field. Merged biomarker data from multiple studies will improve statistical power and enlarge our understanding of the pathophysiology and management of medical syndromes.
Authors: Yue Ma; Derek L Norton; Carol A Van Hulle; Richard J Chappell; Karen K Lazar; Erin M Jonaitis; Rebecca L Koscik; Lindsay R Clark; Rachel Krause; Ulf Andreasson; Nathaniel A Chin; Barbara B Bendlin; Sanjay Asthana; Ozioma C Okonkwo; Carey E Gleason; Sterling C Johnson; Henrik Zetterberg; Kaj Blennow; Cynthia M Carlsson Journal: Alzheimers Dement (Amst) Date: 2021-05-25