There are about 1.8 million people in Germany with impaired kidney function (eGFR <60ml/min/1.73m²) [1,2].
Some people progress to endstage renal disease in which renal replacement therapy needs to be initiated [3]. This can either be dialysis or kidney transplantation. Kidney transplantation offers better outcomes in terms of mortality [4]. However due to organ shortage the waiting time for a kidney transplant is about 10 years so that patients usually need to start dialysis treatment (hemodialysis: 4-5 hours, three times a week, peritoneal dialysis: 3-4 dialysis fluid exchanges daily).
At the end of 2021, over 11.000 people on the waiting list for kidney transplantation in Germany. Kidney transplantation improves quality of life compared to dialysis. In contrast to patient and graft survival improvement short-term (first two years), patient and graft survival long-term remained stagnant. About 5% of patients suffer from graft loss and thus need to return to dialysis. Unfortunately, early detection of those high-risk patients is lacking [5].
Therefore, a project funded by BIH to detect graft loss using an AI approach was designed to fill this void: Patient data from the clinical transplant database “TBase” was retrieved including all patients above 18 years who had undergone kidney only transplantation at Charité, Campus Mitte, after 2000. Data consisted of demographic data of recipient, transplant and donor, examination reports from microbiology, pathology and clinical notes, laboratory values and hospitalizations at Charité. In collaboration with Master students Jan Frick, Rouven Reuter, Rashik Islam, PhD student Jens-Michalis Papaioannou and Prof. Dr.-Ing. habil. Alexander Löser from the Berliner Hochschule für Technik and medical specialists Dr. med. Marcel Naik and Prof. Dr. med. Klemens Budde of the Medical Department of Nephrology at the Charite, Campus Mitte we developed a model that uses multimodal input to automatically predict graft failure up to 8 years before happening.
[1] Girndt M, Trocchi P, Scheidt-Nave C, Markau S, Stang A. The Prevalence of Renal Failure. Results from the German Health Interview and Examination Survey for Adults, 2008-2011 (DEGS1). Dtsch Arztebl Int. 2016 Feb 12;113(6):85-91. doi: 10.3238/arztebl.2016.0085. PMID: 26931624; PMCID: PMC4782264.
[2] Helmut Reichel, Jarcy Zee, Charlotte Tu, Eric Young, Ronald L Pisoni, Bénédicte Stengel, Johannes Duttlinger, Gerhard Lonnemann, Bruce M Robinson, Roberto Pecoits-Filho, Danilo Fliser, Chronic kidney disease progression and mortality risk profiles in Germany: results from the Chronic Kidney Disease Outcomes and Practice Patterns Study, Nephrology Dialysis Transplantation, Volume 35, Issue 5, May 2020, Pages 803–810, https://doi.org/10.1093/ndt/gfz260
[3] Brück K, Stel VS, Gambaro G, Hallan S, Völzke H, Ärnlöv J, Kastarinen M, Guessous I, Vinhas J, Stengel B, Brenner H, Chudek J, Romundstad S, Tomson C, Gonzalez AO, Bello AK, Ferrieres J, Palmieri L, Browne G, Capuano V, Van Biesen W, Zoccali C, Gansevoort R, Navis G, Rothenbacher D, Ferraro PM, Nitsch D, Wanner C, Jager KJ; European CKD Burden Consortium. CKD Prevalence Varies across the European General Population. J Am Soc Nephrol. 2016 Jul;27(7):2135-47. doi: 10.1681/ASN.2015050542. Epub 2015 Dec 23. PMID: 26701975; PMCID: PMC4926978.
[4] Kramer A, Pippias M, Noordzij M, et al. The European Renal Association - European Dialysis and Transplant Association (ERA-EDTA) Registry Annual Report 2015: a summary. Clin Kidney J 2018;11:108-22
[5] Wekerle T, Segev D, Lechler R, et al. Strategies for long-term preservation of kidney graft function. Lancet 2017;389:2152-62.
