Development and validation of a machine learning-supported strategy of patient selection for osteoarthritis clinical trials - the IMI-APPROACH study

Aug 4, 2026

Upload by Eliana

Osteoarthritis clinical trials require progression-enriched patient populations to efficiently evaluate disease-modifying effects of new treatments. The IMI-APPROACH study developed and validated a two-stage, machine learning-supported recruitment strategy that ranked candidates by likelihood of progression. First-stage models used pre-existing cohort data for screening-visit selection, and the second stage used screening data to support final inclusion. Effectiveness was assessed by actual 24-month progression and model performance.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Abstract

Objectives

Design

Results

Conclusions

Introduction

Patient selection challenges in clinical trials

Rationale for enrichment in osteoarthritis

| Development and validation of a machine learning-supported strategy of patient selection for osteoarthritis clinical trials: the IMI-APPROACH study | | | | | --- | --- | --- | --- | | Paweł Widera a, Paco M.J. Welsingb, Samuel O. Danso a, Sjaak Peelen c, Margreet Kloppenburg d, Marieke Loefd, Anne C. Marijnissen b, Eefje M. van Helvoortb, Francisco J. Blanco e,
Joana Magalh~aes e, Francis Berenbaum f, Ida K. Haugen g, Anne-Christine Bay-Jensen h, Ali Mobasherib, i,j, k, l, Christoph Ladel m, John Loughlin n, Floris P.J.G. Lafeber b, Agns Lalande o,
*
Jonathan Larkin p, Harrie Weinans q, Jaume Bacardit a,
a School of Computing, Newcastle University, Newcastle, UK
b Department of Rheumatology & Clinical Immunology, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands c Lygature, Utrecht, the Netherlands
d Department of Rheumatology, Leiden University Medical Center, Leiden, the Netherlands e Institute of Biomedical Research, University Hospital of A Coru~na, A Coru~na, Spain fAPHP Hospital Saint-Antoine, Paris, France
g Division of Rheumatology and Research, Diakonhjemmet Hospital, Oslo, Norway h Nordic Bioscience, Herlev, Denmark
i Research Unit of Medical Imaging, Physics and Technology, Faculty of Medicine, University of Oulu, Oulu, Finland j Department of Regenerative Medicine, State Research Institute Centre for Innovative Medicine, Vilnius, Lithuania k Department of Joint Surgery, First Afﬁliated Hospital of Sun Yat-sen University, Guangzhou, China
l World Health Organization Collaborating Centre for Public Health Aspects of Musculoskeletal Health and Aging, Liege, Belgium m BioBone B. V., Amsterdam, Netherlands
n Bioscience Institute, Newcastle University, International Centre for Life, Newcastle, UK
o Servier International Research Institute, Suresnes, France
p Novel Human Genetics Research Unit, GlaxoSmithKline, Collegeville, United States
q Department of Orthopedics, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands | | | | | A R T I C L E I N F O | | A B S T R A C T | | | Handling Editor: Professor H Madry | | Objectives: To efﬁciently assess the disease-modifying potential of new osteoarthritis treatments, clinical trials need progression-enriched patient populations. To assess whether the application of machine learning results inpatient selection enrichment, we developed a machine learning recruitment strategy targeting progressive patients and validated it in the IMI-APPROACH knee osteoarthritis prospective study.
Design: We designed a two-stage recruitment process supported by machine learning models trained to rank candidates by the likelihood of progression. First stage models used data from pre-existing cohorts to select patients for a screening visit. The second stage model used screening data to inform the ﬁnal inclusion. The effectiveness of this process was evaluated using the actual 24-month progression.
Results: From 3500 candidate patients, 433 with knee osteoarthritis were screened, 297 were enrolled, and 247 completed the 2-year follow-up visit. We observed progression related to pain (P, 30%), structure (S, 13%), and combined pain and structure (P þ S, 5%), and a proportion of non-progressors (N, 52%) ~15% lower vs an unenriched population. Our model predicted these outcomes with AUC of 0.86 [95% CI, 0.81–0.90] for painrelated progression and AUC of 0.61 [95% CI, 0.52–0.70] for structure-related progression. Progressors were ranked higher than non-progressors for P þ S (median rank 65 vs 143, AUC ¼ 0.75), P (median rank 77 vs 143, AUC ¼ 0.71), and S patients (median rank 107 vs 143, AUC ¼ 0.57).
Conclusions: The machine learning-supported recruitment resulted in enriched selection of progressive patients. Further research is needed to improve structural progression prediction and assess this strategy in an interventional trial. | | | Keywords:
Osteoarthritis
Disease progression p

FAQs

What problem does the IMI-APPROACH study address in osteoarthritis clinical trials?

Clinical trials need progression-enriched patient populations so disease-modifying effects can be observed within the trial duration. The study targets the bottleneck of selecting the right patients for expected progression.

How does the machine learning recruitment strategy work?

It uses a two-stage process. Stage one ranks and selects candidates using models trained on pre-existing cohort data, and stage two refines final inclusion using screening-visit data.

What outcomes did the study observe after applying the strategy?

Among screened candidates, progression related to pain and/or structure was measured over 24 months, with non-progressors reduced versus an unenriched population. Model discrimination performance was reported using AUC values for pain- and structure-related progression.

Aug 4, 2026

Upload by Eliana