Patient population:
-20 early stage knee OA patients
-Included via the outpatient clinic of the Department of Orthopedic Surgery of the Erasmus MC,
Rotterdam,
The Netherlands
-Inclusion criteria:
-Knee complaints: > 3 months
-Kellgren & Lawrence grade I or II on radiography
-Exclusion criteria:
-Contra-indications to undergo MRI
-Knee surgery schduled within the coming 6 months
-Viscosupplementation during the last year
Study measurements and intervention:
-Basline measurements consisted of:
-3D dGEMRIC (see Van Tiel,
et al.
Eur Radiol.
2013 Feb;23(2):496-
504 for MR and protocol)
-Knee Osteoarthritis Outcome Score (KOOS) (Roos 2003,
de Groot
2008)
-0=much knee complaints - 100=no knee complaints:
-5 subscales
-Pain
-Symptoms
-Daily activities (ADL)
-Sports activities (sports)
-Quality of life (QOL)
-Pain score
-10=much knee complaints - 0=no knee complaints
-Intervention:
-Hylan G-F 20 (Synvisc,
Genzyme,
USA) was intra-articularly
injected 3 times with a one week interval
-Follow-up measurments:
-Same as baseline measurements
-14 weeks was chosen because significant good clinical results of
viscosupplementation was shwon up to 14 weeks after the
treatment
dGEMRIC analysis:
-Patient motion causes errors in T1 calculation (Miese 2011,
Studler
2010)
-Unrealistic high T1 values with a high error in the T1 fit (Figure
2AB)
-Registration algorithm to correct for (potential) patient motion
(Bron,
Van Tiel,
et al.
Eur Radiol.
2013 Jan;23(1):246-52)
-Realistic T1 values with a low error in the T1 fit (Figure 2CD)
-After registartion,
dGEMRIC T1 values evaluated in 4 cartilage regions (ROIs) per knee compartment (Figure 3)
Statistical analysis:
-For each patiet all dGEMRIC T1 relaxation time per ROI
-KOOS analyzed for each subscale together with the Pain score
-Outcomes were compared using paired testing (Wilcoxon Signed Rank test) with correction for multiple testing (post-hoc Bonferroni)