Patellofemoral pain (PFP) is a common knee pathology with unknown pathogenesis. Potential pathogenic mechanisms are a diminished patellofemoral cartilage composition or vascular problems. Both can be studied in depth with advanced MRI techniques. Cartilage composition can be measured quantitatively with T1GD (delayed gadolinium enhanced MRI of cartilage, dGEMRIC), T1ρ and T2 mapping sequences. Dynamic contrast-enhanced MRI (DCE-MRI) of bone provides a means to quantitatively analyze blood perfusion (Figure 1). Previously, these advanced MRI techniques have neither been applied before in the patella nor in a case-control study in the context of PFP. The purpose of the present study was to apply multiple quantitative MRI techniques to investigate differences in patellofemoral cartilage composition and patellar bone blood perfusion between patients with PFP and healthy controls. This may lead to a better understanding of the pathogenesis of PFP.PFP patients and healthy controls aged 14-40 years underwent MRI at 3T (Discovery MR750, GE Healthcare, Milwaukee, USA) using a dedicated 8-channel knee coil. The MRI protocol included spoiled gradient-echo (SPGR) with high spatial resolution, T1GD, T1ρ and T2 mapping sequences and DCE-MRI. T1GD and DCE-MRI were only conducted in adults due to contrast administration. Contrast was administered after acquisition of the SPGR, T1ρ and T2 sequences. The image protocol for DCE-MRI consisted of 35 phases with intravenous contrast administration (0.2 mmol/kg Magnevist (Bayer, Berlin, Germany)) 2 ml/s starting after the first phase. Other MR parameters are described in Table 1. An experienced observer manually annotated the whole trochlear and patellar cartilage and the patellar bone on the SPGR images with Matlab (R2011a, The MathWorks, Natick, MA, USA). Automated image registration was applied to compensate for subject motion within and between sequences1,2. For each sequence, the separate images were rigidly registered to the image with the highest contrast. Subsequently, T1GD, T1ρ, T2 and DCE sequences were rigidly registered to the SPGR images based on mutual information, such that a comparison between exactly matching VOIs for all sequences was possible. After registration, the relaxation times and perfusion parameters of the VOI were fitted voxel-wise by a Maximum-Likelihood (ML) estimator3 and summarized with the reciprocal Cramér-Rao-Lower-Bound (indicating fit uncertainty) weighted mean. Perfusion parameters (Ktrans and kep) were calculated with the pharmacokinetic model of Tofts4, using an arterial input function computed from the popliteal artery. Differences in perfusion parameters were compared between patients and control subjects after logarithmic transformation by linear regression analyses, adjusted for age, BMI, gender, sports participation. The same applied for relaxation times, with the addition of time of image acquisition as potential confounder.64 patients and 70 controls were included. Mean age was 23.2(6.4) years, mean BMI was 22.9(3.4)kg/m2 and 56.7% was female. T1ρ mapping was only conducted in 52 patients and 64 control subjects and DCE-MRI in 35 patients and 44 control subjects, since these sequences were not yet available at the beginning of the study. Mean T1GD relaxation times of patellar (657.8 vs. 669.4 milliseconds(ms)) and femoral cartilage (661.6 vs. 659.8ms) did not significantly differ between patients and controls. In addition, no significant differences in mean T1ρ relaxation times of patellar (46.9 vs. 46.0ms) and femoral cartilage (50.8 vs. 50.2ms) and mean T2 relaxation times of patellar (33.2 vs. 32.9ms) and femoral cartilage (36.7 vs. 36.6ms) between patients and controls were found. Mean Kep was 0.189 (0.147)min-1 for patients and 0.154 (0.114) min-1 for control subjects (Table 2). Mean Ktrans was 0.019 (0.015)min-1 for patients and 0.014 (0.009) min-1 for control subjects. Both quantitative perfusion parameters were not statistically significantly different between patients and control subjects (Table 2).This is the first study successfully applying multiple advanced quantitative MRI techniques measuring cartilage composition and bone blood perfusion in a large case-control study in the context of PFP. In particular, quantification of bone blood perfusion was challenging, due to the relatively poor vascularization of bone. Relaxation times of cartilage were quite constant within and across groups. More variation was observed in the perfusion parameters. This large inter-subject variability, possibly caused by measurement errors or normal tissue heterogeneity, makes it difficult to detect a possible difference. Therefore, although the differences in mean Ktrans and mean Kep between groups were not significant, we cannot rule out there are none.Multiparametric quantitative MRI showed no significant differences in patellofemoral cartilage composition and patellar bone blood perfusion between patients with PFP and healthy controls.