Evaluation of an automated laminar cartilage T2 relaxation time analysis method in an early osteoarthritis model.

OBJECTIVE: A fully automated laminar cartilage composition (MRI-based T2) analysis method was technically and clinically validated by comparing radiographically normal knees with (CL-JSN) and without contra-lateral joint space narrowing or other signs of radiographic osteoarthritis (OA, CL-noROA). MATERIALS AND METHODS: 2D U-Nets were trained from manually segmented femorotibial cartilages (n = 72) from all 7 echoes (AllE), or from the 1st echo only (1stE) of multi-echo-spin-echo (MESE) MRIs acquired by the Osteoarthritis Initiative (OAI). Because of its greater accuracy, only the AllE U-Net was then applied to knees from the OAI healthy reference cohort (n = 10), CL-JSN (n = 39), and (1:1) matched CL-noROA knees (n = 39) that all had manual expert segmentation, and to 982 non-matched CL-noROA knees without expert segmentation. RESULTS: The agreement (Dice similarity coefficient) between automated vs. manual expert cartilage segmentation was between 0.82 ± 0.05/0.79 ± 0.06 (AllE/1stE) and 0.88 ± 0.03/0.88 ± 0.03 (AllE/1stE) across femorotibial cartilage plates. The deviation between automated vs. manually derived laminar T2 reached up to - 2.2 ± 2.6 ms/ + 4.1 ± 10.2 ms (AllE/1stE). The AllE U-Net showed a similar sensitivity to cross-sectional laminar T2 differences between CL-JSN and CL-noROA knees in the matched (Cohen's D ≤ 0.54) and the non-matched (D ≤ 0.54) comparison as the matched manual analyses (D ≤ 0.48). Longitudinally, the AllE U-Net also showed a similar sensitivity to CL-JSN vs. CS-noROA differences in the matched (D ≤ 0.51) and the non-matched (D ≤ 0.43) comparison as matched manual analyses (D ≤ 0.41). CONCLUSION: The fully automated T2 analysis showed a high agreement, acceptable accuracy, and similar sensitivity to cross-sectional and longitudinal laminar T2 differences in an early OA model, compared with manual expert analysis. TRIAL REGISTRATION: Clinicaltrials.gov identification: NCT00080171.

Unbiased analysis of knee cartilage thickness change over three years after sprifermin vs. placebo treatment - A post-hoc analysis from the phase 2B FORWARD study.

Objective: Post-treatment cartilage morphometry in the FORWARD study was performed without blinding to MRI acquisition order, involving potential reader bias. Here we obtained unbiased estimates of cartilage change post-treatment, reading year (Y)2 and Y5 MRIs with blinding to time point. We studied whether post-treatment cartilage thickness change differed between sprifermin- and placebo-treated knees. Methods: FORWARD was a 5-year randomized control trial in 549 knee osteoarthritis patients. Here, Y2/Y5 images were analyzed with blinding to relative temporal order and treatment group. Cartilage change during Y2→Y5 was obtained in 337 participants: n â€‹= â€‹57 treated with placebo intra-articular injections every 6 months (q6M); n â€‹= â€‹69 with 30 â€‹µg sprifermin every 12 months (q12 â€‹M), n â€‹= â€‹67 with 30 â€‹µg q6M, n â€‹= â€‹73 with 100 â€‹µg q12 â€‹M, and n â€‹= â€‹71 with 100 â€‹µg q6M between baseline (BL) and 18 â€‹M. Total femorotibial joint (TFTJ) cartilage thickness was the primary analytic focus. Results: TFTJ cartilage thickness change during Y2→Y5 was -26µm (SD64; 95%CI -32,-19) across the cohort; no statistically significant difference (p â€‹= â€‹0.80) was observed between Sprifermin treated or placebo arms (one-way ANOVA). All groups lost cartilage, but the treatment-related difference in cartilage thickness in Sprifermin arms relative to placebo at Y2 was maintained until Y5. Annualized cartilage change in placebo participants was -8.2 â€‹µm (SD21; 95%CI -14,-2.5) during Y2→Y5 vs. -5.4 â€‹µm (SD27; 95%CI -13,1.8) during BL→Y2; no significant difference was identified (t-test). Conclusion: FORWARD is the first study evaluating post-treatment benefits of a potential disease modifying osteoarthritis drug. Cartilage thickness gained with 100 â€‹µg sprifermin at Y2 is maintained to Y5 and thus appears viable and sustainable.This is a post-hoc analysis of the FORWARD trial: ClinicalTrials.gov Identifier: NCT01919164.

The design of a sample rapid magnetic resonance imaging (MRI) acquisition protocol supporting assessment of multiple articular tissues and pathologies in knee osteoarthritis.

Objective: This expert opinion paper proposes a design for a state-of-the-art magnetic resonance image (MRI) acquisition protocol for knee osteoarthritis clinical trials in early and advanced disease. Semi-quantitative and quantitative imaging endpoints are supported, partly amendable to automated analysis. Several (peri-) articular tissues and pathologies are covered, including synovitis. Method: A PubMed literature search was conducted, with focus on the past 5 years. Further, osteoarthritis imaging experts provided input. Specific MRI sequences, orientations, spatial resolutions and parameter settings were identified to align with study goals. We strived for implementation on standard clinical scanner hardware, with a net acquisition time ≤30 â€‹min. Results: Short- and long-term longitudinal MRIs should be obtained at ≥1.5T, if possible without hardware changes during the study. We suggest a series of gradient- and spin-echo-sequences, supporting MOAKS, quantitative analysis of cartilage morphology and T2, and non-contrast-enhanced depiction of synovitis. These sequences should be properly aligned and positioned using localizer images. One of the sequences may be repeated in each participant (re-test), optimally at baseline and follow-up, to estimate within-study precision. All images should be checked for quality and protocol-adherence as soon as possible after acquisition. Alternative approaches are suggested that expand on the structural endpoints presented. Conclusions: We aim to bridge the gap between technical MRI acquisition guides and the wealth of imaging literature, proposing a balance between image acquisition efficiency (time), safety, and technical/methodological diversity. This approach may entertain scientific innovation on tissue structure and composition assessment in clinical trials on disease modification of knee osteoarthritis.

Sexual dimorphism in peri-articular tissue anatomy - More keys to understanding sex-differences in osteoarthritis?

Objective: Osteoarthritis prevalence differs between women and men; whether this is the result of differences in pre-morbid articular or peri-articular anatomical morphotypes remains enigmatic. Albeit sex within humans cannot be reduced to female/male, this review focusses to the sexual dimorphism of peri-articular tissues, given lack of literature on non-binary subjects. Methods: Based on a Pubmed search and input from experts, we selected relevant articles based on the authors' judgement of relevance, interest, and quality; no "hard" bibliometric measures were used to evaluate the quality or importance of the work. Emphasis was on clinical studies, with most (imaging) data being available for the knee and thigh. Results: The literature on sexual dimorphism of peri-articular tissues is reviewed: 1) bone size/shape, 2) subchondral/subarticular bone, 3) synovial membrane and infra-patellar fad-pad (IPFP), 4) muscle/adipose tissue, and 5) peri-articular tissue response to treatment. Conclusions: Relevant sex-specific differences exist for 3D bone shape and IPFP size, even after normalization to body weight. Presence of effusion- and Hoffa-synovitis is associated with greater risk of incident knee osteoarthritis in overweight women, but not in men. When normalized to bone size, men exhibit greater muscle, and women greater adipose tissue measures relative to the opposite sex. Reduced thigh muscle specific strength is associated with incident knee osteoarthritis and knee replacement in women, but not in men. These observations may explain why women with muscle strength deficits have a poorer prognosis than men with similar deficits. A "one size/sex fits all" approach must be urgently abandoned in osteoarthritis research.

Anterior cruciate ligament injury and age affect knee cartilage T2 but not thickness.

OBJECTIVE: To investigate the effect of unilateral anterior cruciate ligament (ACL) injury on cartilage thickness and composition, specifically laminar transverse relaxation time (T2) by magnetic resonance imaging (MRI), in younger and older participants and to compare within-person side differences in these parameters between ACL-injured and healthy controls. DESIGN: Quantitative double-echo steady-state 3 Tesla MRI-sequences were acquired in both knees of 85 participants in four groups: 20-30 years: healthy, HEA20-30, n = 24; ACL-injured, ACL20-30, n = 23; 40-60 years: healthy, HEA40-60, n = 24; ACL-injured, ACL40-60, n = 14 (ACL injury 2-10 years prior to study inclusion). Weight-bearing femorotibial cartilages were manually segmented; cartilage T2 and thickness were computed using custom software. Mean and side differences in subregional cartilage thickness, superficial and deep cartilage T2 were compared within and between groups using non-parametric statistics. RESULTS: Cartilage thickness did not differ within or between groups. Only the side difference in medial femorotibial cartilage thickness was greater in ACL20-30 than in HEA20-30. Deep zone T2 was longer in the ACL-injured than in the contralateral uninjured knees and than in healthy controls, especially in the lateral compartment. Most ACL-injured participants had side differences in femorotibial deep zone T2 above the threshold derived from controls. CONCLUSION: In the ACL-injured knee, early compositional differences in femorotibial cartilage (T2) appear to occur in the deep zone and precede cartilage thickness loss. These results suggest that monitoring laminar T2 after ACL injury may be useful in diagnosing and monitoring early articular cartilage changes.

Clinical validation of fully automated laminar knee cartilage transverse relaxation time (T2) analysis in anterior cruciate ligament (ACL)-injured knees- on behalf of the osteoarthritis (OA)-Bio consortium.

Background: Magnetic resonance imaging (MRI) cartilage transverse relaxation time (T2) reflects cartilage composition, mechanical properties, and early osteoarthritis (OA). T2 analysis requires cartilage segmentation. In this study, we clinically validate fully automated T2 analysis at 1.5 Tesla (T) in anterior cruciate ligament (ACL)-injured and healthy knees. Methods: We studied 71 participants: 20 ACL-injured patients with, and 22 without dynamic knee instability, 13 with surgical reconstruction, and 16 healthy controls. Sagittal multi-echo-spin-echo (MESE) MRIs were acquired at baseline and 1-year follow-up. Femorotibial cartilage was segmented manually; a convolutional neural network (CNN) algorithm was trained on MRI data from the same scanner. Results: Dice similarity coefficients (DSCs) of automated versus manual segmentation in the 71 participants were 0.83 (femora) and 0.89 (tibiae). Deep femorotibial T2 was similar between automated (45.7±2.6 ms) and manual (45.7±2.7 ms) segmentation (P=0.828), whereas superficial layer T2 was slightly overestimated by automated analysis (53.2±2.2 vs. 52.1±2.1 ms for manual; P<0.001). T2 correlations were r=0.91-0.99 for deep and r=0.86-0.97 for superficial layers across regions. The only statistically significant T2 increase over 1 year was observed in the deep layer of the lateral femur [standardized response mean (SRM) =0.58 for automated vs. 0.52 for manual analysis; P<0.001]. There was no relevant difference in baseline/longitudinal T2 values/changes between the ACL-injured groups and healthy participants, with either segmentation method. Conclusions: This clinical validation study suggests that automated cartilage T2 analysis from MESE at 1.5T is technically feasible and accurate. More efficient 3D sequences and longer observation intervals may be required to detect the impact of ACL injury induced joint instability on cartilage composition (T2).

Sexual dimorphism in articular tissue anatomy - Key to understanding sex differences in osteoarthritis?

OBJECTIVE: Osteoarthritis (OA) prevalence and incidence varies between women and men, but it is unknown whether this follows sex-specific differences in systemic factors (e.g. hormones) and/or differences in pre-morbid joint anatomy. We recognize that classifications of sex within humans cannot be reduced to female/male, but given the lack of literature on non-binary individuals, this review is limited to the sexual dimorphism of articular morphotypes. METHODS: Based on a Pubmed search using relevant terms, and input from experts, we selected articles based on the authors' judgment of their relevance, interest, originality, and scientific quality; no "hard" bibliometric measures were used to evaluate their quality or importance. Focus was on clinical rather than pre-clinical studies, with most (imaging) data being available for the knee joint. RESULTS: After introducing "sexual dimorphism", the specific literature on articular morphotypes is reviewed, structured by: radiographic joint space width (JSW), meniscus, ligaments, articular cartilage morphology, articular cartilage composition and deformation, and articular tissue response to treatment. CONCLUSIONS: Sex-specific differences were clearly observed for JSW, meniscus damage, ligament size, and cartilage morphometry (volume, thickness, and surface areas) but not for cartilage composition. Ligament and cartilage measures were smaller in women even after matching for confounders. Taken together, the findings indicate that female (knee) joints may be structurally more vulnerable and at greater risk of OA. The "one size/sex fits all" approach must be abandoned in OA research, and all observational and interventional studies should report their results for sex-specific strata, at least in pre-specified secondary or post-hoc analyses.

Is detection of disease-modifying osteoarthritis drug treatment more effective when performing cartilage morphometry without blinding to MR image acquisition order?

OBJECTIVE: We here explore whether observed treatment effects of a putative disease-modifying osteoarthritis drug (DMOAD) are greater when cartilage morphometry is performed with rather than without knowledge of magnetic resonance imaging (MRI) acquisition order (unblinded/blinded to time point). METHODS: In the FORWARD (FGF-18 Osteoarthritis Randomized Controlled Trial with Administration of Repeated Doses) randomized controlled trial, 549 knee osteoarthritis patients were randomized 1:1:1:1:1 to three once-weekly intra-articular injections of placebo, 30 µg sprifermin every 6 or 12 months (M), or 100 µg every 6/12 M. After year 2, cartilage segmentation of BL through 24 M MRIs was performed, with blinding to acquisition order. After year 5, 24 and 60 M MRIs were analyzed together, with unknown relative order, but with segmented BL images as reference (24 M unblinded vs. BL), by the same operators. Total femorotibial joint cartilage thickness (TFTJ_ThC) change was obtained for 352 participants analyzed under both conditions. RESULTS: Twenty-four-month data read unblinded to order revealed a -35 ± 44 µm lower TFTJ_ThC than blinded analysis (all groups: lower/upper bounds -120/+51 µm; correlation r2 = 97%). With unblinded analysis, the placebo group lost -46 ± 57 µm TFTJ_ThC over 24 M, whereas 100 µg/every 6 M lost -2.2 ± 73 µm (difference =44 µm [95% CI: 22, 66]). With blinded analysis, placebo lost -11 ± 53 µm, whereas 100 µg/every 6 M gained 30 ± 62 µm (difference = 40 µm [95% CI: 21, 60]). 100 µg sprifermin injected every 6 M showed statistically significant (p < 0.001) treatment effects on TFTJ_ThC, with Cohen D = -0.66 for unblinded and D = -0.69 for blinded analysis. CONCLUSIONS: These results do not reveal that detection of proposed DMOAD treatment is enhanced with MRIs read unblinded to order; rather, the sensitivity is similar to blinded analysis. Choices on blinded vs. unblinded analysis may thus be based on other criteria.

30 Years of MRI-based cartilage & bone morphometry in knee osteoarthritis: From correlation to clinical trials.

OBJECTIVE: The first publication on morphometric analysis of articular cartilage using magnetic resonance imaging (MRI) in 1994 set the scene for a game change in osteoarthritis (OA) research. The current review highlights milestones in cartilage and bone morphometry, summarizing the rapid progress made in imaging, its application to understanding joint (patho-)physiology, and its use in interventional clinical trials. METHODS: Based on a Pubmed search of articles from 1994 to 2023, the authors subjectively selected representative work illustrating important steps in the development or application of magnetic resonance-based cartilage and bone morphometry, with a focus on studies in humans, and on the knee. Research on OA-pathophysiology is addressed only briefly, given length constraints. Compositional and semi-quantitative assessment are not covered here. RESULTS: The selected articles are presented in historical order as well as by content. We review progress in the technical aspects of image acquisition, segmentation and analysis, advances in understanding tissue growth, physiology, function, and adaptation, and a selection of clinical trials examining the efficacy of interventions on knee cartilage and bone. A perspective is provided of how lessons learned may be applied to future research and clinical management. CONCLUSIONS: Over the past 30 years, MRI-based morphometry of cartilage and bone has contributed to a paradigm shift in understanding articular tissue physiology and OA pathophysiology, and to the development of new treatment strategies. It is likely that these technologies will continue to play a key role in the development and (accelerated) approval of therapy, potentially targeted to different OA phenotypes.

Imaging Biomarkers of Osteoarthritis.

Currently no disease-modifying osteoarthritis drug has been approved for the treatment of osteoarthritis (OA) that can reverse, hold, or slow the progression of structural damage of OA-affected joints. The reasons for failure are manifold and include the heterogeneity of structural disease of the OA joint at trial inclusion, and the sensitivity of biomarkers used to measure a potential treatment effect.This article discusses the role and potential of different imaging biomarkers in OA research. We review the current role of radiography, as well as advances in quantitative three-dimensional morphological cartilage assessment and semiquantitative whole-organ assessment of OA. Although magnetic resonance imaging has evolved as the leading imaging method in OA research, recent developments in computed tomography are also discussed briefly. Finally, we address the experience from the Foundation for the National Institutes of Health Biomarker Consortium biomarker qualification study and the future role of artificial intelligence.

Effects of neuromuscular control and strengthening exercises on MRI-measured thigh tissue composition and muscle properties in people with knee osteoarthritis - an exploratory secondary analysis from a randomized controlled trial.

OBJECTIVE: To investigate the effects of adding strength training to neuromuscular control exercises on thigh tissue composition and muscle properties in people with radiographic-symptomatic knee osteoarthritis (KOA). METHODS: In this exploratory secondary analysis of a randomized controlled trial, using a complete-case approach, participants performed 12 weeks of twice-weekly neuromuscular control exercise and patient education (NEMEX, n = 34) or NEMEX plus quadriceps strength training (NEMEX+ST, n = 29). Outcomes were MRI-measured inter- and intramuscular adipose tissue (InterMAT, IntraMAT), quadriceps muscle cross-sectional area (CSA), knee-extensor strength, specific strength (strength/lean CSA) and 30 s chair-stands. Between-group effects were compared using a mixed model analysis of variance. RESULTS: At 12 weeks, responses to NEMEX+ST overlapped with NEMEX for all outcomes. Both groups reduced InterMAT (NEMEX+ST=25 %, NEMEX=21 %); between-group difference: 0.8cm2 (95 % CI: -0.1, 1.7). NEMEX+ST decreased IntraMAT (2 %) and NEMEX increased IntraMAT (4 %); between-group difference 0.1 %-points (-0.3, 0.5). Both groups increased quadriceps CSA and lean CSA (CSA minus IntraMAT), improved knee-extensor strength and specific strength, and improved chair-stand performance with a trend towards greater effects in NEMEX+ST. CONCLUSION: Adding strength training to 12 weeks of neuromuscular control exercises provided largely similar effects to neuromuscular control exercises alone in decreasing InterMAT and IntraMAT, in improving knee-extensor strength, CSA and in improving performance-based function in KOA persons, with a trend towards greater effects with additional strength training. Notably, both groups substantially reduced InterMAT and improved specific strength (an index of muscle quality). Our hypothesis-generating work warrants exploration of the roles played by InterMAT and IntraMAT in exercise effects in KOA.

Association between knee magnetic resonance imaging markers and knee symptoms over 6-9 years in young adults.

OBJECTIVES: To describe associations between MRI markers with knee symptoms in young adults. METHODS: Knee symptoms were assessed using the WOMAC scale during the Childhood Determinants of Adult Health Knee Cartilage study (CDAH-knee; 2008-2010) and at the 6- to 9-year follow-up (CDAH-3; 2014-2019). Knee MRI scans obtained at baseline were assessed for morphological markers (cartilage volume, cartilage thickness, subchondral bone area) and structural abnormalities [cartilage defects and bone marrow lesions (BMLs)]. Univariable and multivariable (age, sex, BMI adjusted) zero-inflated Poisson (ZIP) regression models were used for analysis. RESULTS: The participants' mean age in CDAH-knee and CDAH-3 were 34.95 (s.d. 2.72) and 43.27 (s.d. 3.28) years, with 49% and 48% females, respectively. Cross-sectionally, there was a weak but significant negative association between medial femorotibial compartment (MFTC) [ratio of the mean (RoM) 0.99971084 (95% CI 0.9995525, 0.99986921), P < 0.001], lateral femorotibial compartment (LFTC) [RoM 0.99982602 (95% CI 0.99969915, 0.9999529), P = 0.007] and patellar cartilage volume [RoM 0.99981722 (95% CI 0.99965326, 0.9999811), P = 0.029] with knee symptoms. Similarly, there was a negative association between patellar cartilage volume [RoM 0.99975523 (95% CI 0.99961427, 0.99989621), P = 0.014], MFTC cartilage thickness [RoM 0.72090775 (95% CI 0.59481806, 0.87372596), P = 0.001] and knee symptoms assessed after 6-9 years. The total bone area was negatively associated with knee symptoms at baseline [RoM 0.9210485 (95% CI 0.8939677, 0.9489496), P < 0.001] and 6-9 years [RoM 0.9588811 (95% CI 0.9313379, 0.9872388), P = 0.005]. The cartilage defects and BMLs were associated with greater knee symptoms at baseline and 6-9 years. CONCLUSION: BMLs and cartilage defects were positively associated with knee symptoms, whereas cartilage volume and thickness at MFTC and total bone area were weakly and negatively associated with knee symptoms. These results suggest that the quantitative and semiquantitative MRI markers can be explored as a marker of clinical progression of OA in young adults.

The Impact of Exercise Training and Supplemental Oxygen on Peripheral Muscles in Chronic Obstructive Pulmonary Disease: A Randomized Controlled Trial.

OBJECTIVE: Exercise training is a cornerstone of the treatment of chronic obstructive pulmonary disease, whereas the related interindividual heterogeneity in skeletal muscle dysfunction and adaptations are not yet fully understood. We set out to investigate the effects of exercise training and supplemental oxygen on functional and structural peripheral muscle adaptation. METHODS: In this prospective, randomized, controlled, double-blind study, 28 patients with nonhypoxemic chronic obstructive pulmonary disease (forced expiratory volume in 1 second, 45.92% ± 9.06%) performed 6 wk of combined endurance and strength training, three times a week while breathing either supplemental oxygen or medical air. The impact on exercise capacity, muscle strength, and quadriceps femoris muscle cross-sectional area (CSA) was assessed by maximal cardiopulmonary exercise testing, 10-repetition maximum strength test of knee extension, and magnetic resonance imaging, respectively. RESULTS: After exercise training, patients demonstrated a significant increase in functional capacity, aerobic capacity, exercise tolerance, quadriceps muscle strength, and bilateral CSA. Supplemental oxygen affected significantly the training impact on peak work rate when compared with medical air (+0.20 ± 0.03 vs +0.12 ± 0.03 W·kg -1 , P = 0.047); a significant increase in CSA (+3.9 ± 1.3 cm 2 , P = 0.013) was only observed in the training group using oxygen. Supplemental oxygen and exercise-induced peripheral desaturation were identified as significant opposing determinants of muscle gain during this exercise training intervention, which led to different adaptations of CSA between the respective subgroups. CONCLUSIONS: The heterogenous functional and structural muscle adaptations seem determined by supplemental oxygen and exercise-induced hypoxia. Indeed, supplemental oxygen may facilitate muscular training adaptations, particularly in limb muscle dysfunction, thereby contributing to the enhanced training responses on maximal aerobic and functional capacity.

Association of quantitative measures of medial meniscal extrusion with structural and symptomatic knee osteoarthritis progression - Data from the OAI FNIH biomarker study.

OBJECTIVE: To study the association of quantitative medial meniscal position measures with radiographic and symptomatic knee osteoarthritis (OA) progression over 2-4 years. METHODS: The FNIH OAI Biomarkers study comprised 600 participants in four subgroups: 194 case knees with combined structural (medial minimum joint space width (minJSW) loss ≥0.7 mm) and symptomatic (persistent Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain subscale increase ≥9 [0-100 scale]) progression; 200 knees with neither structural nor symptomatic progression; 103 knees with isolated structural and 103 with isolated symptomatic progression. Coronal double echo at steady state (DESS) MRIs were used for segmenting five central slices of the medial meniscus. Associations with progression were examined using logistic regression (adjusted for demographic and clinical data). RESULTS: Greater baseline medial meniscal extrusion was associated with combined structural/symptomatic progression (OR 1.59; 95%CI: [1.25,2.04]). No relationship was observed for tibial plateau coverage or meniscal overlap distance. The two-year increase in meniscal extrusion (OR 1.48 [1.21, 1.83]), and reduction in tibial plateau coverage (OR 0.70 [0.58,0.86]) and overlap distance (OR 0.73 [0.60,0.89]) were associated with combined progression. Greater baseline extrusion was associated with isolated structural and less extrusion with isolated symptomatic progression. The longitudinal increase in meniscal extrusion, and reduction in tibial plateau coverage and overlap distance were associated with structural, but not with symptomatic progression. CONCLUSION: Baseline measures of medial meniscal extrusion were consistently positively associated with combined radiographic/symptomatic progression and with isolated structural, but not with isolated symptomatic progression. These measures may therefore allow one to assess the risk of structural knee OA progression and to monitor interventions restoring meniscal position and function.

Comparison of evaluation metrics of deep learning for imbalanced imaging data in osteoarthritis studies.

PURPOSE: To compare the evaluation metrics for deep learning methods that were developed using imbalanced imaging data in osteoarthritis studies. MATERIALS AND METHODS: This retrospective study utilized 2996 sagittal intermediate-weighted fat-suppressed knee MRIs with MRI Osteoarthritis Knee Score readings from 2467 participants in the Osteoarthritis Initiative study. We obtained probabilities of the presence of bone marrow lesions (BMLs) from MRIs in the testing dataset at the sub-region (15 sub-regions), compartment, and whole-knee levels based on the trained deep learning models. We compared different evaluation metrics (e.g., receiver operating characteristic (ROC) and precision-recall (PR) curves) in the testing dataset with various class ratios (presence of BMLs vs. absence of BMLs) at these three data levels to assess the model's performance. RESULTS: In a subregion with an extremely high imbalance ratio, the model achieved a ROC-AUC of 0.84, a PR-AUC of 0.10, a sensitivity of 0, and a specificity of 1. CONCLUSION: The commonly used ROC curve is not sufficiently informative, especially in the case of imbalanced data. We provide the following practical suggestions based on our data analysis: 1) ROC-AUC is recommended for balanced data, 2) PR-AUC should be used for moderately imbalanced data (i.e., when the proportion of the minor class is above 5% and less than 50%), and 3) for severely imbalanced data (i.e., when the proportion of the minor class is below 5%), it is not practical to apply a deep learning model, even with the application of techniques addressing imbalanced data issues.

Machine-learning predicted and actual 2-year structural progression in the IMI-APPROACH cohort.

In the Innovative Medicine's Initiative Applied Public-Private Research enabling OsteoArthritis Clinical Headway (IMI-APPROACH) knee osteoarthritis (OA) study, machine learning models were trained to predict the probability of structural progression (s-score), predefined as >0.3 mm/year joint space width (JSW) decrease and used as inclusion criterion. The current objective was to evaluate predicted and observed structural progression over 2 years according to different radiographic and magnetic resonance imaging (MRI)-based structural parameters. Radiographs and MRI scans were acquired at baseline and 2-year follow-up. Radiographic (JSW, subchondral bone density, osteophytes), MRI quantitative (cartilage thickness), and MRI semiquantitative [SQ; cartilage damage, bone marrow lesions (BMLs), osteophytes] measurements were obtained. The number of progressors was calculated based on a change exceeding the smallest detectable change (SDC) for quantitative measures or a full SQ-score increase in any feature. Prediction of structural progression based on baseline s-scores and Kellgren-Lawrence (KL) grades was analyzed using logistic regression. Among 237 participants, around 1 in 6 participants was a structural progressor based on the predefined JSW-threshold. The highest progression rate was seen for radiographic bone density (39%), MRI cartilage thickness (38%), and radiographic osteophyte size (35%). Baseline s-scores could only predict JSW progression parameters (most P>0.05), while KL grades could predict progression of most MRI-based and radiographic parameters (P<0.05). In conclusion, between 1/6 and 1/3 of participants showed structural progression during 2-year follow-up. KL scores were observed to outperform the machine-learning-based s-scores as progression predictor. The large amount of data collected, and the wide range of disease stage, can be used for further development of more sensitive and successful (whole joint) prediction models. Trial Registration: Clinicaltrials.gov number NCT03883568.

Quantitative CT of the knee in the IMI-APPROACH osteoarthritis cohort: Association of bone mineral density with radiographic disease severity, meniscal coverage and meniscal extrusion.

OBJECTIVE: Osteoarthritis (OA) is a highly prevalent chronic condition. The subchondral bone plays an important role in onset and progression of OA making it a potential treatment target for disease-modifying therapeutic approaches. However, little is known about changes of periarticular bone mineral density (BMD) in OA and its relation to meniscal coverage and meniscal extrusion at the knee. Thus, the aim of this study was to describe periarticular BMD in the Applied Public-Private Research enabling OsteoArthritis Clinical Headway (APPROACH) cohort at the knee and to analyze the association with structural disease severity, meniscal coverage and meniscal extrusion. DESIGN: Quantitative CT (QCT), MRI and radiographic examinations were acquired in 275 patients with knee osteoarthritis (OA). QCT was used to assess BMD at the femur and tibia, at the cortical bone plate (Cort) and at the epiphysis at three locations: subchondral (Sub), mid-epiphysis (Mid) and adjacent to the physis (Juxta). BMD was evaluated for the medial and lateral compartment separately and for subregions covered and not covered by the meniscus. Radiographs were used to determine the femorotibial angle and were evaluated according to the Kellgren and Lawrence (KL) system. Meniscal extrusion was assessed from 0 to 3. RESULTS: Mean BMD differed significantly between each anatomic location at both the femur and tibia (p < 0.001) in patients with KL0. Tibial regions assumed to be covered with meniscus in patients with KL0 showed lower BMD at Sub (p < 0.001), equivalent BMD at Mid (p = 0.07) and higher BMD at Juxta (p < 0.001) subregions compared to regions not covered with meniscus. Knees with KL2-4 showed lower Sub (p = 0.03), Mid (p = 0.01) and Juxta (p < 0.05) BMD at the medial femur compared to KL0/1. Meniscal extrusion grade 2 and 3 was associated with greater BMD at the tibial Cort (p < 0.001, p = 0.007). Varus malalignment is associated with significant greater BMD at the medial femur and at the medial tibia at all anatomic locations. CONCLUSION: BMD within the epiphyses of the tibia and femur decreases with increasing distance from the articular surface. Knees with structural OA (KL2-4) exhibit greater cortical BMD values at the tibia and lower BMD at the femur at the subchondral level and levels beneath compared to KL0/1. BMD at the tibial cortical bone plate is greater in patients with meniscal extrusion grade 2/3.

Worsening of Articular Tissue Damage as Defined by Semi-Quantitative MRI Is Associated With Concurrent Quantitative Cartilage Loss Over 24 Months.

OBJECTIVE: To assess the association of worsening of magnetic resonance imaging (MRI) semi-quantitative (SQ) tissue features with concurrent change in quantitative (Q) cartilage thickness measurements over 24 months within the Foundation for the National Institutes of Health (FNIH) Biomarker Consortium study. METHODS: In all, 599 participants were included. SQ assessment included cartilage damage, meniscal extrusion and damage, osteophytes, bone marrow lesions (BMLs), and effusion- and Hoffa-synovitis. Change in medial compartment Q cartilage thickness was stratified by concurrent ipsicompartmental SQ changes. Between-group comparisons were performed using analysis of covariance (ANCOVA) with adjustment for age, sex, and body mass index (BMI). Results were presented as adjusted mean difference. RESULTS: Knees with any increase in SQ cartilage scores in the medial compartment (n = 268) showed more Q cartilage loss compared to knees that remained stable (mean adjusted difference [MAD] = -0.16 mm, 95% confidence interval [CI]: [-0.19, -0.13] mm). Knees with any increase in meniscal extrusion in the medial compartment (n = 98) showed more Q cartilage loss than knees without (MAD = -0.18 mm, 95% CI: [-0.22, -0.14] mm. Comparable findings were seen for meniscal damage worsening. Regarding BMLs, an increase by one subregion resulted in a MAD of Q cartilage loss of -0.10 mm, 95% CI: [-0.14, -0.06] mm, while this effect almost tripled for change in two or more subregions. Increase in either effusion- and/or Hoffa-synovitis by one grade resulted in a MAD of -0.07 mm, 95% CI: [-0.10, -0.03] mm. CONCLUSION: Worsening of SQ cartilage damage, meniscal extrusion and damage, number of subregions affected by BML, maximum size of BMLs and worsening of effusion- and/or Hoffa synovitis is associated with increased Q cartilage loss.

Exploring the differences between radiographic joint space width and MRI cartilage thickness changes using data from the IMI-APPROACH cohort.

OBJECTIVE: Longitudinal weight-bearing radiographic joint space width (JSW) and non-weight-bearing MRI-based cartilage thickness changes often show weak correlations. The current objective was to investigate these correlations, and to explore the influence of different factors that could contribute to longitudinal differences between the two methods. METHODS: The current study included 178 participants with medial osteoarthritis (OA) out of the 297 knee OA participants enrolled in the IMI-APPROACH cohort. Changes over 2 years in medial JSW (ΔJSWmed), minimum JSW (ΔJSWmin), and medial femorotibial cartilage thickness (ΔMFTC) were assessed using linear regression, using measurements from radiographs and MRI acquired at baseline, 6 months, and 1 and 2 years. Pearson R correlations were calculated. The influence of cartilage quality (T2 mapping), meniscal extrusion (MOAKS scoring), potential pain-induced unloading (difference in knee-specific pain scores), and increased loading (BMI) on the correlations was analyzed by dividing participants in groups based on each factor separately, and comparing correlations (slope and strength) between groups using linear regression models. RESULT: Correlations between ΔMFTC and ΔJSWmed and ΔJSWmin were statistically significant (p < 0.004) but weak (R < 0.35). Correlations were significantly different between groups based on cartilage quality and on meniscal extrusion: only patients with the lowest T2 values and with meniscal extrusion showed significant moderate correlations. Pain-induced unloading or BMI-induced loading did not influence correlations. CONCLUSIONS: While the amount of loading does not seem to make a difference, weight-bearing radiographic JSW changes are a better reflection of non-weight-bearing MRI cartilage thickness changes in knees with higher quality cartilage and with meniscal extrusion.