Objective: To evaluate the efficacy and safety of the anti-catabolic ADAMTS-5 inhibitor S201086/ GLPG1972 for the treatment of symptomatic knee osteoarthritis. Design: ROCCELLA (NCT03595618) was a randomized, double-blind, placebo-controlled, dose-ranging, phase 2 trial in adults (aged 40-75 years) with knee osteoarthritis. Participants had moderate-to-severe pain in the target knee, Kellgren-Lawrence grade 2 or 3 and Osteoarthritis Research Society International joint space narrowing (grade 1 or 2). Participants were randomized 1:1:1:1 to once-daily oral S201086/GLPG1972 75,150 or 300 mg, or placebo for 52 weeks. The primary endpoint was change from baseline to week 52 in central medial femorotibial compartment (cMFTC) cartilage thickness assessed quantitatively by magnetic resonance imaging. Secondary endpoints included change from baseline to week 52 in radiographic joint space width, Western Ontario and McMaster Universities Osteoarthritis Index total and subscores, and pain (visual analogue scale). Treatment-emergent adverse events (TEAEs) were also recorded. Results: Overall, 932 participants were enrolled. No significant differences in cMFTC cartilage loss were observed between placebo and S201086/GLPG1972 therapeutic groups: placebo vs 75 mg, P = 0.165; vs 150 mg, P = 0.939; vs 300 mg, P = 0.682. No significant differences in any of the secondary endpoints were observed between placebo and treatment groups. Similar proportions of participants across treatment groups experienced TEAEs. Conclusions: Despite enrolment of participants who experienced substantial cartilage loss over 52 weeks, during the same time period, S201086/GLPG1972 did not significantly reduce rates of cartilage loss or modify symptoms in adults with symptomatic knee osteoarthritis. & COPY; 2023 The Authors. Published by Elsevier Ltd on behalf of Osteoarthritis Research Society International. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Purpose: To date, no disease-modifying osteoarthritis drugs (DMOADs) have been approved for the treatment of knee osteoarthritis (OA). S201086/GLPG1972 is a potent and selective inhibitor of ADAMTS-5 (a disintegrin and metalloproteinase with thrombospondin motif-5) in development as a DMOAD, hypothesized to reduce cartilage loss via inhibition of the enzymatic cleavage of aggrecan (a key component of cartilage). S201086/GLPG1972 has been shown to reduce cartilage degradation in preclinical models of OA. Therefore, we aimed to evaluate the efficacy and safety of S201086/GLPG1972 in patients with knee OA. Methods: ROCCELLA was a randomized, placebo-controlled, dose-ranging, phase 2 study comprising a 5-week screening period, a 52-week double-blind treatment period and a 2-week safety follow-up period (ClinicalTrials.gov ID: NCT03595618). Patients aged 40-75 years with knee OA and pain severity in the target knee of 40-90 mm on a visual analog scale at screening and baseline were included. Target knees had predominant medial compartment disease, with Kellgren-Lawrence (KL) grade 2 or 3 and OARSI medial joint space narrowing (JSN) grade 1 or 2. Patients were randomized 1:1:1:1 to placebo or 75 mg, 150 mg or 300 mg S201086/GLPG1972 administered orally once daily. Concomitant analgesics (non-steroidal anti-inflammatory drugs and acetaminophen) were permitted. The primary endpoint was change from baseline to week 52 in cartilage thickness of the central medial femorotibial compartment (cMFTC) of the target knee, as measured by quantitative magnetic resonance imaging (qMRI) and analyzed by a central reading facility. Secondary efficacy endpoints included: change from baseline to week 52 in radiographic joint space width of the target knee (JSW; X-ray with central readout); patient-reported outcomes (including WOMAC scores, patient global assessment [PGA] score and pain score, both measured by visual analog scales); and safety outcomes. A mixed-effects model for repeated measures (using all longitudinal observations at each post-baseline visit) was used for the primary analysis. Results: Across 12 countries, 3319 patients were screened and 932 were included in the study. Patients had a mean age of 62.9 years and the majority (69.3%) were women. Baseline characteristics were similar across study groups (Table 1). Overall, 88.8% of knees were KL grade 3 and 67.3% were OARSI medial JSN grade 2. Patients experienced substantial cartilage loss; the mean (SD) change in cMFTC cartilage thickness was −0.12 (0.27) mm from baseline to week 52 in the placebo group. However, no statistically significant differences between treatment groups and placebo were observed for the primary endpoint of cMFTC cartilage thickness loss (placebo vs 75 mg, p = 0.165; vs 150 mg, p = 0.939; vs 300 mg, p = 0.682; Figure 1). These results were confirmed by sensitivity analyses assessing the management of missing data and delayed week 52 qMRI owing to the COVID-19 pandemic. No significant differences between treatment groups and the placebo group were observed at any time point in any of the secondary endpoints, including changes in WOMAC total score and subscores (Figure 2), radiographic JSW, PGA score or pain score. Treatment-emergent adverse events (TEAEs) and serious adverse events were experienced by similar proportions of patients in the placebo and S201086/GLPG1972 groups (Table 2). The most common TEAEs across all S201086/GLPG1972 treatment groups were arthralgia, nasopharyngitis and fall (Table 2). Tolerability was similar across the three S201086/GLPG1972 dose groups, with 6.8% (75 mg) to 8.6% (300 mg) of patients withdrawing from treatment owing to TEAEs, compared with 3.8% in the placebo group. Conclusions: The study successfully selected patients who experienced a substantial decrease in cartilage thickness over 52 weeks; a decrease that would be large enough to demonstrate a sizeable structural benefit of a DMOAD candidate. However, the study failed to meet its primary endpoint (change from baseline to week 52 in cartilage thickness of the cMFTC) and secondary endpoints, and there was no dose-response relationship. S201086/GLPG1972 had a good safety profile and was generally well tolerated.Figure 2Change over time in WOMAC subscores. Data show mean and standard error. W, week.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table 1Baseline patient demographics and characteristics.Demographic/characteristicPlacebo (n = 234)S201086/GLPG197275 mg (n = 234)S201086/GLPG1972150 mg (n = 231)S201086/GLPG1972300 mg (n = 232)Age (years), mean (SD)63.3 (7.1)62.9 (7.5)63.2 (7.2)62.1 (7.4)Age (years), n (%)-40-5528 (12.0)31 (13.2)29 (12.6)33 (14.2)55-6595 (40.6)94 (40.2)96 (41.6)97 (41.6)≥ 65111 (47.4)109 (46.6)106 (45.9)103 (44.2)Women, n (%)163 (69.7)164 (70.1)165 (71.4)154 (66.1)Race, n (%)White171 (73.1)167 (71.4)177 (76.6)168 (72.1)Black or African American25 (10.7)27 (11.5)19 (8.2)25 (10.7)Asian32 (13.7)31 (13.2)28 (12.1)30 (12.9)American Indian, native Alaskan, or native Hawaiian or other Pacific Islander01 (0.4)1 (0.4)2 (0.9)Multiple6 (2.6)8 (3.4)6 (2.6)8 (3.4)Time since first diagnosis (years), mean (SD)7.3 (6.7)6.9 (6.4)7.6 (7.4)7.1 (7.2)KL grade, n (%)229 (12.4)15 (6.4)30 (13.0)29 (12.4)3205 (87.6)219 (93.6)200 (86.6)204 (87.6)4001 (0.4)0OARSI atlas JSN grade, n (%)001 (0.4)1 (0.4)0170 (29.9)74 (31.6)73 (31.6)84 (36.1)2164 (70.1)159 (67.9)156 (67.5)148 (63.5)3001 (0.4)1 (0.4)Cartilage thickness in cMTFC, mean (SD)3.19 (0.82)3.25 (0.76)3.23 (0.76)3.33 (0.80)Joint space width, mean (SD)2.48 (0.86)2.5 (0.78)2.5 (0.78)2.58 (0.84)cMTFC, central medial femorotibial compartment; JSN, joint space narrowing; KL, Kellgren-Lawrence; SD, standard deviation. Open table in a new tab Table 2Summary of safety outcomes.Events, n (%)Placebo (n = 234)S201086/GLPG197275 mg (n = 234)S201086/GLPG1972150 mg (n = 231)S201086/GLPG1972300 mg (n = 232)TEAEs174 (74.4)174 (74.4)177 (76.6)174 (75.0)Severe TEAEs29 (12.4)25 (10.7)27 (11.7)30 (12.9)Treatment-related TEAEs37 (15.8)36 (15.4)30 (13.0)47 (20.3)Serious TEAEs18 (7.7)17 (7.3)17 (7.4)18 (7.8)Serious treatment-related TEAEs2 (0.9)02 (0.9)1 (0.4)TEAEs leading to drug withdrawal9 (3.8)16 (6.8)17 (7.4)20 (8.6)TEAEs occurring in ≥ 5% of patientsArthralgia19 (8.1)27 (11.535 (15.2)26 (11.2)Nasopharyngitis20 (8.5)21 (9.0)16 (6.9)22 (9.5)Fall13 (5.6)15 (6.4)20 (8.7)16 (6.9)Back pain19 (8.1)11 (4.7)10 (4.3)7 (3.0)Headache9 (3.8)15 (6.4)12 (5.2)11 (4.7)Hypertension16 (6.8)6 (2.6)9 (3.9)12 (5.2)Osteoarthritis10 (4.3)8 (3.4)12 (5.2)11 (4.7)Increased blood creatine phosphokinase8 (3.4)12 (5.1)7 (3.0)9 (3.9)Upper respiratory tract infection10 (4.3)7 (3.0)12 (5.2)6 (2.6)Increased gamma-glutamyltransferase4 (1.7)3 (1.3)2 (0.9)16 (6.9)TEAE, treatment-emergent adverse event. Open table in a new tab cMTFC, central medial femorotibial compartment; JSN, joint space narrowing; KL, Kellgren-Lawrence; SD, standard deviation. TEAE, treatment-emergent adverse event.

Background: Osteoarthritis (OA) is a degenerative joint disease involving structural pathology of all joint tissues, and most commonly affecting the knee, hip and hand. Degradation of the cartilage extracellular matrix represents a central feature of OA and is widely thought to be mediated by proteinases that degrade primarily aggrecan and collagen. ADAMTS-5, a Disintegrin And Metalloproteinase with ThromboSpondin-motif-5, is a key aggrecan-cleaving enzyme involved in cartilage degradation. S201086/GLPG1972, a potent and highly selective inhibitor of ADAMTS-5, is an oral Disease-Modifying OsteoArthritis Drug (DMOAD) candidate. Objectives: The primary objective of the ROCCELLA phase 2 clinical trial ( NCT03595618 ) is to evaluate the effect of S201086/GLPG1972 over 52 weeks of treatment (3 dose groups compared to placebo) in reducing cartilage loss. Cartilage thickness of the knee is being measured quantitatively by Magnetic Resonance Imaging. Here, we describe the baseline characteristics of patients included in the ROCCELLA clinical trial. Methods: The main inclusion criteria were: male or female, aged 40 to 75, with a diagnosis of knee OA according to the clinical and radiological criteria of the American College of Rheumatology. The target knee had to meet a pain score between 40 and 90 mm on a 100 mm Visual Analog Scale (VAS), and the following radiographic feature upon central radiographic readings: Kellgren/Lawrence (KL) 2 or 3 and OARSI medial joint space narrowing (JSN) 1 or 2 (for more details see Deckx et al . OARSI 2020). The rationale for these specific radiographic inclusion criteria was to ensure sufficient cartilage loss over 12 months to assess the efficacy of S201086/GLPG1972. Results: Across 12 countries, 3319 patients were screened and 932 were finally included in the study. The screen failure of 72% is mainly due to the radiological criteria. The age of the patients was 62.9 ± 7.3 years (mean ± SD) with a majority of women (69.3%). The BMI was 30.5 ± 4.7 kg/m 2 . The duration of knee OA was 7.2 ± 6.9 years. Five hundred and one (53.8%) patients reported a medical history of musculoskeletal and connective tissue disorders, mainly osteoarthritis in other sites (20.2%), back pain (13.6%), and arthralgia (9.8%). At inclusion, 97.2% of the patients were taking different types of drug treatments, mainly anti-inflammatory and anti-rheumatic products (69.4%) and analgesics (42%). At baseline, 11% of the target knees were KL2 and 89% were KL3; 32% were OARSI medial JSN grade 1 and 68% grade 2. Target knees at inclusion had a pain score on the VAS of 63.5 ± 11.4 mm (range 0-100, with 0 for no and 100 for extreme pain) and a total WOMAC (Likert 3.1) score of 48.0 ± 15.0 (range 0-96). The WOMAC subscores for pain, stiffness and physical function were 10.0 ± 3.2 (range 0-20), 4.2 ± 1.6 (range 0-8) and 33.8 ± 11.2 (range 0-68, indicating functional limitation), respectively. Conclusion: For this clinical trial, patients were selected to present radiological criteria ( i.e. OARSI JSN 1 and 2) to ensure sufficient structural progression (cartilage loss) over 12 months, as well as clinical symptoms. These stringent selection criteria were the main cause for the high screen failure rate. These baseline characteristics should warrant the ability to evaluate the efficacy of S201086/GLPG1972 as a DMOAD candidate. The search for an effective pharmacological treatment that can prevent or cure OA remains a major challenge and unmet medical need. Disclosure of Interests: Katy Bernard Employee of: Institut de Recherches Internationales Servier, Sergey GRANKOV Employee of: Institut de Recherches Internationales Servier, Marjolijne van der Stoep Employee of: Galapagos, Agnès Lalande Employee of: Institut de Recherches Internationales Servier, Olivier Imbert Employee of: Institut de Recherches Internationales Servier, De Phung Employee of: Galapagos, Damien Chimits Employee of: Institut de Recherches Internationales Servier, Karine Muller Employee of: Galapagos, Ellen van der Aar Employee of: Galapagos, Henri Deckx Employee of: Galapagos, Maria Pueyo Employee of: Institut de Recherches Internationales Servier, Felix Eckstein Grant/research support from: Merck, Orthotrphix, Servier, Galapagos, Kolon Tissuegene, Samumed, Novartis, Consultant of: Merck, Bioclinica, Servier, Samumed, Roche, Kolon Tissuegene, Galapagos and Novartis, Employee of: co-owner and employment with Chondrometrics

Purpose: Knee osteoarthritis (OA) is a chronic disease with slow evolution. Kellgren and Lawrence (KL) classification is used to identify the radiographic stages of the disease, but has limited predictive validity of structural disease progression. We hypothesized that, in combination with KL, Osteoarthritis Research Society International (OARSI) Joint Space Narrowing (JSN) grades, permit a superior identification of structural progressors, i.e. patients with subsequent cartilage thickness loss. The objective of the current study therefore was to explore to what extent the combination of KL and JSN is predictive of which knee OA patients are more likely to display structural progression. Methods: For this purpose, the OsteoArthritis Initiative (OAI) public database and cartilage thickness measurements contained therein was used (Chondrometrics GmbH, Ainring, Germany). The data studied included 2782 knees of the progression cohort. Two analysis populations were defined: the KL population consisting of 1140 KL2 or KL3 target knees, and the “KL+JSN” population consisting of 796 KL2 or KL3 target knees with medial OARSI JSN 1 or 2. KL and JSN grades were taken from version v0.8 of the central radiographic readings performed at Boston University. Cartilage loss over one year had been assessed from the difference between baseline and 12 months follow-up values in cartilage thickness as measured in the central medial tibiofemoral compartment (cMTFC) on the both populations. Results: The percentages of KL patients with various JSN grade strata are provided for both populations (Table 1). Descriptive measures of cartilage loss in the various radiographic strata are given in Table 2. cMTFC cartilage loss over one year was −0.134 ± 0.29mm (median= −0.088mm) in the KL population and −0.174 ± 0.29mm (median= −0.135mm) in the KL+JSN population. In the KL population, knees with a JSN0 scoring did not show any cartilage loss (+0,047 ±0,25mm). The mean cartilage loss was −0.074 ± 0.26mm (median = −0.043mm) in KL2 patients, of whom 48% were medial OARSI JSN grade 0. The mean cartilage loss was −0.180 ± 0.30mm (median = −0.138mm) in KL3 patients, of whom 22% were medial OARSI JSN grade 0. Please note that all these KL3 knee had lateral JSN (grade 2). No patients with a medial OARSI JSN score 0 were included in the KL+JSN population. In this population, the mean cartilage loss was −0.084 ± 0.24mm (median = −0.060mm) in KL2 patients and −0.224 ± 0.31mm (median = −0.175mm) in KL3 patients. Conclusions: In addition to the KL classification, the OARSI JSN score contributes to identify structural progressors knees as long as the laterality of the OARSI JSN scoring is chosen in accordance with the cartilage loss in the compartment of interest. These results obtained in the OAI database suggested that the OARSI JSN score as inclusion criteria is useful in clinical trials testing the efficacy of disease-modifying drugs in a reasonable time window of 1-year. In addition, our modelisation provides a useful tool to predict the one year cartilage loss according to the percentage of patients with KL2 or KL3 scoring (resp. patients with JSN1 or JSN2 scoring) in the population at Baseline.Tabled 1Table 1. Repartition of KL2 and KL3 patients according to medial OARSI JSN scoreKL population (N=1140)KL population (N=1140)KL population (N=1140)KL population (N=1140)KL+JSN population (N=796)KL+JSN population (N=796)KL+JSN population (N=796)JSN0JSN1JSN2ALLJSN1JSN2ALLKL247,8%52,2%0,0%100%99,7%0,3%100%KL322,1%1,8%76,1%100%2,0%98,0%100% Open table in a new tab Tabled 1Table 2. Cartilage loss (mm) in cMTFC at one year by KL and medial JSN scoreCartilage loss at one year (cMTFC)KL population (N=1140)KL population (N=1140)KL population (N=1140)KL+JSN population (N=796)KL+JSN population (N=796)KL+JSN population (N=796)NMedianMean (sd)NMedianMean (sd)All patients587-0,088-0,134 (0,29)412-0,135-0,174 (0,29)On KL2 patients251-0,043-0,074 (0,26)149-0,060-0,084 (0,24)On KL3 patients336-0,138-0,180 (0,30)263-0,175-0,224 (0,31)On OARSI JSN 0 patients192-0,0380,047 (0,25)---On OARSI JSN1 patients137-0,050-0,081 (0,24)152-0,053-0,080 (0,24)On OARSI JSN2 patients258-0,178-0,228 (0,31)260-0,180-0,228 (0,31) Open table in a new tab