MARKERS USEFUL IN ENRICHMENT STRATEGIES FOR THE TREATMENT OF OSTEOARTHRITIS

Abstract

The present invention relates to pharmacogenetics, more specifically to strategies involving biomarkers associated with the clinical response to a compound before or during treatment of a cartilage disorder, such as osteoarthritis. The present invention more particularly relates to the combination of JSW measurements and level of specific proteins present in the blood, serum, synovial fluid or in the urine, which can be used in strategies such as patients' enrichment in clinical trials, patients' selection strategy before or during treatment or for adapting the treatment of a patient in the frame of treatments for cartilage disorder, such as osteoarthritis.

Claims

1-16. (canceled)

17. A method of derisking a clinical trial, wherein said clinical trial is related to the treatment of a cartilage disorder in a subject with an active compound, the method comprising the steps of: a) determining, from a sample from said subject, the quantity of at least one of the biomarkers selected from the group consisting of CTX-II and/or ProC2, b) measuring the joint space width (JSW) in at least one knee of said subject, and c) predicting from the result of steps a) and b) high, intermediate, low or no sensitivity of said subject to treatment with said active compound, wherein steps a) and b) are performed in either order.

18. The method according to claim 17, comprising the steps of predicting low or no sensitivity to treatment with an active compound from the presence of higher than 350±2SD ng/mmol of CTX-II, or higher than 4.2±2SD ng/mL of ProC2, and JSW higher than 3.5±2SD mm.

19. The method according to claim 17, comprising the steps of predicting: i) good sensitivity to treatment with an active compound from the presence of lower than 350±2SD ng/mmol of CTX-II and/or, lower than 4.2±2SD ng/mL of ProC2, and JSW lower than or equal to 3.5±2SD mm, or ii) intermediate sensitivity to treatment with an active compound from the presence of: a) lower than 350±2SD ng/mmol of CTX-II and/or, lower than 4.2±2SD ng/mL of ProC2, and JSW higher than 3.5±2SD mm, or b) higher than 350±2SD ng/mmol of CTX-II and/or higher than 4.2±2SD ng/mL of ProC2 and of a JSW lower than 3.5±2SD mm.

20. A method for selecting a subject having a cartilage disorder for inclusion in or exclusion from treatment, or clinical trial, with an active compound, based on the likelihood of their sensitivity to said treatment, comprising the steps of: a) determining, from a sample from said subject, the quantity of at least one of the biomarkers selected from the group consisting of CTX-II and/or ProC2 and measuring the joint space width (JSW) in at least one knee of said subject; wherein the quantity of at least one of these proteins and the width if the JSW are predictive about the subject's risk for being sensitive or not-sensitive to said treatment, and b) selecting the sensitive subjects as being suitable for said treatment or clinical trial.

21. The method according to claim 20, comprising the steps of excluding from the treatment with an active compound the subject presenting higher than 350±2SD ng/mmol of CTX-II, and/or higher than 4.2±2SD ng/mL of ProC2 and JSW higher than 3.5±2SD mm.

22. The method according to claim 21, comprising the steps of including for the treatment with an active compound the subject presenting: a) lower than 350±2SD ng/mmol of CTX-II, and/or lower than 4.2±2SD ng/mL of ProC2 and JSW lower than or equal to 3.5±2SD mm, b) lower than 350±2SD ng/mmol of CTX-II, and/or lower than 4.2±2SD ng/mL of ProC2 and JSW higher than 3.5±2SD mm, or c) higher than 350±2SD ng/mmol of CTX-II and/or higher than 4.2±2SD ng/mL of ProC2 and of a JSW lower than 3.5±2SD mm.

23. The method according to claim 22, comprising the steps of including for the treatment with an active compound, according to an alternative dosing regimen, the subject presenting: a) lower than 350±2SD ng/mmol of CTX-II, and/or lower than 4.2±2SD ng/mL of ProC2 and JSW higher than 3.5±2SD mm, or b) higher than 350±2SD ng/mmol of CTX-II and/or higher than 4.2±2SD ng/mL of ProC2 and of a JSW lower than 3.5±2SD mm.

24. A method for treating a subject with an active compound having a cartilage disorder, comprising the following steps: a) determining, from a sample from said subject, the quantity of at least one of the biomarkers selected from the group consisting of CTX-II and/or ProC2 and measuring the joint space width (JSW) in at least one knee of said subject, wherein said quantities and widths are predictive about the subject's risk for being good sensitive to a treatment with said active compound, b) selecting the subject having: i) lower than 350±2SD ng/mmol of CTX-II, and/or lower than 4.2±2SD ng/mL of ProC2 and JSW lower than or equal to 3.5±2SD mm, ii) lower than 350±2SD ng/mmol of CTX-II and/or lower than 4.2±2SD ng/mL of ProC2 and JSW higher than 3.5±2SD mm, or iii) higher than 350±2SD ng/mmol of CTX-II and/or higher than 4.2±2SD ng/mL of ProC2 and JSW lower than 3.5±2SD mm, and c) administering intraarticularly said active compound to said selected subject.

25. The method according to claim 24, wherein the active compound is selected from the group consisting of FGF-18 compound, BMP-2, BMP-7, GDF-5, FGFβ, FGF-9, SOX-9 enhancers, TGFβ, Wnt inhibitors, anti-MMP13 inhibitors, anti-ADAMTS4 or 5 inhibitors, calcitonin and any variants or fusion proteins thereof.

26. The method according to claim 24, wherein the cartilage disorder is selected from the group consisting of osteoarthritis, cartilage injury, fractures affecting joint cartilage or surgical procedures with impact on joint cartilage.

27. A method for treating a subject with an active compound having a cartilage disorder, comprising the following steps: a) determining, from a sample from said subject, the quantity of at least one of the biomarkers selected from the group consisting of CTX-II and/or ProC2 and measuring the joint space width (JSW) in at least one knee of said subject, wherein said quantities and widths are predictive about the subject's risk for being intermediate sensitive to a treatment with said active compound, b) selecting the subject having: i) lower than 350±2SD ng/mmol of CTX-II, and/or lower than 4.2±2SD ng/mL of ProC2 and JSW higher than 3.5±2SD mm, or ii) higher than 350±2SD ng/mmol of CTX-II and/or higher than 4.2±2SD ng/mL of ProC2 and JSW lower than 3.5±2SD mm, and c) administering intraarticularly said active compound to said selected subject according to an alternative dosing regimen according to which the dose of the active compound is increased, and/or the active compound is administered for a longer treatment period compared to the usual dosing regimen for said active compound.

28. The method according to claim 27, wherein the active compound is selected from the group consisting of FGF-18 compound, BMP-2, BMP-7, GDF-5, FGFβ, FGF-9, SOX-9 enhancers, TGFβ, Wnt inhibitors, anti-MMP13 inhibitors, anti-ADAMTS4 or 5 inhibitors, calcitonin and any variants or fusion proteins thereof.

29. The method according to claim 27, wherein the cartilage disorder is selected from the group consisting of osteoarthritis, cartilage injury, fractures affecting joint cartilage or surgical procedures with impact on joint cartilage.

30. A method of derisking a clinical trial or of determining placebo effect in a subject in a clinical trial, wherein said clinical trial is related to the treatment of a cartilage disorder in a subject with an active compound, the method comprising the steps of: a) measuring the joint space width (JSW) in at least one knee of said subject, and b) predicting from the result of step a) the risk of placebo effect said subject, wherein the presence of a JSW higher than 3.5±2SD mm is predictive of placebo effect and the presence of a JSW lower than or equal to 3.5±2SD mm is predictive of no or low placebo effect.

31. A method of derisking a clinical trial or of determining placebo effect in a subject in a clinical trial, wherein said clinical trial is related to the treatment of a cartilage disorder in a subject with an active compound, the method comprising the steps of: a) determining, from a sample from said subject, the quantity of at least one of the biomarkers selected from the group consisting of CTX-II and/or ProC2, b) measuring the joint space width (JSW) in at least one knee of said subject, and c) predicting from the result of steps a) and b) the risk of placebo effect for said subject, wherein steps a) and b) are performed in either order.

32. The method according to claim 31, wherein the presence of 1) higher than 350±2SD ng/mmol of CTX-II (>240-260% of normal mean) and/or higher than 4.2±2SD ng/mL of ProC2 (>120-280% of normal mean) and a JSW higher than 3.5±2SD mm or 2) lower than 350±2SD ng/mmol of CTX-II and/or lower than 4.2±2SD ng/mL of ProC2 and JSW higher than 3.5±2SD mm or 3) higher than 350±2SD ng/mmol of CTX-II and/or higher than 4.2±2SD ng/mL of ProC2 and JSW lower than or equal to 3.5±2SD mm is predictive of placebo effect, whereas the presence of lower than 350±2SD ng/mmol of CTX-II and/or lower than 4.2±2SD ng/mL of ProC2 and JSW lower than or equal to 3.5±2SD mm is predictive of no or low placebo effect.

Description

DESCRIPTION OF THE FIGURES

[0171] FIG. 1: Scheme of the dosing regimens used for FGF-18 compound in the FORWARD study.

[0172] FIG. 2: Proposed enrichment strategy from FDA draft guidance “Enrichment Strategies for Clinical Trials to Support Approval of Human Drugs and Biological Products” (FDA, 2012).

[0173] FIG. 3: Placebo responses in KL2 and KL3 patients (i.e. placebo effect as a function of the JSW) underlined via total WOMAC score. A) Total WOMAC change from baseline in KL2 and KL3 groups as a function of the elapsed time after first IA injection group means as a function of time elapsed (data are means +/−SD). B) Total WOMAC score in KL2 and KL3 groups as a function of the elapsed time after first IA injection group means as a function of time elapsed (data are medians).

[0174] FIG. 4: Placebo responses as a function of the biomarkers level underlined via total WOMAC score. A) Total WOMAC change in group as a function of elapsed time, based on the CTX-II level (group 1: ≤350 ng/mmol and group 2: >350 ng/mmol) (data are means +/−SD). B) Total WOMAC change in group as a function of elapsed time, based on the ProC2 level (group 1: ≤4.2 ng/mmol and group 2: >4.2 ng/mmol) (data are means +/−SD).

[0175] FIG. 5: Placebo responses combining both JSW measurments) and CTX-II level underlined via total WOMAC score in group means as a function of time elapsed (months). The data is retrieved from the placebo arm of the FORWARD study.

[0176] FIG. 6: A) Difference of active arms to the placebo arm as a function of the KL grade of the patients and as a function of elapsed time after first IA injection, demonstrated with Total WOMAC score. B) Difference of active arms to the placebo arm as a function of the CTX-II level of the patients and as a function of elapsed time after first IA injection, demonstrated with Total WOMAC score. C) Difference of active arms to the placebo arm as a function of the KL grade of the patients and as a function of elapsed time after first IA injection, demonstrated with cartilage thickness level. B) Difference of active arms to the placebo arm as a function of the CTX-II level of the patients and as a function of elapsed time after first IA injection, demonstrated with cartilage thickness level. In FIGS. 6A-D, each group of treated patients (groups 2 to 5) from the FORWARD study were compared to the placebo group (group 1) (comparison based on the difference to the placebo). Data are group medians.

[0177] FIG. 7: Placebo responses combining both JSW (expressed as KL2 and KL3) and CTX-II level underlined via change in WOMAC pain score in group mean (SEM) as a function of time elapsed (months). The number of patients in each subgroup are as follows, n≈380 (KL2 with low CTX-II), n≈90 (KL2 with high CTX-II), n≈40 (KL3 with low CTX-II) and n≈20 (KL3 with high CTX-II).

[0178] FIG. 8: Proposed enrichment strategy for any clinical trials in Osteoarthritis, based on the FDA draft guidance of 2012.

DESCRIPTION OF THE SEQUENCES

[0179] SEQ ID NO.1: Amino acid sequence of the biomarker CTX-II.

[0180] SEQ ID NO.2: Amino acid sequence of the biomarker PROC2.

[0181] SEQ ID NO.3: Amino acid sequence of the native human FGF-18.

[0182] SEQ ID NO.4: Amino acid sequence of the recombinant truncated FGF-18 (trFGF-18).

[0183] SEQ ID NO.5: Amino acid sequence of the salmon calcitonin.

[0184] SEQ ID NO.6: Amino acid sequence of the human BMP-2

[0185] SEQ ID NO.7: Amino acid sequence of the human BMP-7

[0186] SEQ ID NO.8: Amino acid sequence of the human GDF-5.

[0187] SEQ ID NO.9: Amino acid sequence of the human FGFβ.

[0188] SEQ ID NO.10: Amino acid sequence of the human FGF-9.

EXAMPLES

Statistical Methods

[0189] The treatment effect on the primary endpoint was assessed through dose-ranging using a repeated measurement analysis of variance (ANOVA, using PROC MIXED in SAS) on absolute change from Baseline, including the baseline value, the treatment group, the time, and the country as factors and treatment-by-time point as interaction. The primary efficacy analysis consisted of testing the linear dose relationship and the overall treatment effect at 2 years. The significance level was set at 5% 2-sided for both tests. Pairwise comparisons (sprifermin versus placebo, and between sprifermin dose and regimen groups) were performed within the context of this modelling framework. For each pairwise comparison, the difference between treatments and the corresponding 95% confidence interval (CI) and p-value are presented. The same ANOVA model used for the primary endpoint was used to assess the treatment effect on continuous secondary endpoints such as MRI endpoints, WOMAC endpoints (total, pain, function, and stiffness scores), and X-ray endpoints at each time point and over time. Logistic regression was used to assess the treatment effect on the binary efficacy endpoints such as the OMERACT-OARSI responder rate. Point estimates for each pairwise comparison and corresponding 95% Cls and p-values are provided.

Pain and Function Assessments

[0190] The WOMAC is a validated instrument used to assess symptom modification in clinical OA studies. This clinical score was developed in 1981 and is regarded as a valid instrument by both clinical researchers and regulatory authorities. The WOMAC is widely used in clinical studies in hip and knee OA and has been extensively validated.

[0191] Subjects had to answer all of the 24 questions themselves (i.e. 5 for pain, 2 for stiffness and 17 for physical function assessment), using either the 11-box NRS assessment (with categories of 0 to 10) with reference to the past 48 hours for example 1 or 100 mm VAS (visual analogue scales; giving each question a score from 0 to 100) with reference to the past 24 hours for example 2. Different forms of the questionnaire exist for the right and the left knees: in order to reduce confounding of WOMAC responses by symptoms in the contralateral knee, subjects used the WOMAC questionnaire specific to the target knee.

[0192] For administration of the questionnaire, instructions for the WOMAC 3.1 Index were followed for both examples 1 and 2.

[0193] Other instruments for assessment of pain and function are the KOOS (Knee injury and Osteoarthritis Outcome Score, Collins et al. 2016).

X-Ray Assessment of JSW

[0194] Change in JSW as measured by X-ray is a recognized endpoint accepted by the European Medicines Agency and the United States Food and Drug Administration for use in efficacy studies in OA. The JSW was measured using standardized technique. X-ray was also used to assess KL grade in example 2.

qMRI Assessment

[0195] The primary endpoint for the DBPC treatment phase was the change from Baseline in cartilage thickness in the total femorotibial joint as evaluated by qMRI at 2 years in the mITT. Cartilage thickness of the total femorotibial joint were calculated in 2 ways: [0196] 1. Average Cartilage Thickness (Total Volume divided by Total Surface Area), [0197] 2. Total Cartilage Thickness (sum of cartilage thickness in medial and lateral compartment).

[0198] The treatment effect on the primary endpoint was assessed through dose-ranging using a repeated measurement analysis of variance (ANOVA) on absolute change from Baseline, including the treatment group, the time point, and the (pooled) country as fixed factors and the baseline value as covariate and treatment by time point as interaction. Repeated measures over time were accounted for using an “unstructured” covariance pattern.

[0199] Pairwise comparisons of absolute change from Baseline in cartilage thickness (treatment with compound groups versus placebo) were performed within the context of the modelling framework described above. For each pairwise comparison, the difference between treatments and the corresponding 95% confidence interval (CI) and p-value are presented. P-values (corresponding to Type 3 tests of fixed effects) are reported for all covariates in the original “Overall” model for all time points combined (i.e., baseline value, treatment, time point, treatment-by-time point interaction, country) and for all time points. Estimated coefficients, p-values, and 95% Cls are presented overall and at each time point for (i) the dose relationship (linear trend) and (ii) each pairwise comparison between dose level and placebo.

[0200] In order to assess the robustness of the primary results, the tests for linear dose-relationship and for the overall treatment effect were repeated using the PP Analysis Set. For the mITT Analysis Set, a non-parametric analysis was conducted for the ordered data of cartilage thickness in the total femorotibial joint as an alternative method for the primary analysis. Data were ordered by the magnitude of absolute change-from-Baseline over 2 years during DBPC treatment phase using rank transformation.

Biomarkers Measurement

[0201] Serological and urine biochemical markers of bone and joint tissue turnover as well as synovial inflammation were evaluated. Potential biomarkers of cartilage metabolism included, but were not limited to: neo-epitope of collagen type II propeptide (proC2) and C-telopeptide cross-linking of type II collagen (CTX-II). Blood and urine samples for systemic biomarker assessment were collected at the following time points: week 0 (before first injection of sprifermin), week 26, week 54, week 80 and week 104. For time points where injections were also administered, samples were collected before injection. Synovial fluid samples were collected at the time points. These samples were taken just before injection, as part of the i.art. injection procedure and using the same needle that the one used for the injection. For urine collection, second morning void samples were obtained.

[0202] The following assessment were made as exploratory endpoints: [0203] Change from Baseline in serum and urine biomarkers associated with administration of the compound. [0204] Baseline protein markers associated with response to treatment or disease progression (response assessed by MRI and/or questionnaire).

Example 1. Placebo Response and of the Clinical Efficacy in Subjects Treated with an FGF-18 Compound

[0205] The FGF-18 compound used as a treatment in the present examples is sprifermin (as defined in the section “definitions”). Two strengths of sprifermin were supplied for the study: 30 μg and 100 μg. Sprifermin was supplied as a white, sterile, freeze-dried powder in 3-mL glass vials. Each vial contained either 31.5 μg or 105 μg of sprifermin active substance; these quantities included a 5% overage, permitting extraction of respectively 30 μg or 100 μg of sprifermin active substance following reconstitution with 0.9% w/v Sodium Chloride Injection (referred to herein as “saline solution”). Excipients of the formulation were sodium phosphate buffer (pH 7.2), sodium hydroxide, O-phosphoric acid, sucrose, and poloxamer 188. For all treatment groups, the volume administered was 2 mL.

[0206] The present study was based on the FORWARD study (see study EMR700692-006) where five groups of patients were studied: [0207] Group 1 (4 cycles placebo; hereafter referred to as placebo or PBO): 108 subjects. [0208] Group 2 (2 cycles sprifermin 30 μg/injection alternating with 2 cycles placebo; hereafter referred to as sprifermin/placebo 30 μg): 110 subjects. [0209] Group 3 (4 cycles sprifermin 30 μg/injection; hereafter referred to as sprifermin 30 μg): 111 subjects. [0210] Group 4 (2 cycles sprifermin 100 μg/injection alternating with 2 cycles of placebo; hereafter referred to as sprifermin/placebo 100 μg): 110 subjects. [0211] Group 5 (4 cycles sprifermin 100 μg/injection; hereafter referred to as sprifermin 100 μg): 110 subjects.

[0212] According to the FORWARD study, the patients received 4 cycles of treatment (each consisting of 3 once-weekly intra articular injections over 3 consecutive weeks) at intervals of 6 months (see FIG. 1). All injections were intraarticular (done intraarticularly).

[0213] The primary efficacy endpoint was the change from Baseline in cartilage thickness in the total femorotibial joint as evaluated by MRI at 2 years. Exploratory endpoints included Baseline protein markers associated with response to treatment or disease progression (response assessed by MRI and/or questionnaire).

[0214] The study enrolled adult subjects of either sex with primary femorotibial OA according to American College of Rheumatology (ACR) clinical and radiographic criteria who had Kellgren-Lawrence grades (KLG) of 2 or 3 and a minimum joint space width (JSW) of 2.5 mm in the medial compartment. Subjects must have had pain in the target knee on most days and/or require symptomatic treatment of knee pain with paracetamol (acetaminophen), systemic non-steroidal anti-inflammatory drugs (NSAIDs) including COX inhibitors (COXibs), or tramadol on most days of the previous month, and must have had both: 1) A history of pain due to OA in the target knee for at least 6 months, and 2) Pain score for the target knee of 4 to 9 points in response to Question 1 of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain index (“how much pain have you had [in the target knee, over the past 48 hours] when walking on a flat surface?”) at screening and Baseline, after washout of at least 5 half-lives of analgesic medication(s): acetaminophen, topical or oral systemic NSAIDS, COXibs, opioids, and/or tramadol. Women of childbearing potential must have used a form of contraception with a failure rate of less than 1% per year throughout the study.

[0215] Main exclusion criteria included malalignment of >5 degrees in the femorotibial axis of the target knee, clinical signs of inflammation (i.e. redness) in the target knee, i.art. administration of corticosteroids or hyaluronic acid into either knee within 6 months before screening, any plan for knee surgery (affecting either the target or the contralateral knee) within the next 2 years, concomitant conditions or treatments deemed to be incompatible with study participation, contraindications to MRI scanning (including inability to fit in the scanner or knee coil), pregnancy or breastfeeding, participation in another clinical study within the past 30 days, and legal incapacity or limited legal capacity.

[0216] Written informed consent must have been obtained prior to any study-related activity.

[0217] Placebo effect based on JSW alone: As shown in FIG. 3 (both A and B), the placebo effect in KL2 patients (having a JSW >3.5 mm) versus KL3 patients (having a JSW ≤3.5 mm) is very important. As self-evident, it plays an important role on the total WOMAC score, where the simple injection of a placebo (PBO) (i.e. a salt) is enough to reduce drastically over a very long time the WOMAC score in KL2 patients, whereas said WOMAC score increase continuously in KL3 patients, despite the injection.

[0218] Placebo effect based on CTX-II or ProC2 levels: FIG. 4 underlines that the patients having a high metabolic biomarker level such as CTX-II above 350 ng/mmol (FIG. 4A) or ProC2 above 4.2 ng/mL (FIG. 4B) present a higher placebo effect to injection of a placebo.

[0219] Placebo effect based on a combination JSW+CTX-II level: As shown in FIG. 5, different distribution of CTX-II levels at baseline influence outcome measures like MRI and WOMAC. In view of the results, thin cartilage (JSW ≤3.5 mm) and low level of metabolic biomarkers (e.g. CTX-II below or equal to 350 ng/mmol) seems the right study population for a clinical trial to mitigate elevated placebo responses.

[0220] After treatment with FGF18: As evident from FIGS. 6A and 6B, once the placebo effect is taken into consideration, the shape of response in term of total WOMAC score, for both KL2 patients and KL3 patients treated with various doses of sprifermin, change drastically. The same is true when the CTX-2 level are taken into consideration.

[0221] As evident from FIGS. 6C and 6D, once the placebo effect is taken into consideration, the shape of response in term of cartilage thickness also change deeply, for both KL2 patients and KL3 patients treated with various doses of sprifermin. The same is true when the CTX-2 level are taken into consideration. From this example it is clear that stratification can de-risk clinical trial which are otherwise biased by the placebo effect of some subgroups of patients.

Example 2. Placebo Response in Subjects Treated with Placebo from a Study with Compound X

[0222] The study was double blinded, randomized, placebo-controlled and multi-centre clinical trial for the assessment of compound X. There were about 600 patients enrolled in the placebo arm of this study. Patients aged 50-80 years with painful OA of at least one knee, but who were otherwise in good general health were recruited. Compound X being an oral drug, the placebo was supplied in identical packaging. To be included, patients had to meet the ACR criteria for diagnosis of OA. Both knees were assessed during the study, but a target knee was identified prior to randomization for assessment of the primary efficacy endpoint. The target knee had to be painful on most days of the prior month. In addition, the patient had to fulfil at least one of the following criteria: age over 50 years, experience morning stiffness lasting less than 30 min, or knee crepitus. The target knee had to have a JSW ≥2.0 mm at the medial tibio-femoral joint as measured on X-ray and have a KL score of 2 or 3. On WOMAC, patients were to score ≥150 mm for pain and/or ≥510 mm for function. If both knees fulfilled the criteria, the target knee was defined as the knee that meets KL score of 2. If both knees had KL scores of 2 the knee with the highest pain VAS was selected (as long as it was below 80 mm).

[0223] Exclusion criteria included diseases (apart from OA) and medications that affected bone or cartilage metabolism. Pain relief medication was allowed, provided it was taken at least 30 min after the study drug. Intra-articular injection of corticosteroids or hyaluronic acid in the target knee was prohibited during the study and 3 months prior to randomization.

[0224] Randomization was stratified by centre and it was ensured that treatment assignment was unbiased and concealed from patients and investigator staff. The trial was conducted in accordance with the Declaration of Helsinki. The protocol was approved by independent ethics committees or institutional review boards (IEC/IRB). All subjects or legal representatives gave their informed consent to take part.

[0225] Placebo effect based on a combination KL+CTX-II level: As shown in FIG. 7, different distribution of CTX-II levels in subgroups of KL2 and KL3s at baseline influence outcome measures like WOMAC pain. In view of the results, thin cartilage (JSW ≤3.5 mm) and low level of metabolic biomarkers (e.g. CTX-II below or equal to 350 ng/mmol) seems the right study population for a clinical trial to mitigate elevated placebo responses.

Overall Conclusion

[0226] The examples clearly underlined that KL3 enriched and CTX-II balanced design would enable detection of symptomatic benefit up to at least year 3 (see FIG. 8). Therefore, thin cartilage (below or equal to 3.5±2SD mm) and low CTX-II (below or equal to 350±2SD ng/mmol) seems the best study population for a clinical trial to mitigate elevated placebo responses. Of notes, it is obviously possible to use Allcomer data, i.e.no loss of patients: it is just a matter of stratification, of management of the patients. Similarly, based on the results of the example section, ProC2 level could be used instead of CTX-II level. In such a case, thin cartilage and low ProC2 (below or equal to 4.2±2SD ng/ml) would be an equally appropriate study population for a clinical trial to mitigate elevated placebo responses. Thanks to these findings, demonstrated for two very different clinical trial involving two very different active compounds, administered via different routes (one injected intraarticularly and the other administered orally), and to this proposed stratification, it is possible to decrease placebo responses, avoid misbalance between treatment groups and de-risk clinical trial for any drugs targeting OA due to possibility to select the right patient population with structural and symptomatic benefit.

REFERENCES

[0227] 1) WO2008/023063

[0228] 2) WO2004/032849

[0229] 3) WO2014/023703

[0230] 4) http://www.cartilage.org/_files/contentmanagement/ICRS_evaluation.pdf

[0231] 5) Lotz, 2010, Arthritis research therapy, 12:211

[0232] 6) Ellsworth et al., 2002, Osteoarthritis and Cartilage, 10:308-320

[0233] 7) Shimoaka et al., 2002, J. Bio. Chem. 277(9):7493-7500

[0234] 8) Gigout et al., 2017, Osteoarthritis and Cartilage, published online the 18.08.2018 (https://doi.org/10.1016/j.joca.2017.08.004)

[0235] 9) Lohmander et al., 2014, Arthritis Rheumatol., Jul;66(7):1820-31

[0236] 10) Dahlberg et al., 2016, Clin Exp Rheumatol., May-Jun;34(3):445-50

[0237] 11) Deng et al. 2018, Osteoarthritis and Cartilage, in press: 1-9

[0238] 12) Parrish et al., 2017, Osteoarthritis Cartilage., Apr;25(4):554-560

[0239] 13) Ameye and Young 2006. Curr Opin Rheumatol 18:537-547.

[0240] 14) Zhou et al., 2016, Osteoarthritis Cartilage., Dec;24(12):2181-2192

[0241] 15) Wang et al. 2014. Birth defects Res C embryo today 102:37-51

[0242] 16) Karsdal et al., 2016, Osteoarthritis and Cartilage, 24(12): 2013-2021

[0243] 17) Duclos et al., 2010, Osteoarthritis Cartilage. 2010 Nov;18(11):1467-76.

[0244] 18) Gudmann et al., 2016, Arthritis Res Ther., 18(1):141.

[0245] 19) Munk et al., 2016, Rheumatol Int. 36(4):541-9.

[0246] 20) Gudmann et al., 2014, Int J Mol Sci. ,15(10):18789-18803.

[0247] 21)Valdes et al., 2014, Osteoarthritis Cartilage., May;22(5):683-9

[0248] 22) Struglics et al., 2011, Osteoarthritis Cartilage., Aug;19(8):1047-57

[0249] 23) Struglics et al., 2015, Arthritis Rheumatol., Jul;67(7):1816-25

[0250] 24) The Merck Manual, 17.sup.th edition, page 449

[0251] 25) Bellamy et al., 1988, J. Rheumatology, 15:1833-1840

[0252] 26) Wolfe, 1999, Rheumatology, 38:355-361

[0253] 27) Bay-Jensen et al., 2016, Osteoarthritis Cartilage. 24(1):9-20

[0254] 28) Collins et al. 2016. Osteoarthritis Cartilage 24: 1317

ABBREVIATIONS

[0255] BMI=Body mass index; CI=confidence interval, DBPC=double-blind placebo-controlled

[0256] CTX-II=C-telopeptide cross-linking of type II collagen,

[0257] DBPC=Double-Blind Placebo Controlled

[0258] ICOAP=Measure of Intermittent and Constant Osteoarthritis Pain

[0259] ITT=intention-to-treat; JSW=Joint Space Width

[0260] KOOS Symptom Index=Knee Injury and Osteoarthritis Outcome Score symptom index subscale

[0261] KOOS QOL=Knee Injury and Osteoarthritis Outcome Score quality of life subscale

[0262] LOCF=last observation carried forward,

[0263] LFTC=lateral femoro-tibial compartment

[0264] MFTC=medial femoro-tibial compartment

[0265] mITT=modified intention-to-treat

[0266] MOS SF-36=Medical Outcomes Study Short Form-36 General Health Survey

[0267] MRI=magnetic resonance imaging

[0268] NRS pain score=numerical rating scale pain score

[0269] OA=Osteoarthritis

[0270] PGA=Patient's Global Assessment

[0271] PGIC=Patient's Global Impression of Change

[0272] PROC2=neo-epitope of collagen type II propeptide

[0273] WOMAC=Western Ontario and McMaster Universities Osteoarthritis Index; W=Week