OARSI Conference 2017

Elevated levels of CRPM, an inflammatory biomarker correlating with disease activity in RA, are prognostic of radiographic knee OA.

Anne-Christine Bay-Jensen1, Asger Bihlet 2, Inger Byrjalsen2, Jeppe Andersen2, Yi He1, Anne Sofie Siebuhr1, Christian Thudium1, Hans Guehring3, Martin Michaelis3, Bente Juel Riis2, Claus Christiansen2, Christoph Ladel3, Morten A. Karsdal1

Affiliations and emails:

  1. Biomarkers and research, Nordic Bioscience, Denmark
  2. Clinical development, Nordic Bioscience, Denmark
  3. Merck KGA, Germany

Keywords: biomarkers, disease progression, CRPM

Abstract (4023/3700 characters)

Purpose

There are two major needs in clinical development of DMOADs: 1) Identifying a suitable population with an active and progressive disease in order to demonstrate significant improvements by an efficacious intervention; and 2) Phenotyping patients and linking them to a corresponding treatment mode-of-action (e.g. anti-inflammatory, cartilage anabolic). The biochemical marker CRPM is a neo-epitope of C-reactive protein (CRP) and is released from the inflamed tissue when CRP is degraded by proteases such as matrix metalloproteinases. The purpose of this study was to translate the CRPM, from rheumatoid arthritis (RA), where it has been extensively tested, to OA, and to test whether it is predictive of radiographic OA.

Methods

The two phase III OA trials SMC021-2301/02 (clinicaltrial.gov: NCT00486434 and NCT00704847) were used for the investigations. The study included subjects with symptomatic and radiographic knee OA: WOMAC pain ≥150mm and/or WOMAC function ≥510mm, and Kellgren-Lawrence grade (KLG) 2 or 3. KLG were scored for both knees at baseline and year 2 (Y2). The biomarker sub-study Serum CRPM and hsCRP were measured in the placebo groups (N=449/338, fig.1) at baseline (BL) and Y2, in an early RA cohort (N=113) at BL, and in the LITHE study (n=490), which included patients with active, moderate-severe RA (NCT00106535). CRPM was measured at BL and week 4 of the three treatment arms in the LITHE study (placebo, 4 and 8 mg/kg tocilizumab). The association between serum CRPM levels and disease activity score (DAS28) and hsCRP weew investigated by Spearman’s correlations (corrected for multiplicity by Bonferroni). Quartile ranges of CRPM in the early RA cohort were used as references for low, moderate, high and very high CRPM, and was used to define the cut-off between inflammatory OA and non-inflammatory OA at BL and Y2. The case-control study OA knees were divided into cases and controls based on a terminology proposed by the FNIH-OAI consortium (fig.1); knees with KLG≥2 at BL were excluded, and incidence OA at Y2 was defined KLG≥2. Logistic regression was used to compare cases and controls.

Results

Characterization of CRPM There was a significant correlation between disease activity measures and CRPM in both early and moderate to severe RA (tab.1). HsCRP could account for about 40% of the biological variation in CRPM, indicating that although originating from the same protein, hsCRP and CRPM reflect different biological processes, and therefore provide different information. The intra-patient change (CV-%) from BL to week 4 was less for CRPM (36%) as compared to hsCRP (76%), when assessed in the placebo arm. Seventy-five percent of the LITHE patients had high or very high levels of CRPM at BL, which was changed to to a pattern similar to early RA (tab.1). Identification of a cut-off The mean CRPM levels were significantly lower in OA (8.5 [8.3-8.8]ng/mL) compared to the RA patients (15.6 [9.5-21.6] ng/mL); however, a significant subset of OA patients (41% and 31% in SMC2301 and SMC2302) had CRPM levels ≥9ng/mL, as 75% of patients with early RA (tab.1). Incidence knee OA Patients with  BL or Y2 CRPM levels ≥9ng/mL were more likely to develop knee OA than patients with low level of CRPM. Overall, moderate to very high levels of CRPM at BL and Y2 were predictive of incidence OA with odds ratio of 4.6 [1.2-17] and 2.5 [1.2-4.8].

Conclusion

This study provides three major findings: 1) CRPM is associated with disease activity and is modulated in response to anti-inflammatory treatment; 2) a subset of OA patients appear to have tissue inflammation comparable to that of RA; and 3) high CRPM levels is prognostic of incident knee OA. These data suggest that CRPM is a candidate biomarker of disease activity and for patient profiling.

 

 Table 1

 

 Early RA

 Moderate/severe RA

SMC2301

SMC2302

hsCRP

Rho (p-value)

0.54 (<0.0001)

0.66 (<0.0001)

 

 

DAS28

Rho (p-value)

0.43 (<0.0001)

0.24 (<0.0001)

 

 

 

 

 

 

 

 

Interquartile range

No. of patients (%)

No. of patients (%)

No. of patients (%)

 

 

 

Before treatment

After treatment

n1

n2

 

Low

<9 ng/mL

49 (8.2%)

87 (18%)

267 (59%)

233 (69%)

 

Moderate

9-12

100 (17%)

123 (25%)

137 (31%)

71 (21%)

 

High

12 – 15

132 (22%)

105 (21%)

26 (68%)

23 (7%)

 

Very high

>=15

318 (53%)

175 (36%)

19 (4%)

11 (3%)

 

 

 

 

A biomarker of aggrecanase degraded aggrecan is associated with key characteristics of OA  - Investigation in a phase III clinical study

Anne Sofie Siebuhr, Sabine Hoielt, Morten Asser Karsdal, Anne-C. Bay-Jensen

 

Purpose: Aggrecanase degraded aggrecan is known to be the first event in cartilage degradation. However, what association the event has in OA is still to be described. In this study we developed a serological aggrecanase degraded aggrecan biomarker and investigated it in a phase 3 clinical study.

Methods: The aggrecanase degraded aggrecan neo-epitope 374ARGSVI to G2 was used in a sandwich ELISA (huARGS) utilizing 2 monoclonal antibodies. huARGS was firstly validated technically. Specificity was determined using synthetic peptides and in vitro cleavage of purified bovine aggrecan. Intra- and inter-assay variation were determined by 10 individual runs of 8 samples in double determinations. Linearity was determined by 2-fold dilutions of serum and recovery was calculated with the undiluted sample as reference. Interference was tested by spiking human serum with hemoglobin, biotin or lipids. Analyte stability was tested by determining recovery of the analyte after 5 freeze-thaw cycles and after storing human serum at 4⁰C or 20⁰C for 48hours. Biological relevance was investigated by bovine cartilage explants (BEX) and human osteoarthritis cartilage explants (HEX) cultured for 14 days with or without (w/o) catabolic cytokines (oncostatin M [10ng/mL] + TNFα [20ng/mL] (O+T)). huARGS and AGNxI (N-terminal of aggrecanase degraded aggrecan cleavage site) were assessed and Mann-Whitney test was used to test the at the individual days.

huARGS was evaluated clinically by assessing serum level in OA patients and investigating the predictive capacity over 2-years. THE SMC01 study was a phase 3 clinical placebo-controlled study. The current study investigated the placebo group only (n=573). The correlation between the baseline level of huARGS and baseline clinical parameters and change in clinical parameters to year 2 was investigated (Pearsons). Patient tertiles of equally size depending on the baseline huARGS level was made and difference in clinical parameters and change in clinical parameters to year 2 between the tertiles groups were investigated (t-test).

Results: A sandwich ELISA (huARGS) was developed using monoclonal antibodies recognizing 374ARGSVI in the interglobular domain and the G2 domain of aggrecan. huARGS showed low intra- and inter-assay variation (7.5% and 6.5%, respectively) and interference (103.5%), and god linearity (102.6%) and analyte stability (90.4% and 94%; see table). The huARGS was specific of aggrecanase degraded aggrecan. O+T treatment of cartilage increased the huARGS level significantly with peak at day 5, whereas AGNxI peaked at day 7.

The SMC01 cohort was a regular OA cohort with a mean age of 64.5 years, a BMI of 28.8 and WOMAC total was 1134.8 (assessment in mm). huARGS was at baseline correlated to 2 other serological biomarkers of ECM destruction: MMP-degraded type II (C2M, rho: 0.24)  and III collagen (C3M, rho: 0.15). huARGS was associated with cartilage volume assessed by MRI at the femoral condyle at baseline (figure). The cartilage volume level was lowest in the low huARGS tertile and highest in the high huARGS tertile. The WOMAC level was higher in the middle huARGS tertile compared to the highest tertile (total: p=0.013, function: p=0.0091 and pain: p=0.046). The huARGS middle tertile had the significantly lowest change in 2-year in WOMAC function (low: p=0.047, high: p=0.025).

Conclusions: We developed a robust biomarker assay specific for aggrecanase degraded aggrecan and had indications of this being the earliest biomarker of aggrecan degradation. huARGS was associated with cartilage volume and WOMAC both at baseline and with the change to year 2. Thus, huARGS is associated with the key characteristics of OA and may therefore be an early biomarker of treatment efficacy.

 

The effect of small molecule anti-inflammatory inhibitors on cartilage degradation is dependent on specific cytokine pathways

Christian Thudium, kjelgaard, Ali Mobasheri, Morten Karsdal, Anne-Christine Bay-Jensen

aRheumatology,  Biomarkers and Research, Nordic Bioscience A/S, Herlev, Denmark

bD-BOARD EU Consortium for Biomarker Discovery, University of Surrey, UK

Purpose: 

Inflammatory osteoarthritis (iOA) is a degenerative joint disease with an inflammatory component allowing for the development of novel anti-inflammatory treatments for patient benefit. A number of different signaling pathways have been associated with the inflammation-driven degradation of the extracellular matrix (ECM) and targeted in drug development with varying clinical results. 

A better understanding of the signaling pathways’ modulation of the ECM and how it is reflected in the biomarker profile may therefore aid in selection and development of novel anti-inflammatory disease-modifying osteoarthritis drugs (DMOADs) for iOA.

The aim of this study was to investigate the differential effect of the anti-inflammatory inhibitors SB203580 (p38 inhibitor), Tofacitinib (Jak inhibitor), TPCA-1 (IKK inhibitor) and R406 (syk inhibitor) on cartilage turnover biomarkers driven by two different pro-inflammatory cytokines, IL-1α and TNF-α.

Methods: 

Full depth cartilage ex vivo cultures were cultured for 3 weeks stimulated with either OSM [10 ng/ml] and TNF-α[2 ng/ml] [O+T] or IL-1α [10 ng/mL] alone, together with SB203580, Tofacitinib, TPCA-1, or R406 at 3 µM, 1 µM and 0.3 µM. Untreated explants (w/o) were included as negative controls. The cartilage ECM turnover was assessed by measuring the tissue fingerprint biomarkers C2M and AGNx1 in the conditioned medium with ELISA.

Results: 

Aggrecanase mediated degradation of aggrecan was measured by AGNx1. The JAK inhibitor Tofacitinib and TPCA-1 inhibited O+T and IL-1α induced AGNx1 release in a dose dependent manner with 91-95 % at the highest dose (3µM), and R406 inhibited 58-81% of the O+T and IL-1α induced AGNx1 release at the highest dose. Whereas the p38 inhibitor SB203580 had no effect. The degradation of type II collagen was measured by the MMP-mediated degradation of type II collagen (C2M) (Figure 1a and b). Tofacitinib, TPCA-1 and R406 significantly inhibited 88-99% of the C2M release in O+T and IL-1α stimulated cultures (Tofacitinib: Figure 1d). SB203580 significantly inhibited C2M in O+T stimulated conditions in a dose dependent manner, with 87% in the highest concentration (Figure 1c). In contrast, SB203580 failed to inhibit C2M release in IL-1α stimulated conditions (Figure 1c).

Conclusion: 

The four inhibitors tested had an inhibitory effect on the degradation of type II collagen; however, for the p38 inhibitor only in a TNFα driven setting. Furthermore, the p38 inhibitor was not able to inhibit aggrecan degradation in either TNFα nor IL-1α stimulated conditions, while the others dose dependently or completely inhibited this in both inflammatory settings. These findings suggest that anti-inflammatory effects tested ex vivo are dependent on type of cytokine activation, and, thus, indicate that signaling pathways driving inflammatory joint diseases needs to be given consideration when assessing the effect of potential anti-inflammatory treatments and interpreting biomarker analysis. 

 

 

 

Pre-treatment of synovial fluid enable precise and accurate measurement of neo-epitope biomarkers

Authors: CFKP, CST, BJU, Per Hägglund, Ole Simonsen (Aalborg hospital), MK, ALI, ACBJ

 

Purpose:

The synovial fluid (SF) provides lubrication and nutrients to the cartilage, as well as removing waste products from the cartilage. Biomarkers measured in SF may provide direct measure of cartilage and synovial turnover of the joint from where the SF was retrieved. In contrast, serum measurements of joint-related biomarker may reflect cartilage and synovial turnover of multiple joints. However, the matrix of the synovial fluid is different than the matrix of serum or plasma as it contains a high degree of hyaluronic acid that makes the SF viscous and sticky. The matrix effect of synovial fluid is therefore also different, which makes it challenging to measure small neo-epitope biomarkers herein. Accuracy and precision are key in measuring biomarkers in all types of samples. Thus, it is critical to thoroughly test and validate a biomarker in the matrix it is being measured in. The purpose of this study was to optimize and validate the measurement of four joint-related neo-epitope biomarkers, CRPM (chronic tissue inflammation), C2M and AGNx1 (cartilage degradation), and C3M (connective tissue remodeling) in SF. 

 

Methods:

Human and bovine SF was centrifuged and pre-treated with 0.4 mU chondroitin ABC and 0.4 mU Endo-β-galactosidase pr. µL synovial fluid for 3 hours at 37°C prior to measurement of the biomarkers. Each biomarker tested for dilution recovery and spike-in recovery. Dilution recovery: Samples were diluted in a 2-fold dilution in the assay buffers from the biomarker measured. C3M enrichment of synovial fluid: The standard peptide from C3M assay were added to the untreated synovial fluid sample. The synovial fluid were then treated and used for dilution recovery as described above. Spike-in recovery: The standard peptide from the biomarker measured were diluted in a 2-fold dilution and four points were mixed 1+1 with bovine synovial fluid or assay buffer. Acceptance range for dilution- and spike-in recovery was 80-120%. As control synovial fluid were mixed 1+1 in assay buffer and also included. The biomarkers were measured with competitive ELISA specific for C2M, C3M, CRPM, and exAGNx1 (NITEGE).  The DBOARD consortium supported this study.

Results:

The deglycosylated pre-treated synovial fluid samples had a dilution recovery between 98-120% compared to neat for C2M (Fig.1A), 89-110% for CRPM (Fig.1C), and 85-93% for AGNx1. The levels of C3M were just around lower limit of measuring range (LLMR) and the dilution recovery could not be determined in the synovial fluid. We therefore enriched the synovial fluid with C3M to measure dilution recovery in the synovial fluid matrix. When enriched for C3M the dilution recovery was 99-104%. Additional the spike-in recovery was measured to test the precision of the assay in synovial fluid. The spike-in recovery for C2M was 97-106% (Fig.1B), for CRPM within measuring range 114-107% (Fig.1D), for AGNx1 97-121%, and 109-122% for C3M within measuring range.

Conclusion:

Pre-treatment with 0.4 mU chondroitin ABC and 0.4 mU Endo-β-galactosidase pr. µL synovial fluid enabled accurate and precise measurement for the four joint-related biomarkers: C2M, C3M, CRPM, and AGNx1. These data show that it is possible to reliably measure neo-epitope biomarkers of cartilage and synovial turnover in the synovial fluid.

 

 

A Novel, High Sensitivity Marker, hsPro-C2, Of Cartilage Formation, Was Developed And Tested In A Phase II Clinical Trial Of PTH

Yunyun Luo1, Yi He1, Inger Byrjalsen1, Kim Henriksen1, Natasja Stæhr Gudmann1, Ali Mobasheri2, Gitte Hansen1, Morten A. Karsdal1, Anne C. Bay-Jensen1

1Nordic Bioscience A/S, Herlev, Denmark

2D-BOARD EU Consortium for Biomarker Discovery, University of Surrey, Surrey, United Kingdom

 

 

Purpose:

Presently, measurement of cartilage formation rely on magnetic resonance imaging (MRI), which is a sensitive measure, but which need long follow-up time. Thus, there is an unmet need, in disease-modifying osteoarthritis drug (DMOAD) development, for an objective and non-invasive marker of cartilage formation, which can provide early indication of drug efficacy. The objective was to enable assessment of type IIB collagen synthesis in serum from human subjects. Such tool may be applicable as a non-invasive biomarker of cartilage repair and growth in the development of cartilage anabolic drugs. No cartilage anabolic DMOAD have to date been approved thus for proof of concept, we used an osteoporosis (OP) trial testing teriparatide (human parathyroid hormone (PTH) 1-34). Teriparatide has been shown to have potential chondro-protective and chondro-regenerative effects on articular cartilage in vitro and in vivo. However, it remains unclear whether the pro-anabolic effect of teriparatide translate to human OA.

Methods:

A high sensitivity competitive electro-chemiluminescence immunoassay for detection of PIIBNP (hsPro-C2 ECLIA) was developed and the technical performance evaluated. From a randomized, double-blind placebo-controlled study with an open-label active comparator/positive control (teriparatide) in postmenopausal women with OP (clinicaltrials.gov: NCT01321723), the biomarker sub-study included 64 Caucasian postmenopausal women (age 45-80 years) with OP duration of at least 5 years. Thirty-two women were treated with teriparatide, and 32 with placebo. Biomarkers of bone formation (PINP; procollagen type I N-terminal propeptide) and cartilage formation (hsPro-C2) were analyzed retrospectively at baseline, week 4, 12 and 24. Correlation between PINP and hsPro-C2 at baseline and change at week 4 relative to baseline were investigated by Pearson’s correlation.

Results:

The technical performance of hsPro-C2 ECLIA was summarized in Table 1. The intra-assay CV was 5.4% and the inter-assay CV was 5.5%. The measurement range was 1-32 ng/ml. The spiking and dilution recovery tested in human serum were 100 ± 20% within the measurement rang of the assay.

The mean percent change in serum hsPro-C2 level was higher in teriparatide treated group compared to the placebo group at week 4 (9%), 12 (3%) and 24 (10%), although it was not statistically significant (Fig.1A). There were no statistically significant correlation between hsPro-C2 and PINP before treatment initiation (Fig.2A), however an increase in hsPro-C2 was significantly associated with increase in PINP in the teriparatide treated group at week 4 (R2=0.1976, p<0.05, Fig.2B). This was not observed in the placebo group (data not shown).

Conclusions:

In spite of the small sample size of this study there was a clear trend toward increased cartilage formation in the PTH treated group over time. Furthermore, hsPro-C2 changes correlated with the changes in PINP, which is believed to be a pharmacodynamics biomarker of teriparatide treatment, indicating that hsPro-C2 reflect a possible chondro-anabolic effect of PTH. It is concluded that hsPro-C2 may be a promising and novel marker of cartilage formation to be used in DMOAD development.

Table 1 Summary of technical performance for two biomarkers assays

Parameters

hsPro-C2 Competition ECLIA

PINP Sandwiches ECLIA

Linear range of standard (ng/mL)

1.0 - 38.5

5-1200

LLOD (ng/ml)

0.13

<5

Intra-assay % CV

5.4

2.2

Inter-assay % CV

5.5

2.8

Spiking Recovery, % range

98

NA

Dilution Recovery, % range

99

NA

            LLOD: lower limit of detection

 

Figure 1 Response as median percent change from baseline (± interquartile range) in hsPro-C2 (A) and PINP (B) in women receiving teriparatide () or placebo (). Teriparatide was administered subcutaneously with 20 mg/day. 

  

Figure 2 Correlation between the serum procollagen type IIB N-terminal propeptide (Pro-C2) change and serum procollagen type I N-terminal propeptide (PINP) change in patients receiving teriparatide (A, B) and placebo (A)Person’s correlation coefficient (R2)

 

Anabolic induction of cartilage can be achieved in a novel explant co-culture model of bovine cartilage and synovial membrane by TGF-β1

 

Anne Sofie Siebuhr, Sarah Dahab, Cristina Acro Cabanes, Anne-C. Bay-Jensen, Christian Thudium

Background and aim: Currently, there is no treatment reversing or event halting cartilage degradation. One of the obstacles in development of drug reversing or halting cartilage degradation is the lack of model systems for early drug testing. We therefore investigated if models of bovine explants, both as co-culture and alone could be anabolic stimulated with growth factors.

Method: Synovium (bSME) from healthy bovine (<24months old) hind knees were isolated and cut into explants of 30 ±5 mg and bovine cartilage (BEX) was cut into equally sized explants using a 6mm diameter biopsy puncher. bSME and BEX was cultured together (bCC) or alone. Explants were cultured for 14-35 days in DMEM-GlutaMAX™ with or without continuous stimuli with 4 replicates per treatment: Oncostatin M [10 ng/mL] + TNF-α [20 ng/mL] (O+T), GM6001 [1uM], IGF-1 [100ng/mL] or TGFβ-1 [50-0.5 ng/mL]. Media were changed 3 times a week and the viability was assessed with Alamar Blue every week. The reversibility of cartilage degradation was investigated by 10days incubation with O+T followed by 21 days of TGFβ-1 [50-0.5 ng/mL] treatment.

In the conditioned medium following biomarkers were assessed: Type I, II and III collagen degradation (C1M, C2M and C3M, respectively), formation of type I and II collagen (P1NP and PRO-C2, respectively), aggrecanase degraded aggrecan (AGNxI) and matrix metalloproteinase degraded aggrecan (exFFGV). These biomarkers are in-house competitive ELISAs.

Results: Explants were viable throughout the experiments, albeit the bSME lost some viability over time. bSME treated with O+T had increased C1M and C3M (p-values respectively) compared to w/o from day 10. In the bCC O+T increased C1M from day 21, and C2M and C3M from day 14 (p-value, respectively). TNF-a alone also increased C1M, C2M and C3M, but to a lower extend than when co-treated with OSM. The BEX alone had increased C1M at day 21, but the level about 25% of the level in the bCC. C3M levels increased to the same extend in the BEX as in the bCC. Co-treatment with GM6001 completely inhibited this signal and even decreased the C1M and C3M compared to w/o (p-value). In the bSME the O+T treatment did not affect the P1NP level at any time point, but in bCC O+T and TNF-a alone lowered the PRO-C2 level (p-value, respectively). In the bCC the AGNxI was increased at day 7 and 10 in the O+T groups and TNF-a also increased AGNxI, but to a lower extend. exFFGV was increased from day 21 in the bCC (p-value).

TGF-b1 was continuously and dose-dependently increased P1NP from day 7 compared to w/o (p-value) in the bSME. In the explants which was treated with O+T for 10 days TGF-b1 increased P1NP 7 days after TNF-b1 treatment was initiated. IGF-1 did not affect the P1NP level at any time point. In the bCC both IGF-1 and TGF-b1 (dose-dependently) sustained the PRO-C2 level at the level of baseline throughout the study periods.

Conclusion: We developed two bovine explants model (bSME and bCC), which both could be anabolic and catabolic stimulated and anabolic rescued after catabolic treatment. Previous explant models of human synovial membrane and human co-culture of cartilage and synovial membrane lacked the anabolic capacity. Anabolic stimulation was achieved with TGF-b1 in both bSME and bCC, but only in the bCC did IGF-1 have anabolic capacities. This indicates a secondary signal from the cartilage to achieve anabolic stimulation in synovial membrane. These newly developed explants models can be used in the early development of anabolic drugs for cartilage degenerative diseases.

  

 

Establishment of a human synovium and cartilage co-culture

Authors: CFKP, ACBJ, Thorbjørn, MK, Per Hägglund, Siebuhr, CST

Purpose:

Osteoarthritis (OA) is a degenerative joint disease with a low-grade inflammatory component that leads to an altered turnover of extracellular matrix (ECM), not only in the cartilage, but also in the synovium and bone. The diseased tissues are believed to interact with each other and initiate and drive OA. However, the mechanism behind this interaction and the effect on the ECM turnover are unknown. The aim of this study was to establish an ex vivo co-culture model of the cartilage and synovium to study the interaction between the two tissues and its effect on the ECM turnover.

Methods:

Human synovium and cartilage were obtained from end-stage OA patients undergoing total knee replacement. The synovium were cut into explants of 30±4 mg and the cartilage were punched with a 5 mm biopsy punches. After isolation of the two tissues, they were immediately added together in one well. Additionally both tissues were also cultured alone. The tissues were cultured for 14 days with OSM [10ng/mL] and TNFα [20ng/mL] (O+T), alone (w/o), or with O+T and GM6001 10 µM. GM6001 was only included in the co-culture system. The metabolic activity was measured with alamar blue weekly. Conditioned media were removed three times a week and fresh treatment added. The conditioned media were used for biomarker measurement. Four biomarkers, C3M, AGNx1, FFGV, and C2M, were measured by ELISA in the conditioned media.

Result:

The human co-culture of the synovium and cartilage were cultured for 14 days. The explants were metabolic active throughout the study. However, the metabolic activity of the synovium dropped after 7 days on culture. Four biomarkers of the joint ECM turnover were measured in the conditioned medium and the accumulated biomarker over 14 days were calculated based on measurements of four time points. O+T increased the release of C3M 3.2-fold compared to w/o in the co-culture and 8.6-fold (P=0.016) compared to w/o in the synovium alone (Fig.1a). O+T increased the release of C2M 4.9-fold compared to w/o in the co-culture (P=0.012), 2-fold compared to non-treated synovium (P=0.041), and 2.4-fold compared to non-treated cartilage (P<0.001) (Fig. 1b). The MMP-mediated aggrecanse degradation, FFGV, was increased in 12.9-fold in response to O+T compared to w/o in the co-culture (P=0.003) and 7.1-fold in cartilage alone (P=0.002) (Fig. 1c). AGNx1, aggrecanase mediated degradation of aggrecan, was similarly released in response to O+T treatment compared to w/o, 2.2-fold increase in the co-culture (P=0.008) and 3.7-fold increase from cartilage alone (P<0.001) (Fig. 1d). Neither FFGV nor AGNx1 were released from synovium alone (Fig. 1c and d). No increased biomarker release was measured in the co-culture without external cytokine stimuli. The release of C3M, C2M and FFGV were MMP-depended as GM6001 inhibited the release (Fig. 1).

 

Conclusion:

It was possible to culture the synovium and cartilage together for 14 days. The pro-inflammatory cytokines OSM and TNFα induced release of C3M, C2M, FFGV, and AGNx1 from the co-culture, where C3M originated from the synovium, C2M originated primarily from the cartilage, and FFGV and AGNx1 only originated from the cartilage. This model provides a tool to study effect of potential drugs on the synovium and cartilage together.

 

 

Bone marrow lesions are associated with pain, but not with inflammatory markers in end-stage knee osteoarthritis patients

 

Maja R Radojčić1,2, Christian S Thudium1, Kim Henriksen1, Keith Tan3, Rolf Karlsten4, Amanda Dudley3, Iain Chessell3, Morten A Karsdal1, Anne-Christine Bay-Jensen1, Michel D Crema5,6, Ali Guermazi5

 

1Nordic Bioscience Biomarkers & Research, Nordic Bioscience A/S, Herlev, Denmark

2Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark

3Neuroscience Innovative Medicines, AstraZeneca, Cambridge, United Kingdom

4Department of Surgical Sciences, Anaesthesiology and Intensive Care, Uppsala University, Uppsala, Sweden

5Quantitative Imaging Center, Department of Radiology, Boston University School of Medicine, Boston, Massachusetts, USA

6Department of Radiology, Hôpital Saint-Antoine, University Paris VI, Paris, France

 

Purpose

Bone marrow lesions (BMLs) are important and intriguing features of knee osteoarthritis (OA) that have been linked to disease progression and pain. However, the pathological process that drives this OA feature remains unclear. Although inflammation has been suggested as a potential driver of OA, its relationship with BMLs is uncertain. Circulating biomarkers can assist in measuring and understanding different pathological processes, including inflammation. C-reactive protein (CRP) is a systemic acute inflammatory biomarker, while matrix metalloproteinase-cleaved CRP (CRPM), which is derived from the inflamed tissue, is a local chronic inflammatory biomarker. We aimed to investigate the association between BMLs and pain in end-stage knee OA patients and associations between biochemical markers of acute and chronic inflammation with BMLs.

Methods

One hundred four patients (mean age 67 years, 61.5% women, and average body mass index 30.4) with OA undergoing a total knee replacement were included in the study. The study protocol was approved by the Ethics Committee and written informed consent was obtained from all patients. The biochemical markers high-sensitivity CRP and CRPM (neo-epitope of CRP) were measured in serum using immuno-assays. Magnetic resonance imaging (MRI) scans were performed using 3-plane fat-suppressed, proton density-weighted sequences, and BMLs were read according to the MRI osteoarthritis knee score (MOAKS). BML size (sBML), percent of BML accounted for a cyst (pBML), and the number of BML (nBML) were semi-quantitatively scored at different subregions across the knee and summed into three total scores each representing the specific characteristic of BML. The pain was assessed using the Western Ontario & McMaster Universities Osteoarthritis Index (WOMAC) Pain scale and the Neuropathic Pain Questionnaire (NPQ) total score. We used linear regression to investigate associations between BMLs and pain, and biomarkers and BMLs. All models were adjusted for age, gender and body mass index.

Results

We found that sBML was associated with NPQ (B=0.84, 95% CI 0.01-1.68), but not with WOMAC pain. pBML was associated with both pain dimensions, WOMAC (B=0.32, 95% CI 0.03-0.60 ) and NPQ (B=1.06, 95% CI 0.43-1.69). Similarly, we also observed a significant association between nBML and WOMAC (B=0.79, 95% CI 0.02-1.56), and between nBML and NPQ (B=2.30, 95% CI 0.57-4.03). In contrast, we did not observe any statistically significant association between inflammatory biochemical markers (CRP and CRPM) and any of three BML characteristics.  

Conclusions

This cross-sectional study showed that BMLs are associated with WOMAC pain and neuropathic features in end-stage OA patients, and that of three BML characteristics, the number of BMLs is the strongest predictor of pain. Further, acute and chronic inflammatory biochemical markers were not associated with BMLs. Our findings suggest that inflammation is less likely the process that describes BMLs pathology at this stage of the disease, and that markers evaluating BMLs are needed. This OA feature should be further investigated because of its relation to clinical symptoms and because BMLs could be a target for pharmacological treatment that modifies the disease and relieves pain.

 

Acknowledgments

We would like to thank the patients that agreed to participate in this study and the staff from the participating centres in Canada and Sweden.

The present study has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 642720. The CRPM measurements were partly supported by the D-BOARD consortium funded by European Commission Framework 7 programme (EU FP7; HEALTH.2012.2.4.5-2, project number 305815, Novel Diagnostics and Biomarkers for Early Identification of Chronic Inflammatory Joint Diseases). The funding agency had no role in the design and conduct of the study, collection, management, analyses and interpretation of the data.

The clinical study was carried out by AstraZeneca (Clinicaltrials.gov ID: NCT01611441).

Disclosures

MRR is Marie Sklodowska-Curie fellow and has no competing financial interest in relation to the described work. CST, KH, MAK, and ACBJ are Nordic Bioscience employees; KH, MAK and ACBJ own stock/stock options at Nordic Bioscience. AD, IC, and KT are AstraZeneca employees. RK was AstraZeneca employee when clinical part of the study was conducted. MDC is a shareholder of Boston Imaging Core Lab (BICL). AG is president of BICL, a company providing radiologic image assessment services; and consultant to MerckSerono, OrthoTrophix, AstraZeneca, Genzyme and TissueGene.

 

 

EFFECT OF NAPROXEN AND KBP-056 IN A COLLAGEN INDUCED ARTHRITIS RAT MODEL OF ARTHRITIC PAIN

  1. Katri1,2, H. Löfvall1, C. S. Thudium1, M.A. Karsdal1 ,K. Henriksen1

1Nordic Bioscience A/S, Herlev, Denmark

2Department of Drug Design and Pharmacology, University of Copenhagen, Denmark

 

  1. Purpose

Pain is one of the most challenging symptoms for patients with rheumatoid arthritis (RA) and is frequently considered to be solely a consequence of inflammation in the joints that eventually leads to cartilage and bone destruction. Agents such as analgesics, non-steroidal anti-inflammatory drugs (NSAIDs) are used to give symptomatic relief but alone do not change the course of the disease of RA or prevent joint destruction. Salmon calcitonin has been widely used in the treatment of bone diseases characterized by increased bone resorption mediated by the osteoclasts. KBP-056 is a novel Dual Amylin Calcitonin Receptor Agonist (DACRA), developed for the treatment of osteoarthritis. We aimed to see if the combination of the standard of care treatment (Naproxen) and an inhibitor of bone turnover, KBP-056, might further reduce disease activity. For that purpose, we evaluated clinical effects and measured the pain levels in a CIA animal rat model of RA, with regards to both structure and pain.

  1. Methods

Collagen induced arthritis (CIA) was induced in 47 female Lewis rats by immunization with type II collagen. A booster injection was given to the rats one week later at the base of the tail. The study contained 5 treatment groups. Treatment groups were CIA+Napr, CIA+KBP-056+Napr, CIA+KBP-056, CIA control and Sham. Evaluation of disease progression, which consists of monitoring of weight, behavioral assessment (fur grooming, posture and exploratory behavior) and daily joint swelling scoring was done 7 days a week. Score of joints >10 and body loss <20% were indicative of severe arthritis and the animals needed to be terminated. Burrowing, which is a model that relies on innate animal behavior, was used to assess pain-like behavior and the efficacy of the drugs.

  1. Results

The level of inflammation in the paws of the CIA control animals was significantly higher in comparison with all the other groups, while there was no difference in the combination of the two drugs when compared to the group which received Naproxen only. Median survival for the CIA control was 20.5 days after the induction of the immunoreaction while median survival for the CIA+KBP-056+Napr was 43 and it was found to be significantly higher than those who received Naproxen only (p<0.01). The behavioral assessment showed a significant effect 34 days after immunization between the CIA+KBP-056+Napr and the CIA+Napr with the animals that received both drugs having a better health score (p<0.030). However, the joint score reached significance between the CIA+KBP-056+Napr and CIA+Napr with the first having less swollen joints at day 39 and the difference was more pronounced at day 42 (p<0.0009). The burrowing performance of all groups, except the sham treated, decreased significantly 16 days after the induction of the disease. The CIA+KBP-056+Napr and CIA+Napr animals showed increased burrowing performance, p<0.005 and p<0.015 respectively, when compared to the CIA control.

  1. Conclusion

This study shows that combination therapy with NSAIDs and an inhibitor of bone resorption (KBP-056) improved the health and behavior levels of the animals as well as decreased the pain outcome compared with the diseased non treated group. What’s interesting is that CIA+KBP-056+Napr group showed a slightly better pattern in the health-survival score as well as in the joint score, when compared to CIA+Napr treatment alone. There is no indication that KBP-056 provides an anti-inflammatory benefit in RA disease but we speculate that it has an effect on the bone structure. However, careful histological analysis will be needed to confirm the hypothesis.

 

Pain of knee osteoarthritis is associated with a collagen degradation

 

Anne C. Bay-Jensen, Steven B Abramson, Jonathan Samuels, Svetlana Krasnokutsky, Tina Manon-Jensen, Morten A. Karsdal, Mukundan Attur

(1)    Rheumatology, Biomarkers and Research, Nordic Bioscience

2) Division of Rheumatology, Department of Medicine, New York University Hospital for Joint Diseases, New York University School of Medicine New York, New York

Abstract

Background:

Osteoarthritis (OA) is heterogeneous disease characterized by pain and tissue destruction which is in some case concomitant with inflammation. However the link between pain and tissue destruction is yet unknown. Collagens are the main structural proteins of the joint extracellular matrix (ECM). The degradation of especially type I (connective tissue), II (cartilage), III (synovium) and IV (basement membrane) collagens have been shown to be elevated in joint degenerative diseases.

Objective:

To investigate whether biomarkers reflecting collagen degradation were associated with symptomatic knee OA and different pain and inflammatory phenotypes.

Methods:

111 knee OA patients, 62% women, from NYUHJD progression cohort study with varying degree of OA was included: mean (SD) age, 32 (10); mean BMI (SD), 27 (4); NSAID users, 23%; radiographic OA (Kellgren-Lawrence (KL) ≥2) 68%; and bilateral knee OA; 87%. Pain was assessed by visual analogue scale (VAS, 100 mm) pain score and by Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) at baseline at a 2-year follow-up visit. Median (interquartile range) were 39 (13-69) and 37 (13-52) for baseline VAS pain and WOMAC pain at baseline respectively.

Four biomarkers of type I, II, III and IV collagen degradation (C1M, C2M, C3M and C4M), as well as the two inflammatory biomarker CRPM and hsCRP, were assessed in baseline serum sample. The biomarkers were log2 transformed. Associations between baseline biomarkers, baseline pain scores and delta pain scores were assessed by multivariate linear mixed model. The model included gender, age, BMI, KL of the signal knee, bilateral knee OA (0[no]/1[yes]) and NSAID use (0/1). Pain progression were represented by the delta change.

Phenotype definitions: Patients with i) a mild/moderate phenotype had a baseline VASpain<54 and FU VASpain<30, ii) a moderate/severe phenotype had VASpain>30 at both baseline and FU, and iii) a transitional phenotype had either VASpain_BL<30 and VASpain_FU>54 or VASpain_BL>54 and VASpain_FU<30. Patients with iv) non-inflammatory phenotype were low in CRPM (<12nM), v) inflammatory OA were high in CRPM but low in hsCRP (<5), and vi) a highly active phenotype were high in both CRPM and hsCRP.

Results:

Baseline association between pain and biomarkers: C2M (β -17.9, p<0.0001) and KL of the signal knee (β -5.41, p=0.0031) were significantly associated with WOMAC pain at baseline, whereas (β -12.4, p=0.0033), C3M (β -19.9, p<0.059), age (β -0.84, p<0.0018), signal knee KL (β 8.93, p<0.0021) and bilateral knee OA (β -12.2, p=0.087) were associated with VAS pain at baseline.

Association between baseline biomarkers and change pain: C2M (β 13.3, p=0.0016), age (β 0.49, p=0.029) and bilateral (β -12.0, p=0.043) were significantly associated with delta WOMAC pain, whereas only age, BMI and NSAID at baseline was associated with delta VAS pain.

Association between pain phenotypes and baseline biomarkers. Only C2M were significantly associated with different phenotypes. Patients with mild to moderate pain at both baseline and follow-up had significantly higher levels of C2M compared to both patients with patient with transitional severe pain (p=0.0014) and always moderate to severe pain (p=0.04).

Biomarker levels, pain and delta pain in patients with or without inflammatory OA: Patient with a non-inflammatory phenotype had significantly lower WOMAC (p<0.05) and boarder line significant VAS (p<0.1) pain at baseline. C1M was higher (p<0.05) in the highly active group compared to the non-inflammatory OA. C3M was higher (p<0.05) in the inflammatory OA group compared to the non-inflammatory group. There was a trend towards C4M (p<0.1) being lower in the non-inflammatory OA vs the two other groups. There were no difference in C2M.

Conclusion: Different collagen degradation products are linked differentially to different phenotypes. Especially C2M and C3M were associate with levels at baseline. Cartilage degradation, was consistently linked to pain changes and phenotypes, whereas it was not associated with an inflammatory phenotype. In contrast, C1M and C3M were linked to an inflammatory phenotype, however differentiated as C1M was higher in the highly active phenotype and C3M higher in the inflammatory OA phenotype.

 

Variable

WOMAC pain (0-100) (n=97)

VAS pain (0-100) (n=97)

Delta WOMAC pain

Delta VAS pain

 

Coefficient

SE

P-value (<0.1)

Coefficient

SE

P-value (<0.1)

Coefficient

SE

P-value (<0.1)

Coefficient

SE

P-value (<0.1)

Gender (female)

-2.98

3.87

 

2.66

6.32

 

-0.88

4.39

 

4.21

5.72

 

Age

-0.32

0.20

 

-0.84

0.26

0.0018

0.49

0.22

0.029

0.60

0.28

0.032

BMI

-0.07

0.53

 

0.19

0.63

 

-0.97

0.60

 

-1.74

0.73

0.020

Sig_KL_BL

5.41

1.78

0.0031

8.93

2.81

0.0021

1.30

2.58

 

-1.95

3.19

 

BilateralOA (yes)

1.85

4.72

 

-12.2

7.04

0.087

-12.0

5.82

0.043

-1.86

7.29

 

NSAIDuse (yes)

1.87

3.90

 

9.34

5.76

 

0.70

5.73

 

-16.8

5.90

0.0056

Ln_sC1M

4.61

5.76

 

2.91

6.43

 

-0.49

5.59

 

-3.90

6.94

 

Ln_sC2M

-17.9

3.45

<0.0001

-12.4

4.10

0.0033

13.3

4.08

0.0016

8.71

5.71

 

Ln_sC3M

-7.52

9.33

 

-19.9

10.4

0.059

13.0

10.2

 

15.7

13.9

 

Ln_sC4M

-4.96

5.64

 

-6.81

7.94

 

-4.29

6.12

 

-3.18

7.48

 

Ln_hsCRP

-2.48

3.13

 

-0.65

4.18

 

-0.62

0.73

 

0.37

0.76

 

ln_sCRPM

9.90

8.57

 

9.16

13.0

 

0.09

10.3

 

11.6

13.7

 

 

 

 

Variable

WOMAC pain (0-100) (n=97)

VAS pain (0-100) (n=97)

Delta WOMAC pain

Delta VAS pain

 

Coefficient

SE

P-value (<0.1)

Coefficient

SE

P-value (<0.1)

Coefficient

SE

P-value (<0.1)

Coefficient

SE

P-value (<0.1)

Gender (female)

-2.98

3.87

 

2.66

6.32

 

-0.88

4.3874

 

4.2082

5.7158

 

Age

-0.32

0.20

 

-0.84

0.26

0.0018

0.49

0.2191

0.029

0.6024

0.2766

0.032

BMI

-0.07

0.53

 

0.19

0.63

 

-0.97

0.5974

 

-1.7401

0.7341

0.020

Sig_KL_BL

5.41

1.78

0.0031

8.93

2.81

0.0021

1.30

2.5750

 

-1.9498

3.1921

 

BilateralOA (yes)

1.85

4.72

 

-12.20

7.04

0.087

-12.0

5.8192

0.043

-1.8647

7.2901

 

NSAIDuse (yes)

1.87

3.90

 

9.34

5.76

 

0.70

5.7342

 

-16.7639

5.8950

0.0056

Ln_sC1M

4.61

5.76

 

2.91

6.43

 

-0.49

5.5910

 

-3.8967

6.9386

 

Ln_sC2M

-17.93

3.45

<0.0001

-12.4

4.10

0.0033

13.3

4.0780

0.0016

8.7127

5.7087

 

Ln_sC3M

-7.52

9.33

 

-19.9

10.4

0.059

13.0

10.1693

 

15.7076

13.8965

 

Ln_sC4M

-4.96

5.64

 

-6.81

7.94

 

-4.29

6.1208

 

-3.1759

7.4749

 

Ln_hsCRP

-2.48

3.13

 

-0.65

4.18

 

-0.62

0.7252

 

0.3710

0.7563

 

ln_sCRPM

9.90

8.57

 

9.16

13.0

 

0.09

10.3423

 

11.6094

13.7315

 

 

 

Group_BL

 

Outcome_24

1

2

3

 

1

34

13

13

60 (55.0%)

2

6

8

8

22 (20.2%)

3

4

4

19

27 (24.8%)

 

44

25

40

109

 

 

Student-Newman-Keuls test for all pairwise comparisons

Factor

n

Mean

SD

Different (P<0.05) from factor nr

Different (P<0.001) from factor nr

(1) 1

53

-1.0533

0.4226

(2)(3)

(3)

(2) 2

37

-1.2885

0.4763

(1)(3)

 

(3) 3

16

-1.5972

0.7826

(1)(2)

(1)

 

 

 

Group

 

Outcome_24mo

1

2

3

 

1

Normal (1)

Normal (1)

Transitional (3)

60 (55.0%)

2

Normal (1)

Chronic (2)

Chronic (2)

22 (20.2%)

3

Transitional (3)

Chronic (2)

Chronic (2)

27 (24.8%)

 

44

25

40

109

 

Summary statistics table

 

hsCRP_CRPM_cat

 

OA

Inflammatory OA

OA with acute inflammation

Highly active inflammatory OA

 

N

Mean

95% CI

N

Mean

95% CI

N

Mean

95% CI

N

Mean

95% CI

VASpain100_BL

74

38

32 to 46

13

44.231

24.593 to 63.868

6

37.000

17.559 to 56.441

14

57.857

41.945 to 73.769

WOMACpain100_BL

74

34.076

28.487 to 39.664

13

32.323

16.806 to 47.840

6

34.433

10.045 to 58.822

14

50.943

37.070 to 64.816

VASpain_delta

72

-5.333

-11.841 to 1.175

13

-14.231

-29.397 to 0.935

6

7.167

-19.035 to 33.368

14

-16.071

-33.380 to 1.237

WOMACpain_delta

72

-5.8

-10.6 to -1.1

13

-10.4

-20.412 to -0.357

6

0.367

-26.770 to 27.503

14

-13.371

-29.337 to 2.594