WO2007038501A2

WO2007038501A2 - Rheumatoid arthritis markers

Info

Publication number: WO2007038501A2
Application number: PCT/US2006/037461
Authority: WO
Inventors: Peter K. Gregersen; Franak Batliwalla
Original assignee: Feinstein Institutes for Medical Research
Current assignee: Feinstein Institutes for Medical Research
Priority date: 2005-09-27
Filing date: 2006-09-27
Publication date: 2007-04-05
Anticipated expiration: 2008-03-27
Also published as: WO2007038501A3

Abstract

Provided are methods of predicting responsiveness to TNF blockade therapy in a subject with rheumatoid arthritis (RA). Also provided are methods of determining the effectiveness of TNF blockade therapy in a subject with RA. Additionally, methods of diagnosis of RA in a subject and method of determining whether a subject is at risk for RA are provided.

Description

RHEUMATOID ARTHRITIS MARKERS

CROSS-REFERENCE TO RELATED APPLICATION

This application claims the benefit of U.S. Provisional Application No. 60/721,026, Filed September 27, 2005.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT The U.S. Government has a paid-up license in this invention and the right in limited circumstances to require the patent owner to license others on reasonable terms as provided for by the terms of Contract No. NOl-ARl -2256 awarded by The National Institutes of Health.

BACKGROUND OF THE INVENTION (1 ) Field of the Invention

The present invention generally relates to biological markers for diagnosis of disease and determining prognosis in disease treatment. More specifically, the invention is directed to utilization of biological markers for diagnosing rheumatoid arthritis (RA) and for evaluating the prognosis of using TNF blockade treatment for RA treatment. (2) Description of the Related Art

References cited

1. Gabriel S. The epidemiology of rheumatoid arthritis. Rheum. Dis. Clin. North Am. 2001; 27:269-281

2. Firestein GS. Evolving concepts of rheumatoid arthritis. Nature. 2003 May 15;423(6937):356-61.

3. Gregersen PK. Genetics of rheumatoid arthritis: confronting complexity. Arthritis Res 199; 1 :37-44, (http://artlmtis-research.eom/content/l/l/37^').

4. Seldin M, Amos C, Ward R, and Gregersen PK. The genetics revolution and the assault on rheumatoid arthritis. Arthritis Rheum 1999; 42:1071-1079, 1999 5. Vyse TJ, Todd JA. Genetic analysis of autoimmune disease. Cell 1996;85:311-8.

6. Wolfe F. The effect of smoking on clinical, laboratory, and radiographic status in rheumatoid arthritis. J Rheumatol 2000;27 :630-7.

7. Lipsky PE, van der Heijde DM, St Clair EW et al. Infliximaband methotrexate in the treatment of rheumatoid arthritis. Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group. N Engl J Med 2000; 343: 1594-1602. 8. W. Shao et al., J Biomed Biotechnol 2003, 299 (2003).

9. L. Perlee et al., Proteome Sci 2, 9 (2004).

10. Phillips DJ et al. Activin A: From sometime reproductive factor to genuine cytokine. Vet Immunol Immunopathol. 2005 Oct 18;108(l-2):23-7. 11. Abe, M., Shintani, Y., Eto, Y., Harada, K., Kosaka, M., Matsumoto, T., 2002. Potent induction of activin A secretion from monocytes and bone marrow stromal fibroblasts by cognate interaction with activated T cells. J. Leukoc. Biol. 72, 347-352.

12. Cho, SJH., Yao, Z.,Wang, S.-W., Alban, R.F., Barbers, R.G., French, S.W., Oh, C.K., 2003. Regulation of activin A expression in mast cells and asthma: its effect on the proliferation of human airway smooth muscle cells. J. Immunol. 170, 4045-4052.

13. Gribi, R., Tanaka, T., Harper-Summers, R., Yu, J., 2001. Expression of activin A in inflammatory arthropathies. MoI. Cell. Endocrinol. 180, 163-167.

14. Tong, S., Wallace, E.M., Burger, H. G., 2003. Inhibins and activins: clinical advances in reproductive medicine. Clin. Endocrinol. (Oxford) 58, 115-127. 15. Nu^'sing, R.M., Barsig, J., 1999. Induction of prostanoid, nitric oxide, and cytokine formation in rat bone marrow derived macrophages by activin A. Br. J. Pharmacol. 127, 919-926. 16. Ohguchi, M., Yamato, K., Ishihara, Y., Koide, M., Ueda, N., Okahashi, N., Noguchi, T., Kizaki, M., Ikeda, Y., Sugino, H., Nisihara, T., 1998. Activin A regulates the production of mature interleukin-lb and interleukin-1 receptor antagonist in human monocytic cells. J. Interferon Cytokine Res. 18, 491^-98.

Rheumatoid arthritis (RA) is a heterogeneous chronic inflammatory disorder that affects approximately 1% of the U. S. adult population (1). The cause is unknown, but autoimmune phenomena, in the form of autoantibodies, are characteristic of the disease (2). There is a genetic component to the disorder, and it aggregates in families with several other autoimmune disorders, including type 1 diabetes and autoimmune thyroid disease (3-5). Environmental factors, such as smoking, are also implicated in the etiology of RA, although the mechanism for this is unknown

(6).

Over the last decade or more, it has become widely accepted that certain inflammatory cytokines are involved in the pathogenesis of RA (2). The most prominent of these is tumor necrosis factor (TNF). TNF levels are elevated in inflammatory joints, and more importantly, therapeutic inhibition of TNF action (TNF blockade therapy) is an effective treatment for a substantial fraction of patients (7). However, there is wide inter-individual variation in the response to TNF inhibitors. In controlled clinical trials, approximately one-third of patients achieve a robust therapeutic response, achieving an ACR 50 response or better. In contrast, up to one-quarter of patients do not exhibit any significant clinical response to anti-TNF therapy. The remaining patients achieve various intermediate degrees of clinical benefit. There is currently no way to distinguish these clear "responder" and "non-responder" patient groups, much less to predict the degree of response to TNF blockade therapy in any particular individual.

Thus there is a need to determine whether a subject having RA is likely to respond to TNF blockade therapy. The present invention addresses that need.

SUMMARY OF THE INVENTION

Accordingly, the inventors have identified markers that are useful for diagnosing RA and for determining whether a subject with RA is likely to respond to TNF blockade therapy. Thus, in some embodiments, the invention is directed to methods of predicting responsiveness to TNF blockade therapy in a subject with rheumatoid arthritis (RA). The methods comprise determining whether the level of at least one marker is elevated in the subject compared to the corresponding average level in a sample of RA patients that are not responsive to TNF blockade therapy. In these methods, the marker is (a) transforming growth factor β receptor 3 (TGFβR3) protein in serum, or

(b) interleukin-6 receptor (IL-6R) mRNA in peripheral blood cells, or

(c) IL-6R protein in peripheral blood.

The elevated level of the marker is predictive that the subject will be responsive to TNF blockade therapy. In other embodiments, the invention is directed to methods of determining the effectiveness of TNF blockade therapy in a subject with rheumatoid arthritis (RA). In these methods, the subject has an elevated level of serum IL-6 before TNF blockade therapy. The methods comprise determining whether the subject's serum IL-6 level has fallen to the normal range within six weeks of commencement of TNF blockade therapy. In these methods, the normalization of IL-6 within six weeks of commencement of TNF blockade therapy indicates that the TNF blockade therapy is effective.

In further embodiments, the invention is directed to methods of diagnosis of rheumatoid arthritis (RA) in a subject. The methods comprise determining the level of IL-6 and/or gpl30 in the serum of the subject. In these methods, a level of IL-6 and/or gpl30 that is elevated in the serum above the normal level indicates that the subject has RA.

The present invention is additionally directed to methods of determining whether a subject is at risk for rheumatoid arthritis (RA). The methods comprise determining the level of IL-6 and/or gpl30 in the serum of the subject, where a level of IL-6 and/or gpl30 that is elevated in the serum above the normal level indicates that the subject is at risk for RA. A-

The invention is also directed to additional methods of predicting responsiveness to TNF blockade therapy in a subject with rheumatoid arthritis (RA). These methods comprise determining whether IL-6 pathway activity is elevated in the subject compared to the corresponding average IL-6 pathway activity in a sample of RA patients that are not responsive to TNF blockade therapy, where the elevated level of IL-6 pathway activity is predictive that the subject will be responsive to TNF blockade therapy.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is graphs of experimental results comparing IL-6R gene expression in peripheral blood RNA from ACR50 responders (higher expression) and non-responders (lower expression.

Two datasets were examined independently using the Affymetrix Ul 33 A (dataset 1) and the

Affymetrix U133 2.0 (dataset 2). Of 22,645 probesets in common between the two chips, IL-6R showed significant differences in both dataset 1 and 2 (P=0.009 and P=0.005 respectively,

Student's t-test). FIG 2 is a graph of experimental results showing serum gpl30 protein expression differs significantly when comparing RA patients (higher expression) and control subjects (lower expression), PO.0001 (t-test).

FIG. 3 is a graph of experimental results comparing serum TGFβR3 (betaglycan) protein expression in responder (ACR50) and non-responders at baseline, 6 weeks and 12 weeks after starting anti-TNF therapy. Values for control subjects are shown at right. ACR50 responders differ significantly (higher expression) from both non-responders and controls at baseline (P=0.04 and P=0.02, respectively), and when all time points are considered (P=0.0003 ACR50 vs NR;

P=0.0002 ACR50 vs. control)

FIG. 4 is a graph of experimental results showing that serum IL-6 is increased in RA patients mean (19.0) compared with controls (mean 6.5), P=0.029.

FIG.5 is a graph of experimental results showing serum IL-6 in responders (ACR50) and non-responders at baseline, 6 weeks and 12 weeks after starting anti-TNF therapy. Normalization of serum IL-6 occurs in a fraction of patients who are ACR50 responders and occurs by 6 weeks after start of therapy. Controls are shown at right. FIG. 6 is a graphic summarizing putative interactions between Activin/Inhibin and IL-6 pathways. The mechanism of inhibition of IL-6 action by activin has not been elucidated.

DETAILED DESCRIPTION OF THE INVENTION

Accordingly, the inventors have identified markers that are useful for diagnosing RA and for determining whether a subject with RA is likely to respond (show improvement) to TNF blockade therapy. See Example. Thus, in some embodiments, the invention is directed to methods of predicting responsiveness to TNF blockade therapy in a subject with rheumatoid arthritis (RA). The methods comprise determining whether the level of at least one marker is elevated in the subject compared to the corresponding average level in a sample of RA patients that are not responsive to TNF blockade therapy. In these methods, the marker is

(a) transforming growth factor β receptor 3 (TGFβR3) protein in serum, or

(b) interleukin-6 receptor (IL-6R) mRNA in peripheral blood cells, or

(c) IL-6R protein in peripheral blood.

The elevated level of the marker is predictive that the subject will be responsive (show a subjective or a clinical improvement in RA disability or severity) to TNF blockade therapy.

The TNF blockade therapy in these methods can be any such therapy known, including treatment with muscarinic agonists (see, e.g., U.S. Patent 6,610,713), fetuin (see, e.g., U.S. Patent Application 11/129,672, Filed May 13, 2005), adrenomedullin and/or adrenomedullin binding protein (see, e.g., U.S. Patents 6,864,237 and 6,884,781), or, preferably a treatment comprising administration of Infliximab (Remicade©), D2E7 (Humira™) and/or Etanercept (Enbrel©).

The skilled artisan could determine whether levels of any of the markers is elevated in a particular subject by comparing the level to the average level of the marker from (i) a sample of nonresponsive RA patients or (ii) a sample of subjects that do not have RA. If the level in the subject is higher at a set statistical significance (usually PO.05, but can be more or less stringent, e.g., PO.01 or P<0.10, depending on the Type I or Type II error to be tolerated), the subject would be deemed a responder, and TNF blockade therapy would likely be effective in the subject. The levels of the markers can be determined by any means known in the art. Preferably, the level of TGFβR3 protein is measured using an immunoassay. The level of EL-6R mRNA is preferably measured using a hybridization assay, as in the Example, or using an RT-PCR assay, as are known in the art. The level of IL-6R protein is preferably measured using an immunoassay. For the present invention, the levels of the marker need not be quantified (e.g., in μg/ml) if a control that represents the normal level is included in the test.

In some aspects of these methods, the marker is TGFβR3 protein in serum. In other aspects, the marker is IL-6R mRNA in peripheral blood cells. In still other aspects, the marker is IL-6R protein in peripheral blood. Additionally, any two or all three of these markers can be evaluated. In preferred embodiments, both (a) TGFβR3 protein and (b) IL-6R mRNA or IL-6R protein levels are determined.

The inventors have also discovered that RA patients that have an elevated level of IL-6, where the patients respond to TNF blockade therapy, have reduced levels of IL-6 within six weeks of commencement of the therapy, whereas nonresponders continue to have elevated levels after six weeks of therapy. See Example. Thus, in other embodiments, the invention is directed to methods of determining the effectiveness of TNF blockade therapy in a subject with rheumatoid arthritis (RA). In these methods, the subject has an elevated level of serum IL-6 before TNF blockade therapy. The methods comprise determining whether the subject's serum IL-6 level has fallen to the normal range within six weeks of commencement of TNF blockade therapy. In these methods, the normalization of IL-6 within six weeks of commencement of TNF blockade therapy indicates that the TNF blockade therapy is effective.

The level of serum IL-6 can be measured by any known method. Preferably, the IL-6 level is determined using an immunoassay, as in the Example. The inventors have additionally discovered that IL-6 or gpl30 levels are often elevated in patients that have or are at risk for RA. Thus, in further embodiments, the invention is directed to methods of diagnosis of rheumatoid arthritis (RA) in a subject. The methods comprise determining the level of IL-6 and/or gpl30 in the serum of the subject. In these methods, a level of IL-6 and/or gpl30 that is elevated in the serum above the normal level indicates that the subject has RA.

In some of these embodiments of these methods, IL-6 levels are determined; in other embodiments gpl30 levels are determined. However, in the most preferred embodiments, both IL-6 and gpDO levels are determined. These levels can be determined by any known method. Preferably, the method is an immunoassay. The present invention is additionally directed to methods of determining whether a subject is at risk for rheumatoid arthritis (RA). The methods comprise determining the level of IL-6 and/or gpl30 in the serum of the subject, where a level of EL-6 and/or gpl30 that is elevated in the serum above the normal level indicates that the subject is at risk for RA.

In some of these embodiments of these methods, IL-6 levels are determined; in other embodiments gpl30 levels are determined. However, in the most preferred embodiments, both IL-6 and gpl30 levels are determined. These levels can be determined by any known method. Preferably, the method is an immunoassay.

As discussed in the Example, the elevation of IL-6 and/or TGFβR3 levels in TNF blockade therapy responders likely reflects a general activation of the IL-6 pathway in those patients. Thus, the invention is also directed to additional methods of predicting responsiveness to TNF blockade therapy in a subject with rheumatoid arthritis (RA). These methods comprise determining whether IL-6 pathway activity is elevated in the subject compared to the corresponding average IL-6 pathway activity in a sample of RA patients that are not responsive to TNF blockade therapy, where the elevated level of IL-6 pathway activity is predictive that the subject will be responsive to TNF blockade therapy. In preferred embodiments, elevated IL-6 pathway activity is determined by measuring the level of at least one marker in the subject and comparing that level to the corresponding average level in a sample of RA patients that are not responsive to TNF blockade therapy, where the marker is (a) transforming growth factor β receptor 3 (TGFβR3) protein in serum, or

(b) interleuldn-6 receptor (IL-6R) mRNA in peripheral blood cells, or

(c) IL-6R protein in peripheral blood.

In these embodiments, the elevated level of the marker is indicative that the subject has elevated IL-6 pathway activity. In some aspects of these methods, the marker is TGFβR3 protein in serum. In other aspects, the marker is IL-6R mRNA in peripheral blood cells. In still other aspects, the marker is IL-6R protein in peripheral blood. Additionally, any two or all three of these markers can be evaluated. In preferred embodiments, both (a) TGFβR3 protein and (b) IL-6R mRNA or IL-6R protein levels are determined. Preferred embodiments of the invention are described in the following Example. Other embodiments within the scope of the claims herein will be apparent to one skilled in the art from consideration of the specification or practice of the invention as disclosed herein. It is intended that the specification, together with the examples, be considered exemplary only, with the scope and spirit of the invention being indicated by the claims, which follow the examples.

Example 1. Identification of biomarkers that are predictive of clinical response to anti-TNF therapy in rheumatoid arthritis

We now report that a group of analytes in serum, combined with patterns of gene expression in peripheral blood, is strongly associated with responder status. These data provide insight into the pathways that underlie TNF dependent rheumatoid arthritis, and form the basis of clinical diagnostic tests that will be useful for stratifying patients into those who will respond, or not respond, to TNF inhibition. Materials and Methods

Patient population. One hundred and twenty six rheumatoid arthritis patients were enrolled in a longitudinal study of response to TNF inhibition. Table 1 summarizes the characteristics of the patient cohort. All subjects were studied at baseline, 6 weeks and 12 weeks after beginning anti-TNF therapy (40% Remicade, 40% Enbrel and 20% Humira). Out of 97 subjects with full data, 27 subjects achieved an ACR50 response at 12 weeks, while 20 subjects failed to achieve an ACR20 response and were classified as "non-responders". Gene expression analysis in peripheral blood. Peripheral blood RNA was collected in PaxGene tubes (Qiagen Inc.) and prepared according to the manufacturer's standard protocol. Gene expression was examined using the Affymetrix U133A or U133 2.0 platform. RNA amplification was performed using protocols and reagents supplied by Affymetrix using regents from Invitrogen and Enzo for the U133A array and from Ambion for the U133 plus arrays. Affymetrix microarray suite (MAS) 5.0 Software was used to obtain gene expression (signal) values for each gene. To permit accurate comparison between chips and to correct for minor variations in the overall intensity of hybridization, each chip was scaled to an intensity of 1,500 units. Sandwich ELISA assays on serum. The levels of 173 serum protein analytes were measured in serum aliquots (lOOμl) from the cases and controls using custom dual antibody sandwich immunoassay arrays, as described (8,9) (see Table 2). Briefly, monoclonal capture antibodies specific for each analyte were fixed to glass slides, with 12 replicate spots for each analyte. Duplicate samples of sera were incubated for 2 hours, and then washed. Slides were then incubated with secondary biotinylated polyclonal antibodies, and signals were amplified using a 'rolling circle' method (8,9). Quality control measures included optimization of antibody pairs, the use of internal controls to minimize array-to-array variation, and standardized procedures of chip manufacturing (8,9). Arrays were scanned using a Tecan LS200 unit and Mean Fluorescence Intensities (MFIs) were generated with customized software. To convert MFIs to concentration values, 15 serial dilutions of recombinant analytes at known concentrations (studied in parallel on each slide) were used to develop best-fit equations for each analyte. The upper and lower limits of quantitation were defined to ensure a dynamic working range. There were an additional 12 analytes measured on the protein microarrays (beyond the 160) for which > 80% of samples were at the upper or lower limits of detection; these were excluded from further consideration.

Statistical analysis. Gene expression data was analyzed using logistic regression. Logistic regression is a statistical discriminant model, in which the probability for a sample to be in one of the two classes (for example responder or non-responder groups of RA patients) is given by an equation ("model") that contains adjustable parameters. We utilized the equation given as:

1 Prob(sample labeHRA) =

_{1 + e}-a- (ox)

where a and c are the two adjustable parameters in the model, and x is a quantity related to the rnRNA expression level (in this case, the logarithm of the fluorescence intensity). After the training, the two parameters choose the value a = S and c= S. In order to measure the classification performance of the fitted statistical model, we determined the "deviance", defined as the sum of squares of residuals (difference between the fitted-model- predicted and the true sample label value). The smaller the deviance, the better the model fits the data. Since at the expression level x = - a / c, the probability of the sample label to be both RA and control, according to the model, is 1/2, - a^" / 6 is the threshold value. If the model classifies the data perfectly, the expression levels of RA samples and control samples would be on two sides of the threshold value, respectively, without exception. Logistic regression has been applied previously to microarray data analysis. Results

IL-6 receptor gene expression distinguishes ACR50 responders vs. non responders to anti-TNF therapy. A total of 20 responders (ACR50), and 12 non-responder (NR) subjects were analyzed for gene expression at baseline, prior to beginning anti-TNF therapy. RNA samples from 16 subjects (8 ACR50, 8 NR) were analyzed using the U133A Affymetrix chip, and RNA samples from the remaining 16 subjects (12 ACR50, 4 NR) were analyzed using the U133 2.0 Affymetrix chip. For the initial analysis, only probe sets that corresponded to genes in common between the two chips were analyzed (22,469 probe sets). Each comparison between ACR50 responders and NR was performed within each dataset for each chip type, using logistic regression. For each dataset, genes were ranked according to performance in logistic regression and we selected probesets that achieved a combined ranking of <200 in the two datasets. A probeset corresponding to IL-6 receptor met this criterion. A subsequent analysis of the two datasets using a students t-test shows a clear difference in IL-6 receptor expression, as shown in Figure 1. Both datasets show significantly higher levels of IL-6 receptor gene expression in ACR50 responders, compared with non-responders (P=0.009 and P=0.005 in datasets 1 and 2, respectively).

Serum levels of TGFβR3 (betaglycari) distinguish ACR50 responders vs. non responders to anti-TNF therapy. We carried out a broad survey of a panel of 160 serum analytes in 16 subjects with rheumatoid arthritis. These subjects included ACR50 responders (n=7) and non- responders (n=5). In addition, 16 control subjects were assayed for these analytes. Patients were analyzed at baseline, 6 weeks and 12 weeks for all analytes. When comparing all cases vs. controls, soluble gpl30 was found to be most significantly different among the case and control groups (pO.OOOl, Student's t-test), as shown in Figure 2. However, ACR50 responder and non responder groups did not differ with respect to the levels of gpl30, either at baseline or at 6 weeks and 12 weeks after TNF therapy (data not shown). In contrast, when comparing ACR50 responder and non responder groups, the serum levels of TGFβR3 (betaglycan) were significantly different, with marked elevation of this analyte in the ACR50 responder group. For baseline samples, the mean TGFβR3 level in the responder group was 37,000, compared with 20,000 in the non-responder group (PO.04). This difference persisted at all time points, so that the combined data increased significance to PO.0003). In addition, compared with control subjects, ACR50 responders had significantly elevated TGFβR3 levels at both baseline (P<0.3) and when all time points were combined (PO.0002). These data are shown in Figure 3.

IL-6 levels change in response to anti-TNF therapy in a subset of ACR50 responders. The serum concentration of IL-6 was found to be significantly different when comparing RA patients and control subjects (P=0.029), although the majority of RA patients did not show substantial elevation of this cytokine, as shown in Figure 4. There was no significant difference in baseline IL-6 serum levels in ACR50 responder vs. non-responder subjects. However, among the three ACR50 responders with high IL-6 serum levels, all returned to normal levels after TNF therapy by 6 weeks. In contrast, 2/5 non-responder subjects exhibited high IL-6 serum levels even after 6 weeks of anti-TNF therapy, and elevation of IL-6 levels persisted at 12 weeks in non- responders. These data are shown in Figure 5. Overall, these data suggest that early reduction of IL-6 levels may be a marker for anti-TNF responsiveness in a subset of RA patients who exhibit high serum levels of this cytokine. Discussion

In this study, we have combined the analysis of peripheral blood gene expression with measurement of serum proteins in order to develop a set of biomarkers that may be useful in identifying subsets of RA patients, and providing insight into the mechanisms of disease pathogenesis in this heterogeneous disorder. Strikingly, we have identified a group of biomarkers that collectively suggest that alterations in IL-6 and activin/inhibin pathways may be useful for distinguishing subjects who will respond to TNF blockade.

First, peripheral blood gene expression of IL-6 receptor appears to distinguish the ACR50 responder vs. non-responder subgroups. IL-6 receptor may exist as both a membrane-bound and soluble form, both of which result in positive signaling when combined with IL-6, acting through the gpl30 membrane-bound signaling chain of the complete IL-6 receptor complex. It is currently unclear whether this elevation in gene expression leads to either increased membrane or soluble IL-6R, or both. This issue can be addressed by developing assays for measurement of serum IL-6R, whether bound to circulating IL-6 or free. In contrast, soluble gpl30 is elevated in all RA patients, regardless of responder status. Since soluble gpl30 negatively regulates IL-6 activity, this may reflect an apparently ineffective compensatory mechanism to downregulate IL-6 action in patients with active inflammation. Interestingly, soluble gpl30 levels are not affected by anti-TNF therapy, suggesting that this pathway may be upstream of TNF action.

IL-6 itself is modestly elevated in a subset of RA patients. Elevations of IL-6 are present in the inflamed joint, and serum levels may reflect production at the inflammatory site. However, if so, it is an insensitive indicator of such inflammation, since it not elevated in most RA patients, even in the setting of active disease. Although serum IL-6 normalizes in responders to anti-TNF therapy, IL-6 levels cannot be used to distinguish responders vs. non-responders prior to starting therapy.

The presence of elevations in TGFβR3 levels in RA patients is an entirely novel finding, and is consistent with a role for IL-6 pathways in this disease. TGFβR3, also known as betaglycan, is a coreceptor for inhibin, a member of the TGFβ family of cytokines (10,14). A major role for the inhibin/betaglycan complex is to inhibit the action of activin, another member of the TGFβ family. The inhibin-activin pathway is known to have a profound effect on regulating the action of IL-6 on a variety of cell types, including B cells, monocytes and liver cells (10). Thus, the primary action of activin is to inhibit IL-6 action on these cells, and a role of inhibin/betaglycan is in part to counter-regulate this inhibitory action of activin. These interactions are summarized in Figure 6. Thus, an elevated level of inhibin/betaglycan would be predicted to release IL-6 from activin inhibition, and thus should be a marker for increased activity in the IL-6 pathway. However, research into the role of activin in the inflammatory response is in its infancy

(10), and it remains possible that these effects on the IL-6 pathway are not the primary explanation for the association between elevated betaglycan levels and response to TNF inhibition. For example, activin is also associated with pro-inflammatory effects, including the stimulation of release of TNF and ILl from macrophages (15,16). Indeed, activin is one of the first serum constituents to appear after LPS injection in experimental animals, even before the appearance of TNF. Activin has also been described in the inflamed joint tissue of rheumatoid arthritis patients, and induction and release of activin from mast cells has been described in asthma models (11-13). Thus, the elevation of betaglycan in a subset of RA patients may reflect a compensatory mechanism to control activin-mediated inflammation. Again, our data suggests that this subset is particularly responsive to therapeutic TNF inhibition.

Since activin and inhibin were originally described as regulators of FSH release (14), it is interesting to speculate on the relationship of this pathway to the female predominance of RA, and particularly on their possible role in the phenomenon of RA disease remission during pregnancy. Both activin and inhibin are elevated during pregnancy, and it will be of great interest to examine whether these molecules, or others in the pathway, such as follistatin, may mediate this effect. This may point the way to the development of new therapeutics based on the activin/inhibin system.

As a whole, these data point the potential use of betaglycan, as well as other serum analytes such as activin, follistatin and the newly described follistatin-like protein as diagnostic biomarkers to predict response and manage treatment of rheumatoid arthritis. In addition, exploration of these compounds as potential therapeutics in animal models of inflammatory arthritis are clearly warranted.

Table 1. Characteristics of 97 RA subjects undergoing treatment with TNF inhibitors

Mean age 55.8 years (range 23-89)

Anti-CCP 40% positive

Shared epitope + 67%

Female/Male 74/23 = 3.2

Table 2. List of serum protein analytes examined in this study. Only TGFβR3 distinguished ACR50 responders from non-responders to anti-TNF therapies.

4-1BB

6Ckine

Activated leukocyte cell adhesion molecule

Agouti-related protein

Alpha fetoprotein

Amphiregulin

Angiogenin

Angiotensin converting enzyme

Angiotensin I converting enzyme-2

Betacellulin beta-nerve growth factor

B-lymphocyte chemoattractant

Brain-derived neurotrophic factor

Cancer antigen 125

CD27

CD30

CD40

Ciliary neurotrophic factor

Ciliary neurotrophic factor receptor alpha

C-reactive protein

Cutaneous T-cell attracting chemokine

D-Dimer

Death receptor 6

Endostatin

Endothelin 3

Eotaxin

Eotaxin-2

Eotaxin-3 Epidermal growth factor

Epidermal growth factor receptor 1

Epidermal growth factor receptor 2

Epithelial cell-derived neutrophil-activating peptide

E-selectin

Fas (CD95)

Fas ligand

Fibroblast growth factor acidic

Fibroblast growth factor receptor 3 IHb isoform

Fibroblast growth factor receptor 3 IHc isoform

Fibroblast growth factor-2 (FGF-basic)

Fibroblast growth factor-4

Fibroblast growth factor-6

Fibroblast growth factor-7

Fibroblast growth factor-9 fins-like tyrosine kinase-3 ligand

Follistatin

Fractalkine

Glial cell line derived neurotrophic factor

Granulocyte chemotactic protein 2

Granulocyte colony stimulating factor

Granulocyte macrophage colony stimulating factor

Growth related oncogene beta

Growth related oncogene gamma

Heat shock protein 70

Hemofiltrate CC chemokine 1

Hemofiltrate CC chemokine 4

Heparin-Binding EGF-like Growth Factor

Hepatocyte growth factor

Herpes virus entry mediator

Human chorionic gonadotropMn

Human Protein C

Human Protein S

1-309

Insulin-like growth factor binding protein 1

Insulin-like growth factor binding protein 2

Insulin-like growth factor binding protein 3

Insulin-like growth factor binding protein 4

Insulin-like Growth Factor Binding Protein 6

Insulin-like growth factor I receptor

Insulin-like growth factor II

Intercellular adhesion molecule 1

Intercellular adhesion molecule 3

Interferon alpha

Interferon gamma

Interferon gamma-inducible T cell alpha chemoattractant

Interferon omega

Interleukin 1 alpha Interleukin 1 beta

Interleukin 1 receptor 4

Interleukin 1 receptor antagonist

Interleukin 1 soluble receptor I

Interleukin 1 soluble receptor II

Interleukin 10 receptor beta

Interleukin 12 p40

Interleukin 13

Interleukin 15

Interleukin 16

Interleukin 17

Interleukin 18

Interleukin 2

Interleukin 2 receptor beta

Interleukin 2 receptor gamma

Interleukin 2 soluble receptor alpha

Interleukin 3

Interleukin 4

Interleukin 5

Interleukin 5 receptor alpha

Interleukin 6

Interleukin 7

Interleukin 8

Interleukin 9

Leptin

Leukemia inhibitory factor

Leukemia inhibitory factor soluble receptor alpha

Leukocyte selectin

Lymphotactin lymphotoxin-beta receptor

Macrophage colony stimulating factor

Macrophage colony stimulating factor receptor

Macrophage inflammatory protein 1 alpha

Macrophage inflammatory protein 1 beta

Macrophage inflammatory protein 1 delta

Macrophage inflammatory protein 3 alpha

Macrophage inflammatory protein 3 beta

Macrophage migration inhibitory factor

Matrix metalloproteinase 1

Matrix metalloproteinase 10

Matrix metalloproteinase 2

Matrix metalloproteinase 7

Matrix metalloproteinase 9

Matrix Metalloproteinase-8

Monocyte chemotactic protein 1

Monocyte chemotactic protein 2

Monocyte chemotactic protein 3

Monocyte chemotactic protein 4 Monokine induced by interferon gamma

Myeloid progenitor inhibitory factor 1

Neurotrophin 3

Neurotrophin 4

Neutrophil Activating Peptide 2

Neutrophil elastase

Oncostatin M

Osteopontin

Pigment epithelium-derived factor

Placental growth factor

Plasminogen activator inhibitor type 1

Plasminogen activator inhibitor-II

Platelet endothelial cell adhesion molecule- 1

Platelet-derived growth factor receptor alpha

Platelet-derived growth factor receptor beta

Prolactin

P-Selectin

Pulmonary and activation-regulated chemokine

Receptor activator of NF-kappa-B

Regulated upon activation, normal T expressed and presumably secreted

Soluble glycoprotein 130

Soluble Vascular Adhesion Protein- 1

Soluble CD44var (v6)

Soluble VCAM-I

Stem cell factor

Stem cell factor receptor

Stromal cell-derived factor beta

Survivin

Thrombomodulin/CD141

Thymus and activation regulated chemokine

Thyroid stimulating hormone

Tissue inhibitors of metalloproteinases 1

Tissue inhibitors of metalloproteinases 2

TNF-related apoptosis-inducing ligand receptor 1

TNF-related apoptosis-inducing ligand receptor 4

Transforming growth factor alpha

Transforming growth factor beta receptor III

Transforming growth factor beta-3

Tumor necrosis factor alpha

Tumor necrosis factor beta

Tumor necrosis factor receptor I

Tyrosine kinase with Ig and EGF homology domains 2

Urokinase plasminogen activator

Urokinase plasminogen activator receptor

Vascular Endothelial Cadherin

Vascular endothelial growth factor

Vascular endothelial growth factor receptor 2

Vascular endothelial growth factor-D VEGF receptor 3

In view of the above, it will be seen that the several advantages of the invention are achieved and other advantages attained.

As various changes could be made in the above methods and compositions without departing from the scope of the invention, it is intended that all matter contained in the above description and shown in the accompanying drawings shall be interpreted as illustrative and not in a limiting sense.

All references cited in this specification are hereby incorporated by reference. The discussion of the references herein is intended merely to summarize the assertions made by the authors and no admission is made that any reference constitutes prior art. Applicants reserve the right to challenge the accuracy and pertinence of the cited references.

Claims

What is claimed is:

1. A method of predicting responsiveness to TNF blockade therapy in a subject with rheumatoid arthritis (RA), the method comprising determining whether the level of at least one marker is elevated in the subject compared to the corresponding average level in a sample of RA patients that are not responsive to TNF blockade therapy, wherein the marker is

(a) transforming growth factor β receptor 3 (TGFβR3) protein in serum, or

(b) interleukin-6 receptor (IL-6R) mRNA in peripheral blood cells, or

(c) IL-6R protein in peripheral blood; and wherein the elevated level of the marker is predictive that the subject will be responsive to TNF blockade therapy.

2. The method of claim 1, wherein the marker is TGFβR3 protein in serum.

3. The method of claim 1 , wherein the marker is IL-6R mRNA in peripheral blood cells.

4. The method of claim 1, wherein the marker is IL-6R protein in peripheral blood.

5. The method of claim 2, wherein the level of TGFβR3 protein is measured using an immunoassay.

6. The method of claim 3, wherein the level of IL-6R mRNA is measured using a hybridization assay.

7. The method of claim 3, wherein the level of IL-6R mRNA is measured using an RT-

PCR assay.

8. The method of claim 4, wherein the level of IL-6R protein is measured using an immunoassay.

9. The method of claim 1, wherein (x) TGFβR3 protein and (y) IL-6R mRNA or IL-6R protein levels are determined.

10. A method of determining the effectiveness of TNF blockade therapy in a subject with rheumatoid arthritis (RA), where the subject has an elevated level of serum IL-6 before TNF blockade therapy, the method comprising determining whether the subject's serum IL-6 level has fallen to the normal range within six weeks of commencement of TNF blockade therapy, wherein such normalization of IL-6 within six weeks of commencement of TNF blockade therapy indicates that the TNF blockade therapy is effective.

11. The method of claim 10, wherein IL-6 levels are determined using an immunoassay.

12. A method of diagnosis of rheumatoid arthritis (RA) in a subject, the method comprising determining the level of IL-6 and/or gpl30 in the serum of the subject, wherein a level of IL-6 and/or gpl30 that is elevated in the serum above the normal level indicates that the subject has RA.

13. The method of claim 12, wherein IL-6 levels are determined.

14. The method of claim 12, wherein gpl30 levels are determined.

15. The method of claim 12, wherein both IL-6 and gpl30 levels are determined.

16. The method of claim 12, wherein levels of IL-6 and/or gpl30 are determined using by immunoassay.

17. A method of determining whether a subject is at risk for rheumatoid arthritis (RA), the method comprising determining the level of IL-6 and/or gpl30 in the serum of the subject, wherein a level of IL-6 and/or gpl30 that is elevated in the serum above the normal level indicates that the subject is at risk for RA.

18. The method of claim 12, wherein IL-6 levels are determined.

19. The method of claim 12, wherein gpl30 levels are determined.

20. The method of claim 12, wherein both EL-6 and gpl30 levels are determined.

21. The method of claim 12, wherein levels of IL-6 and/or gpl30 are determined using by immunoassay.

22. A method of predicting responsiveness to TNF blockade therapy in a subject with rheumatoid arthritis (RA), the method comprising determining whether IL-6 pathway activity is elevated in the subject compared to the corresponding average IL-6 pathway activity in a sample of RA patients that are not responsive to TNF blockade therapy, wherein the elevated level of IL- 6 pathway activity is predictive that the subject will be responsive to TNF blockade therapy.

23. The method of claim 22, wherein elevated IL-6 pathway activity is determined by measuring the level of at least one marker in the subject and comparing that level to the corresponding average level in a sample of RA patients that are not responsive to TNF blockade therapy, wherein the marker is

(a) transforming growth factor β receptor 3 (TGFβR3) protein in serum, or (b) interleukin-6 receptor (IL-6R) mRNA in peripheral blood cells, or

(c) IL-6R protein in peripheral blood; and wherein the elevated level of the marker is indicative that the subject has elevated IL-6 pathway activity.

24. The method of claim 23, wherein the marker is TGFβR3 protein in serum.

25. The method of claim 23, wherein the marker is IL-6R mRNA in peripheral blood cells.

26. The method of claim 23, wherein the marker is IL-6R protein in peripheral blood.