Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.

Post-Treatment Changes in Serum C-Reactive Protein Levels and Clinical Response in Rheumatoid Arthritis.

Emery8,  Paul, Matteson6,  Eric L., Genovese7,  Mark C., Sague1,  Sarah, Hsia3,  Elizabeth C., Doyle4,  Mittie K., Fan2,  Hongtao

Centocor Research and Development, Inc., Malvern, PA
Centocor Research and Development, Inc.
Centocor Research and Development, Inc./Univ of Pennsylvania School of Medicine, Malvern, PA
Centocor Research and Development, Inc./Univ. of Pennsylvania School of Medicine, Malvern, PA
Elashoff Consulting
Mayo Clinic, Rochester, MN
Stanford Universit, Sunnyvale, CA
University Leeds, Leeds, United Kingdom


With the availability of many effective therapies for RA, biomarker(s) that can reliably predict response to a treatment (tx) would be useful in clinical practice. Serum CRP levels have been known to be associated with disease activity including radiographic progression. It is however not clear if early changes in serum CRP levels in response to tx can predict sustained clinical response at later time points.


To evaluate if early post-tx changes in CRP can reliably predict clinical response, we evaluated the association of early post-tx changes in serum CRP levels with clinical response at a later time point.


Sera were collected at wk0, 4 & 24 from GO-BEFORE (MTX naïve RA pts; 477 golimumab[GLM]+/-MTX, 160PBO+MTX), GO-FORWARD (active RA pts despite MTX; 311GLM+/-MTX, 133PBO+MTX) and GO-AFTER (active RA pts previously tx with TNF inhibitors; 304GLM, 155PBO) studies. Samples were tested for CRP levels using Roche Tinaquant assay. CRP levels at baseline (BL), wk4 & 24 in GLM±MTX and PBO+/-MTX pts achieving and not achieving multiple measures of clinical response (ACR & EULAR/DAS response criteria) were evaluated. Logistic regression models were used to test for marker associations with clinical endpts. Positive predictive value (PPV) & Negative predictive value (NPV) were calculated.


The figure shows CRP levels at wk0, 4&24 among remission (DAS28-ESR<2.6) achievers and non-achievers at wk24 (as a representative data; data for other clinical measures of response were similar but not shown). Pts achieving remission tend to have lower CRP at BL, wk4&24, compared to pts not achieving remission, & pts treated with GLM had in general lower CRP levels than pts treated with MTX or PBO. Changes from BL in CRP levels were analyzed in 3 RA populations and decreases from BL at wk4 in pts with BL CRP>1 mg/dL were significantly associated with remission at wk24 in GLM±MTX pts from GO-BEFORE (OR=1.4; p<0.002; PPV=67%; NPV=53%), GO-FORWARD (OR=2.1; p<0.000;PPV=70%; NPV=68%) and GO-AFTER (OR=1.3; p=0.06; PPV=51%; NPV=64%). Reductions from BL CRP of >50% at wk4 in these pt subsets were also signif. associated with multiple measures of clinical response including remission (data not shown).


While changes in CRP levels associate with clinical response and may be a good measure of disease activity in RA at any time point, it is not a reliable predictor of clinical response to tx.

To cite this abstract, please use the following information:
Emery, Paul, Matteson, Eric L., Genovese, Mark C., Sague, Sarah, Hsia, Elizabeth C., Doyle, Mittie K., et al; Post-Treatment Changes in Serum C-Reactive Protein Levels and Clinical Response in Rheumatoid Arthritis. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :326
DOI: 10.1002/art.28095

Abstract Supplement

Meeting Menu