Page 30 - THE EVOLUTION OF EARLY ARTHRITIS AND CARDIOVASCULAR RISK Samina A. Turk
P. 30

CHAPTER 2
  Risk Factor
Autoantibodies
Other biomarkers in blood
Imaging
Symptoms
Comments
Status and levels of (isotypes of) RF and ACPA associate with RA risk(131– 143)
Higher levels and the combination of RF and ACPA confer a higher risk(144,145)
Additional predictive ability independent of RF and ACPA was shown for anti–carbamylated protein antibodies(146) and anti–peptidyl arginine deiminase type 4 antibodies(147)
Several acute phase reactants and cytokines are increased in pre-RA or at- risk cohorts(1,148–162)
TNF (receptor), cartilage oligomeric matrix protein, and a high interferon gene score are quantified risk factors(163,164)
Ultrasonography abnormalities (mainly power Doppler signal) in seropositive patients with arthralgia were predictive of arthritis at the joint level in 1 study(165) and at the patient level in another study(166) Technetium bone scintigraphy is predictive of RA in patients with arthralgia(167) and can exclude inflammatory joint disease(168) Macrophage-targeted positron emission tomography predicts arthritis in ACPA-positive patients with arthralgia(169)
The predictive capacity of MRI in arthralgia is not yet clear(170,171)
Predictive symptoms in combination with the presence of autoantibodies: duration <12 mo, intermittent symptoms, arthralgia in upper and lower extremities, morning stiffness 1 h, self-reported joint swelling,(145) tenderness of hand or foot joints, and morning stiffness 30 min(166)
          Abbreviations: ACPA, anti–citrullinated protein antibody; HIV, human immunodeficiency vi- rus; MRI, magnetic resonance imaging; RF, rheumatoid factor; TNF, tumor necrosis factor
PREDICTION RULES: PUTTING THE BLOCKS TOGETHER
In a manner similar to the way clinical characteristics, signs, and symptoms can be combined to diagnose a disease in a patient, the potential risk factors for a given disease can be combined by statistical modeling of variables measured in an at-risk population in order to produce prediction rules. The advantage of such models is that they clarify the relative impact of the individual variables and quantify the overall risk for individuals coming from that population. The validity of these models can then be further confirmed by testing them in other populations.
Recently, several prediction models have been published that attempt to quantify progression to RA (Table 2). Two of these models were based on large population studies, of which 1 was designed for investigating other diseases as well. One of these used clinical characteristics to predict either seropositive or seronegative RA,(29) the other used the combination of clinical characteristics, autoantibodies, and a genetic risk score containing multiple genes (see Table 2 for the variables in the models).(172) Both studies achieve good prediction. However, it is uncertain whether these values can be reproduced in smaller populations.
28















































































   28   29   30   31   32