The diagnostic utility of serology testing for predicting the risk of cancer-associated myositis

Background: The study objective was to determine whether the risk of developing myositis in association with a cancer is predictable by serological profiling.Methods: A cross-sectional study of 280 UK Caucasian adults with polymyositis (PM, n=106), dermatomyositis (DM, n=103) and myositis in overlap with another connective tissue disease (myositis/CTD-overlap, n=71) was performed. Patients were tested for myositis-specific antibodies (MSAs) (anti-Jo-1, PL-7, PL-12, EJ, OJ, KS, Mi-2, SRP), myositis-associated autoantibodies (MAAs) (anti-U1-RNP, U3-RNP, Ku, PM-Scl) and the recently described antibody (Ab) directed against 155kDa and 140kDa specificities (anti-155/140 Ab) (Targoff et al, 2006; Kaji et al, 2006). Cancer-associated myositis (CAM) was defined as cancer occurring in myositis patients within three years of disease onset (Troyanov et al, 2005). Sensitivity and specificity analyses of various combinations of serological testing were performed for optimal identification of cancer risk.Results: Sixteen patients (5.7%) had cancer-associated myositis (CAM) (15 [14.6%] DM, 0 PM, 1 [1.4%] myositis/CTD-overlap). CAM patients were significantly older compared to non-CAM patients (CAM 59 years vs. non-CAM 48 years, p=0.02), and most did not possess an MSA/MAA. In contrast, anti-155/140 Ab represented a significant risk factor for CAM compared to patients without this Ab (risk of CAM: anti-155/140 Ab positive vs. anti-155/140 Ab negative, p=0.0009, odds ratio 23, 95% confidence interval 6.1-83.8), and was exclusively found in DM. No patients with anti-155/140 Ab had interstitial lung disease. A positive anti-155/140 Ab result proved highly specific and moderately sensitive with a high negative predictive value (NPV) for CAM. A negative result on a routine myositis Ab panel (anti-Jo-1, PM-Scl, U1/U3-RNP, Ku) was highly sensitive with high NPV for CAM. The combination of testing a negative routine myositis Ab panel with a positive anti-155/140 Ab result was 100% sensitive, detecting all CAM patients with 100% NPV.Conclusions: We have demonstrated serological differences between CAM and non-CAM patients, where an absence of MSA/MAAs on routine testing increases the likelihood of CAM. The addition of anti-155/140 Ab testing further aids the prediction of CAM. These results will help identify those myositis patients at greater risk of malignancy and who require more intensive cancer surveillance. When routine myositis serology in adult myositis patients is negative, extra vigilance is required in screening for co-existent cancers.Utility of diagnostic serology tests to predict cancer-associated myositis Sensitivity Specificity Positive predictive value Negative predictive value ROC area Detection of anti-155/140 Ab 50.0 95.8 42.1 96.9 0.73 Negative result on routine Ab panel 93.8 45.5 9.4 99.2 0.70 Negative result on routine Ab panel and detection of anti-155/140 Ab 100 45.1 9.9 100 0.73 ROC=receiver-operator characteristic (sensitivity + specificity / 2)