Background Cell Bound Complement Activation Products (CB-CAPs), are stable form of classical complement activation, ex-vivo, and sensitive and specific marker of SLE. In the present study, we sought to compare the performances of CB-CAPs to gold standard low complement C3 or C4.

Methods All subjects (n=1200) were adults (18 years) and enrolled from multiple academic Centers in the United States. All SLE fulfilled the 1997 ACR criteria for SLE (n=498). Patients with Other Rheumatic Diseases (n=450) consisted of 189 rheumatoid arthritis, 88 Sjogrens, 90 Fibromyalgia and 83 patients with other connective tissues diseases. A group of healthy normal individuals was also enrolled (n=252). Abnormal CB-CAPs status (EC4d or BC4d>99 th percentile of normal) were determined using flow-cytometry. Complement C3 and C4 levels were determined using immunoturbidimetry (Binding Site, San Diego, CA) assay kits. Performances of the markers, either alone or in combination to distinguish SLE from other rheumatic diseases and controls were established using Sensitivity, Specificity, Odds Ratio and Area under the Curve (AUC) of the Receiver Operating Characteristic curve (ROC). Youden Index (Sensitivity +Specificity - 100) was also calculated. The combination of 4 complement marker abnormalities were also evaluated using logistic regression and unweighted composite score cumulating the presence of these abnormalities was calculated.

Results Abnormal CB-CAPs status yielded 62% sensitivity with 88% specificity in distinguishing SLE from the group of patients with other diseases (table 1). Youden Index was 0.492±0.027. Low C3/C4 status yielded 38% sensitivity and 93% specificity in distinguishing SLE from the group of patients with other diseases. Youden Index was 0.313±0.025 for Low C3 or C4 and significantly lower than Youden score associated with abnormal CB-CAPs (p<0.01). Specificity of Low C3/C4 and abnormal CB-CAPs in distinguishing SLE from normal healthy individuals was 93% and 99%, respectively. AUC was also significantly higher with BC4d (0.718) than with low C3 (0.620; p<0.01), low C4 (0.618; p<0.01) and Low C3 or C4 status (0.656; p<0.01). A composite score (unweighted) cumulating all 4 abnormalities, was higher in SLE (1.47±0.06) than disease control group (0.21±0.02) (p<0.01) and normal healthy individuals (0.01±0.02) (p<0.01). The complement scoring system yielded higher AUC (0.812), higher OR (36.0 CI95%: 18.8–69.0), lower AIC (1037) and greater R2 (0.403) than any other combinations

Conclusions Our data suggests that CB-CAPs have greater diagnostic yield than low complement C3/C4. The combination of these complement abnormalities in composite complement score has high yield in distinguishing SLE from other rheumatic diseases and normal healthy individuals.

Funding Source(s): Exagen.