Table 2.
Frequency of sera being positive in one test system and negative in another test systema
| SLE patients (N = 207) | NcX+ (n = 138) |
dsDNA+ (n = 86) |
Nuc+ (n = 115) |
Farr+ (n = 115) |
CLIF+ (n = 58) |
|---|---|---|---|---|---|
| NcX- (n = 69) | - | 1% | 0% | 13% | 7% |
| dsDNA- (n = 121) | 44% | - | 33% | 35% | 14% |
| Nuc- (n = 92) | 25% | 12% | - | 21% | 8% |
| Farr- (n = 92) | 35% | 14% | 21% | - | 5% |
| CLIF- (n = 149) | 57% | 30% | 43% | 57% | - |
aSLE, systemic lupus erythematosus; NcX, anti-dsDNA-nucleosome-complexed enzyme-linked immunosorbent assay (ELISA); dsDNA, double-stranded DNA; Nuc, anti-dsDNA-nucleosome ELISA; Farr, radioimmunoassay; CLIF, Crithidia luciliae immunofluorescence assay. Cutoffs for all test systems were read out of receiver operating characteristic curve analysis at a specificity of 98.15% to include the CLIF assay in that analysis. Frequencies of sera being positive in one selected test (column headings) and negative in another test system (left column stub headings) were filled to demonstrate the extent of cross-reactivity between tests. The positive sera were compared with the negative sera to calculate percentages. For example, among all anti-dsDNA-NcX ELISA-negative sera (n = 69), 1% were positive on the basis of the anti-dsDNA ELISA.