Tumour-associated autoantibodies may diagnose lung cancer
Tumour-associated autoantibodies (TAAbs), whether alone or in combination, may detect lung cancer at any stage, according to a new meta-analysis.
The meta-analysis involved 65 eligible articles, of which 31 focused on single autoantibodies while 39 examined multiplex autoantibodies. Five studies involved both single and multiplex autoantibodies. Enzyme-linked immunosoassay (ELISA) was the most common technique for detecting autoantibodies.
Pooled analysis of 38 tests showed that the sensitivities of 34 specific single TAAbs for diagnosing lung cancer ranged from 13.8 to 99 percent, with a mean sensitivity of 55.2 percent. The mean specificity was 84.4 percent and values ranged from 19.7 to 100 percent.
Results from 33 tests showed that the mean sensitivity of mixed TAAbs for detecting lung cancer was 70.3 percent with values ranging from 30 to 100 percent. Mean specificity was 86.3 percent and values ranged from 43 to 97.3 percent.
The diagnostic value of a panel of six TAAbs (p53, SOX2, Annexin-1, GBU4-5, CAGE and NY-ESO-1) was assessed in eight studies. A meta-analysis showed that the pooled sensitivity estimate was 38 (34 to 46) percent while the specificity estimate was 89 (83 to 91) percent. Mean diagnostic accuracy was 65.9 (62.5 to 81.8) percent.
Overall, the six-TAAb panel had a relatively low diagnostic accuracy (area under the curve [AUC], 0.52; 95 percent CI, 0.48 to 0.57). On the other hand, a panel of seven TAAbs (Hu-D, A4, MAGE, SOX2, GBU4-5, NY-ESO-1, CAGE, p5) yielded pooled sensitivity and specificity estimates of 47 (37 to 66) and 90 (84 to 91) percent, respectively. Mean diagnostic accuracy was 78.4 (67.5 to 88.8) percent.
The seven-TAAb panel had moderate diagnostic accuracy (AUC, 0.90; 0.87 to 0.93).