Discrepant anti-dsDNA testing between EliA and Crithidia in ANA negative samples: just a false positive?

We read the insightful article of Trujillo Aguilera et al¹ concerning discrepancies in anti-dsDNA test results, a common issue in the routine of immunology laboratories, and we share the interest to approach the discrepant results and their clinical significance. This valuable longitudinal study was performed on discrepant dsDNA autoantibody (DA) samples in ANA positive by indirect immunofluorescence (IIF) patients. Interestingly, we have been working on a study on a complementary subset of patients, those IIF ANA negative.

It is usual to receive samples for ANA and dsDNA testing in immunology laboratory. But, as ANA negative by IIF result does not completely rule out the possibility of DA,² we performed the requested DA tests independently of IIF results. In the event that IIF were negative (or had any unrelated metaphase-negative pattern) but EliA DA were positive, we performed a subsequent Crithidia luciliae indirect immunofluorescence test (CLIFT) assay in order to verify DA by a method of higher specificity, as it has been usually recommended.³ A second discrepancy may then occur when DA is negative by CLIFT.

We were interested in knowing what happens in those patients IIF ANA negative, DA EliA positive but CLIFT negative. So, we studied a series of 173 patients who tested negative for ANA by IIF (HEp-2000 cells), positive for DA by EliA but CLIFT negative since 2016, and reviewed their previous clinical records and their evolution until now looking for those who got to develop a systemic autoimmune rheumatic disease (SARD).

From the total group of 173 patients, 15 developed a SARD while 158 were not. We found some significant differences between the groups having or not a final diagnostic of SARD. The SARD group had a lower mean age, a much higher female ratio and higher mean values of EliA DA (67.6 IU/mL vs 37.10 IU/mL in non-SARD group). Moreover, when stratifying the EliA DA results, the higher the DA levels, the more patients had a diagnosis of SARD. We found one patient with SARD in the lower positive range of EliA DA (up to 25 IU/mL), while 18.6% of patients were diagnosed with SARD when EliA DA values were over 25 IU/mL.

In consequence, we agree that CLIFT must be performed to confirm positive solid-phase DA assays if the clinical purpose is SARD diagnosis, but not for monitoring. We also fully agree that it is important to take into account possible false positives in discrepant samples, mainly in the low positive range of values. In addition, we think that the EliA dsDNA cut-off values should be upregulated without a significant loss of sensitivity in order to improve its clinical relevance. Further studies are needed in this way.

Finally, we emphasise the importance of maintaining a close collaboration between autoimmunity laboratory specialists and clinicians, to enhance mutual understanding of their respective needs and findings—specially in the case of discrepant results—ultimately benefiting the patient management.