Letter to the Editor
False-positive TSH receptor antibody results in an infant with activating TSHR mutation
2018 Volume 65 Issue 7 Pages 793-794
Details
2018 Volume 65 Issue 7 Pages 793-794
Dear Editor,
Wada et al. [1] reported the occurrence of false-positive TSH receptor antibody results in neonates and healthy volunteers using lithium-heparin plasma in a third-generation anti-TSHR platform (Elecsys® Anti-TSHR, Roche Diagnostics Ltd., Basel, Switzerland). We encountered a case of discrepant TSH receptor antibody results in an infant with activating TSH receptor mutation. The Chinese male infant was delivered preterm (33 week 5 days) for suspected fetal distress. He had persistent tachycardia after birth and thyroid function tests on day 10 showed a suppressed TSH (<0.01 mIU/L) and a grossly elevated free T4 (>100 pmol/L). The mother was clinically and biochemically euthyroid with no personal history of thyroid disease. Family history of thyroid diseases was negative. The serum TSH receptor antibody of the mother on postnatal day 10 was slightly elevated (1.3 IU/L, reference interval <1 IU/L) on a third generation enzyme-linked immunosorbent assay (ELISA) (Anti-TSH receptor (TRAb) Fast ELISA (IgG), Euroimmun, Lübeck, Germany). The serum TSH receptor antibody results of the infant were negative for three times initially but became persistently elevated and fluctuating afterwards (Table 1). The infant’s thyrotoxicosis was difficult to control and eventually required block-and-replace regimen. A diagnosis of neonatal Graves’ disease was considered to be less likely in view of the prolonged and persistent thyrotoxicosis, the initially-negative TSH receptor antibody levels in the infant and the mild elevation of the TSH receptor antibody result of the mother. Sanger sequencing of the TSHR gene was performed and heterozygous TSHR c.1358T>C, pMet453Thr, which is a known pathogenic variant associated with Persistent Sporadic Nonautoimmmune Hyperthyroidism (PSNAH), was detected [2]. We re-analyzed all sera from the infant and the mother on another third generation electrochemiluminescence immunoassay (ECLIA) (Elecsys® Anti-TSHR, Roche Diagnostics Ltd., Basel, Switzerland). All results were negative (<0.8 IU/L, Table 1). Patients with PSNAH are typically negative for TSH receptor antibody [3], although positive results have been reported [4]. Thus, we consider that the Euroimmun assay results were likely false positives. Third generation TSH receptor antibody assays are characterized by the use of a monoclonal human TSH receptor-stimulating antibody, M22. It is known that inter-method variability in TSH-Receptor antibody measurement exists for different third generation assays, despite calibration against the same reference standard (NIBSC 90/672), leading to discrepant results around cut-off levels [5, 6]. We performed a comparison of results between the two assays using specimens from pregnant women with present or past history of hyperthyroidism (n = 23). The Passing-Bablok regression equation was [Euroimmun Result] = 1.57 * [Roche Result] – 0.35 and the Spearman rank correlation coefficient was 0.93 (95% CI: 0.85 to 0.97). Thus, although the Euroimmun assay results were generally higher than the Roche assay results, the grossly elevated Euroimmun assay results in the present case were clearly unexpected. To our knowledge, this is the first case report of grossly discrepant TSH receptor antibody results associated with a delay in reaching the correct diagnosis as the false positive TSH receptor antibody results suggested a diagnosis of neonatal Graves’ disease initially. The use of one or more alternative TSH receptor antibody assay(s) should always be considered when the results of one assay are not compatible with the clinical presentation. The cause of the false-positive results and fluctuation of Euroimmun TRAb results in this case remained unexplained. Further research on the frequency of occurrence of false-positive TSH receptor antibody results in neonates are warranted.
| Age | Euroimmun Anti-TSHR Ab (IU/L) | Roche Anti-TSHR Ab (IU/L) |
|---|---|---|
| 11 days | 0.6 | <0.8 |
| 19 days | 0.3 | <0.8 |
| 1.2 months | 0.1 | <0.8 |
| 1.9 months | 8.4 | <0.8 |
| 2.2 months | 7.1 | <0.8 |
| 2.4 months | 5.9 | <0.8 |
| 2.6 months | 5.4 | <0.8 |
| 2.9 months | 14.9 | <0.8 |
| 3.1 months | 34.3 | <0.8 |
| 3.3 months | 31.9 | <0.8 |
| 3.6 months | 20.6 | <0.8 |
| 3.8 months | 10.0 | <0.8 |
| 4.0 months | 10.7 | <0.8 |
| 4.3 months | 10.9 | <0.8 |
| 4.5 months | 4.8 | <0.8 |
| 4.7 months | 6.3 | <0.8 |
We would like to acknowledge Dr Liz YP Yuen for her comments on the manuscript, and Dr CC Ho and Dr WT Poon for the genotyping result.
