Endocrine Abstracts

Background: The purpose of this study was to register the frequency and cause of method-dependent interference in the Roche thyroid immunoassays observed in our laboratory.

Methods: Serum samples to confirm discordant thyroid results were collected from June 2013 to September 2018, in samples sent to the Hormone Laboratory, Oslo University Hospital. The laboratory serves as a reference laboratory on thyroid diseases for about 2/3 of the inhabitants of Norway. We registered 106 samples with discordant pathological thyroid results when analyzed with the Roche methods, but with normal results when analyzed with alternative methods. 42 of the samples were sent to Roche Diagnostics for investigation of the interfering substance.

Results: In the 42 samples investigated we found interference by anti-streptavidin antibodies in 34 cases, antibodies against ruthenium or the idiotype of the ruthenium labelled antibody in three cases and human mouse antibodies (HAMA) in one case. We estimated the frequency of method-dependent interference to 0.2%. Euthyroid persons with streptavidin antibodies may have falsely elevated freeT4, freeT3, anti-thyroxin peroxidase and TSH receptor antibodies and decreased thyroid-stimulating hormone when analyzed with the Roche methods, mimicking a hyperthyroid disease.

Conclusion: Method-dependent interference in Roche thyroid immunoassays is caused mainly by streptavidin-antibodies, which occur more frequently than previously anticipated. This may lead to misdiagnosis and inappropriate medical treatment. Method-dependent errors should be reported to the supplier of the assay. If diagnostic tests showing method-dependent interference are not promptly improved, the laboratory should consider alternative methods.