Verification of ferritin on Roche cobas® 8000 C702 and comparison with E801

von | Okt. 17, 2025 | Original Papers

Ferritin is an iron-binding protein responsible for storing iron mainly in the spleen, liver and bone marrow. The plasma concentration of ferritin reflects the iron status in the patient and is determined for monitoring iron metabolism and iron therapy in patients [1].

The laboratory of the university hospital UZ Brussel (Brussels, Belgium) performs approximately 3,5 million cobas® 8000 (Roche Diagnostics, Rotkreuz, Switzerland) analyses on a yearly basis on two mirrored cobas® 8000 ISE-c702-e801 systems, with one of the most requested cobas® e801 analyses being ferritin (ca. 40,000 out of 850,000 per year). Since 2000, we used the electrochemiluminescence assay (ECLIA) on Roche cobas® 8000 e801 for the determination of ferritin in plasma. This method has been standardized against the Elecsys Ferritin Assay (REF11820982122). Due to the optimization of turn-around-time (reduction of analysis time from 20 to 10 min) and the limitation of available reagent positions on the e801 rotor, we have recently made the switch to the immunoturbidimetric assay on Roche cobas® 8000 c702. This method has been standardized against the Elecsys Ferritin assay (immunological assay) which is traceable to NIBSC (WHO).

Journal of Laboratory Medicine. Credits: De Gruyter
Journal of Laboratory Medicine. Credits: De Gruyter

Until now, no method comparison has been published between these two assays for the determination of ferritin. Therefore, the aim of this study is to examine whether there is a difference between the two methods in the low, normal and/or high range of plasma ferritin and whether the methods can be used interchangeably. We initially conducted a method verification of the Tina-quant Ferritin Gen. 4 on Roche cobas® 8000 c702 by determining the imprecision and accuracy (CLSI EP5-A2), by the verification of the auto-dilution 1:50 (EP34) and by the investigation of the high-dose hook effect [2], 3]. Secondly, we carried out a comprehensive method comparison of 115 samples between both methods using Passing–Bablok regression analysis and we determined the clinical sensitivity and specificity of the c702 method with the e801 method as a reference (EP09) [4]. We used frozen stored samples and ferritin data of 115 subjects that had been obtained between November 2023 and January 2024 at UZ Brussel. UZ Brussel Ethical Committee agreement was obtained for the study.

Full article