Search

Background Biological diagnosis of hemoglobin disorders is a complex process relying on the combination of several analytical techniques to identify Hb variants in a particular sample. Currently, hematology laboratories usually use high-performance liquid chromatography (HPLC), capillary electrophoresis and gel-based methods to characterize Hb variants. Co-elution and co-migration may represent major issues for precise identification of Hb variants, even for the most common ones such as Hb S and C. Methods We adapted a top-down selected reaction monitoring (SRM) electron transfer dissociation (ETD) mass spectrometry (MS) method to fit with a clinical laboratory environment. An automated analytical process with semi-automated data analysis compatible with a clinical practice was developed. A comparative study between a reference HPLC method and the MS assay was performed on 152 patient samples. Results The developed workflow allowed to identify with high specificity and selectivity the most common Hb variants (Hb S and Hb C). Concordance of the MS-based approach with HPLC was 71/71 (100%) for Hb S and 11/11 (100%) for Hb C. Conclusions This top-down SRM ETD method can be used in a clinical environment to detect Hb S and Hb C.

Background The use of targeted LC-MS/MS methods for protein quantitation in clinical laboratories implies a careful evaluation of potential sources of analytical interference. In this study, we investigated whether inflammation, which is associated with both the release of proteolytic enzymes and increased expression of acute phase protease inhibitors, is affecting the accuracy of a haptoglobin selected reaction monitoring (SRM) assay. Results A SRM assay was developed and used to quantify haptoglobin in 57 human serum samples. The SRM assay had CVs (n = 6) of 12.9% at 698 mg/L and 11.8% at 1690 mg/L. Results of the SRM assay were compared to those of a commercial immunonephelometric test. Passing-Bablok regression gave a proportional bias of 0.92 (95% CI: 0.82 to 1.04) and a constant bias of 75.40 (95% CI: −71.09 to 251.04), indicating that SRM and immunonephelometric assays provided comparable results. We then investigated whether the accuracy of the SRM assay was influenced by the patient’s inflammatory state by assessing the relationship between the serum CRP concentration and the bias between the two methods. No correlation was found between the SRM/immunoassay bias and the CRP concentration (Pearson correlation coefficient r = 0.0898). Conclusions These data indicate that neither the release of proteolytic enzymes nor the increased level of protease inhibitors occurring during inflammation processes have a significant impact on the haptoglobin SRM assay accuracy. Such studies provide important information about potential sources of analytical interferences in protein SRM assays. Electronic supplementary material The online version of this article (doi:10.1186/1559-0275-11-38) contains supplementary material, which is available to authorized users.