Authorship:
Laura Owen, Philip J Monaghan and Brian Keevil
MAHSC, University Hospital of South Manchester and The Christie Pathlogy Partnership. Manchester. UK.
| Short Abstract Oestradiol measurement is traditionally performed by immunoassays which have poor performance at lower concentrations. It is important to accurately measure oestradiol at low concentrations in certain patient groups e.g. females with breast cancer. Mass spectrometry is an attractive alternative due to its better specificity and performance at the low concentrations. We have investigated different mechanisms for improving the sensitivity of our current oestradiol assay. Further investigation into the interference of breast cancer medications has shown that one in particular, fulvestrant causes direct interference in two immunoassays. The other medications (anastrozole, exemestane and tamoxifen) do not appear to directly interfere. |
Long Abstract
Introduction
Oestradiol measurement is traditionally performed by immunoassays which have poor performance at lower concentrations. It is clinically important to be able to accurately measure oestradiol at low concentrations in certain patient groups such as females undergoing treatment for breast cancer. In this population the measurement of oestradiol is important for determining menopausal status as this may guide treatment choices. Immunoassay in this setting may be compromised due to lack of sensitivity. Additionally, some treatments for breast cancer have been reported to cause interferences in immunoassays.
Mass spectrometry is an attractive alternative due to its superior specificity and performance at the low concentrations that are required in this patient population group. However, this is still challenging from an analytical pespective due to the low picomolar circulating concentrations and because oestradiol preferentially forms negative ions.
Methods & results
We have investigated different mechanisms for improving the sensitivity of our current oestradiol assay. The aim was to continue to use a modest volume of sample (250 µL), as volumes greater than this are more difficult to handle and also not applicable to our routine clinical service, but still improve on the current sensitivity of 10 pmol/L. The presence of ammonium fluoride in the mobile phase improves the sensitivity by at least 30%. A novel ionisation technique, the Waters Unispray has also shown an improvement in sensitivity. This technique is a variation in ESI in which the eluate from the column is nebulised and sprayed at a charged target pin. The maintenance procedure of the instrument has also been optimised for this analyte as well as modifications to our sample extraction process.
Further investigation into the interference of breast cancer medications has shown that one in particular, fulvestrant causes direct interference in two immunoassays when spiked into serum. The degree of interference is different between the two immunoassays. Of the two immunoassays tested, the Siemens Centaur exhibits the most interference with a 70% increase of measured oestradiol with a 4 nmol/L spike which was equivalent to 2.7% cross-reactivity. The Abbott immunoassay had a 25% increase on measured oestradiol with the same sample which was equivalent to 1.3% cross-reactivity. The interference of fulvestrant in the immunoassays has also been confirmed in a clinical sample. In this case the mass spectrometry measurement was <10 pmol/L, the Abbott immunoassay gave a result of 117 pmol/L and the Centaur gave 371 pmol/L. Although fulvestrant does interfere directly with the immunoassay, the affect seen in this patient sample exceeds this, suggesting that metabolites of this drug may also have an effect. The other medications (anastrozole, exemestane and tamoxifen) do not appear to directly interfere when 4 nmol/L concentrations are spiked into serum as differences were within the measurement uncertainty of the analytical procedure. Previous reports in the literature of interference are therefore likely to be mediated by metabolites.
Conclusions
We have investigated different approaches to improve the sensitivity of an existing mass spectrometry assay for oestradiol and have succeeded in improving the sensitivity by approximately 50%. Due to the low concentrations and the potential for interferences, we recommend that all patient with breast cancer, having an oestradiol measurement performed, access a mass spectrometry assay for reliable results.