[SELDI-TOF MS versus prostate specific antigen analysis of prospective plasma samples in a nested case-control study of prostate cancer.]
>>Int J Cancer. 2007 Apr 4; [Epub ahead of print]
SELDI-TOF MS versus prostate specific antigen analysis of prospective plasma samples in a nested case-control study of prostate cancer.
Skytt A, Thysell E, Stattin P, Stenman UH, Antti H, Wikstrom P.
Department of Medical Biosciences, Pathology, Umea University, Umea, Sweden.
There is an urgent need for better biomarkers for detection of clinically significant prostate cancer (PCa). Recent studies suggest that surface enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF MS) analysis of serum may provide diagnostic information. The aim of this study was to investigate if PCa cases could be identified by applying predefined SELDI-TOF analysis conditions on prospectively, uniformly collected plasma samples from PCa cases and matched controls. Prospective samples from 387 incident PCa cases and an equal number of controls, matched for age and time for recruitment, were analyzed by SELDI-TOF MS (IMAC30/Cu chip) and multivariate classification analysis. Prospective prostate specific antigen levels were subjected to ROC curve analysis giving an AUC of 0.87 for the total cohort with a median lag time between blood sampling and diagnosis of 6.1 years. No markers were found by SELDI-TOF MS that significantly discriminated between cases and controls in the total cohort or in subanalysis of cases with less than 2 years between blood donation and diagnosis (n = 42). Cases with aggressive disease at the time of diagnosis who gave blood less than 4 years prior to diagnosis (n = 23) could however be separated from their controls (sensitivity 70%, specificity 83%) by a model based on SELDI-TOF MS and OPLS-DA data analysis. We were thus not able to confirm previous results with SELDI-TOF MS identifying men with PCa from healthy individuals, but we report an optimal experimental set-up for verification of markers for early detection of cancer in prospectively collected samples.<<
This seems well controlled and done, but the parameters seem rather ambitious. That they got signal in men with aggressive disease is somewhat promising (the sensitivity, specificity, and -- by inference -- predictive value aren't looking too good here). But they couldn't confirm someone's studies. Guess we'd have to buy the article to know which ones they are referring to.
Cheers, Tuck |
