GSBS Student Publications


Monitoring the serological proteome: the latest modality in prostate cancer detection

Publication Date


UMMS Affiliation

Graduate School of Biomedical Sciences; Division of Urology; Department of Medicine, Division of Oncology

Document Type



Life Sciences | Medicine and Health Sciences


PURPOSE: Various strategies have recently emerged to improve the diagnostic prediction of prostate cancer (CaP). One such strategy includes the mass profiling of serum protein fractions selectively adsorbed onto chemically modified probes. In the current study we further validated this approach, while offering a more versatile, less expensive and yet equally predictive alternative to existing technologies.

MATERIALS AND METHODS: A solid core lipophilic C-18 resin was used to extract and enrich the low molecular weight protein fraction from patient serum for further analysis by mass spectrometry. Mass spectra generated from a 48 patient training set were data mined using multivariate analysis to identify diagnostically significant protein peaks. These peaks were then used to test a blinded study set comprising 168 patients with common statistical algorithms and commercially available software packages.

RESULTS: A total of 36 peaks generated from the training set were used to test the combined set of 168 serum samples obtained from 98 healthy individuals and 70 patients with CaP. We report a sensitivity of 94.1% and a specificity of 99.0% with 1 false-positive, 4 false-negative and 5 nondiagnosed cases.

CONCLUSIONS: Our results further indicate that mass profiling of serological proteins provides a means for the accurate detection of CaP. In addition, our approach was found to be superior to chip based protocols, generating rich, sharp, highly reproducible spectra attainable in a high throughput manner and at minimal cost. This technique is also scaleable for subsequent protein characterization using multidimensional protein identification technologies. Finally, analyses of mass spectra with commercially available statistical applications was found to be highly effective in generating highly discriminatory m/z values for CaP diagnosis.

DOI of Published Version



J Urol. 2004 Jul;172(1):331-7. Link to article on publisher's site

Journal/Book/Conference Title

The Journal of urology

Related Resources

Link to article in PubMed

PubMed ID