In the past 5 years, significant progress has been made in identifying novel biomarkers for ovarian cancer early detection. A field of recent interest is clinical proteomics, which has been reported to lead to high sensitivity and specificity for early detection of cancer (Petricoin and Liotta, 2002; Wulfkuhle, 2003; Zhu et al., 2003). This emerging field uses mass spectrometry-based protein profiles/patterns of easy accessible body fluids to distinguish cancer from non-cancer patients. It is postulated that the blood proteome constantly changes as a consequence of the perfusion of the diseased organ adding, subtracting, or modifying the circulating proteome. These differences might be the result of proteins being abnormally produced or shed and added to the serum proteome, proteins being clipped or modified as a consequence of the disease process, or proteins being subtracted from the proteome owing to disease-related proteolytic degradation pathways. Therefore, protein pattern diagnostics would provide easier and more reliable tools for detection of cancer (Petricoin and Liotta, 2002; Wulfkuhle et al., 2003; de Noo et al., 2006).

However, after the initial positive announcements in early 2002 on ovarian cancer early detection, questions were raised about whether the results were reproducible and reliable enough for application in practice (Diamandis, 2004; Check, 2004; Garber, 2004). There is a sound basis for optimism that novel and robust approaches to cancer detection and screening will emerge in the near future, but further progress in refining the reproducibility and sensitivity of the technology will be required and the question about whether the approach of discovery-based serum proteomics can accurately and reliably diagnose ovarian cancer – or any cancer – has not been resolved (Ransohoff, 2005; Jacobs and Menon, 2004). The challenges in translating the potential of serum proteomic technology into a robust approach to clinical ovarian cancer screening are numerous and criticisms were raised about several aspects of serum proteomics. Among them, non-disease-related confounding factors are important sources of bias and variations, which may lead to misleading diagnosis. In this work, we focus on the effects of confounding factors and how they can be addressed in proteomic studies.

Confounding factor affects sensitivity and specificity of identified biomarkers because any confounding factor could conceivably cause a phenotypic response that might be confused with a specific characteristic of the disease process under study (Boguski and McIntosh, 2003). Confounding factors can be classified as biological confounding factors and technological confounding factors. Biological confounding factors include clinical or biological factors such as age, race, and diet. Technological confounding factors include sample collection, sample quality, analysis procedures etc. This is important to notice that technological confounding factors can be eliminated or reduced by advances in technology and standardization of procedures. However, biological confounding factors would always exist. Therefore, any effective proteomic methods should be designed to eliminate or reduce the effects of biological confounding factors.

In this work, some existing statistical methods are investigated regarding their capabilities of confounding factor handling. New approaches are proposed to address this issue more effectively. The proposed methods are evaluated using both simulations and publicly available cancer proteomic datasets from national cancer institute (NCI).

References:

1. Petricoin EF, Liotta LA. Proteomic analysis at the bedside: early detection of cancer. Trends Biotechnol 2002; 20: S30-S34

2. Wulfkuhle JD, Liotta LA, Petricoin EF. Proteomic applications for the early detection of cancer. Nat Rev Cancer 2003; 3: 267-275

3. Ransohoff D.F. (2005), Lessons from controversy: ovarian cancer screening and serum proteomics. J Natl Cancer Inst, Vol. 97 (4), 315 – 319.

4. Boguski M.S., McIntosh M.W. (2003), Biomedical informatics for proteomics. Nature, Vol. 422: 233 – 237

5. Wolkenhauer O., Ghosh B.K., and Cho K.H. (2004), Control and coordination in biochemical networks. IEEE Control Systems Magazine, 24(4):30-34.

6. Check E. (2004 ), Running before we can walk? Nature; 429 : 496 – 7.

7. Diamandis E.P. (2004), Analysis of serum proteomic patterns for early cancer diagnosis: drawing attention to potential problems. J Natl Cancer Inst, 96(5):353 – 356.

8. Garber K. (2004 ), Debate rages over proteomic patterns. J Natl Cancer Inst; 96 : 816 – 8.

9. Jacobs I.J., Menon U. (2004), Progress and challenges in screening for early detection of ovarian cancer, Molecular & Cellular Proteomics, Vol. 3.4, 355 – 366.

10. Zhu W., Wang X., Rao M., Glimm J., Kovach J.S. (2003), Detection of cancer-specific markers amid massive mass spectral data, Proceedings of National Academy of Science, Vol. 100 (25), 14666 – 71