Applications of proteomics to clinical questions in breast cancer

The greatest hope for long, active lives for people with breast cancer rests in the creation of highly sensitive and specific tools to further optimize early diagnosis and treatment. Mammography is the international gold standard for breast cancer screening. It has positively affected our ability to detect small noninvasive breast cancer. However, this test has a high false-negative rate and a positive predictive value of only 25% [27]. The utility of mammography is even more questionable in young women, particularly those with dense breasts and with risk factors for the development of breast cancer. The inadequate sensitivity and specificity of mammograms provokes unnecessary anxiety and increases healthcare costs through the need for additional testing [12]. It is logical to expect that successful earlier detection should be possible, since it is estimated to take 6-8 years for a breast tumor to reach the 0.5 cm size threshold needed for detection by mammography [17]. The age of genomics has brought new promises through the discovery of germline aberration in genes that confer a high risk for breast cancer development, such as BRCA1, BRCA2, p53 (Li Fraumenni syndrome), and PTEN (Cowden's syndrome). This allows the direction of high-risk patients to increased surveillance, chemoprevention, lifestyle modifications, and prophylactic surgery. A large number of genetic changes such as mutations in p53, amplification of HER2/neu, and amplification of Rab25 have been demonstrated to correlate with patient prognosis [41]. Analysis of single genes, however, has insufficient power to alter patient management based on their effects on prognosis. In contrast, amplification of HER2/neu is predictive of response to herceptin, with a powerful negative predictive value, and may be predictive of responsiveness to anthracyclines. Gene transcription has been studied in depth using cDNA technology, and genomic analysis has advanced our understanding of the etiology and progression of breast cancer. Furthermore, transcriptional profiling, focusing on a limited number of genes, has refined our ability to classify breast cancers, at least in part based on the cell of origin (Luminal/basal) and potentially on genomic aberrations (HER2/neu amplification). While having great potential to impact on patient management, these technologies have not been sufficiently validated to justify implementation in clinical practice. Practical evaluation of these new insights has revealed that gene transcription does not necessarily reflect translation into protein or protein activity. The structure and function of trans lated proteins is studied through proteomics. The relationship between genomics and proteomics is analogous to a play: genomics reads the script while proteomics takes a snapshot of the script coming to life through the actors. Breast cancer, like all cancers, is a disease of aberrant function of protein products of abnormal genes. Dysregulation of genes, through mutation, rearrangement, loss, amplification, or silencing yields changes in proteins that alter the ability of the proteome to function normally. Cells are unable to wage war when confronted with neoplastic insult in the absence of a fully functioning protein infantry. Proteomics documents the functional consequences of genetic change by adding methods that complement existing technology. The proteomic approach is uniquely suited to answer clinical questions in breast cancer by moving the focal point from the microscope to the microenvironment. Intracellular signaling pathways and post-translational modifications can be exploited to discover molecular markers of breast cancer [35]. The products of these pathways in cancer cells can leave a molecular fingerprint that is shed into the blood supply. This led to our hypothesis that circulating blood may contain protein signatures reflecting organ-confined or small-volume disease [35]. Serum biomarkers, coupled with imaging approaches, may be developed to detect breast cancer during the 6- to 8-year window before it is apparent on a screening mammogram and has potentially acquired metastatic competence. Biomarkers may also be used to detect early relapse or monitor the response to therapy. Identification of the peptides involved in the biomarker signature may lead to better understanding of the pathophysiology of breast cancer. Finally, identification and study of upregulated or downregulated proteins in key signaling cascades will lead to novel therapeutic targets, molecular therapeutics, and molecular imaging.