Decoding cancer’s molecular signature

The clinical diagnostics of cancer has seen significant changes in recent years. This is mainly due to three major and mutually influencing developments. First, the increasing knowledge of genetic alterations that accumulate during development of specific tumor types has led to a deeper understanding of tumor biology in general, and has allowed researchers and clinicians to assign individual tumor properties to individual patients. In consequence, the number of promising molecular biomarkers is constantly increasing both on the DNA/RNA and protein level.¹

In addition, the elaboration of new pathways for tumor genesis has generated a huge therapeutic potential by accelerating the development of new drugs designed to act on key events in these pathways. Second, the necessity to evaluate increasing numbers of tumor markers with diverse biochemical structures has given rise to enormous technological leaps, allowing for parallel evaluation of multiple parameters in a shorter time. These include optimization of the well-established ELISA- or qPCR-based techniques (e.g., homogeneous formats and detection systems), introduction of multiplexing concepts (e.g., microarrays), and development and adaptation of instruments (e.g., LC-MS/MS and Next Generation Sequencing). Third, the evaluation of a panel of potential tumor markers allows clinicians to generate a personalized tumor profile for each patient. As a consequence, molecular biomarkers help them to stratify the disease status and predict a patient's prognosis and the likelihood of an individual tumor's responsiveness.

The concept of so-called companion diagnostics paves the way for a personalized cancer medicine that aims to establish individualized treatment algorithms and to tightly monitor the efficacy of therapy as well as a recurrence of the disease. The effort to combine new biomarkers with complementary therapy has also led to the concept of theragnostics (a portmanteau of therapeutics and diagnostics). The key applications of theragnostics are identification of subgroups of patients presenting a profile likely to be responders (efficacy) or nonresponders (safety) to an intended treatment, and monitoring the response to this treatment. Theragnostics may therefore be considered as the combined result of new advances made in pharmacogenomics, drug discovery, and diagnostics.

The history of diagnosing prostate cancer may serve as a good example that reflects these developments. Prostate cancer is the second leading cause of cancer deaths in men in the United States. Based on rates from 2007 through 2009, the incidence rate of prostate cancer in the U.S. was 154.8 per 100,000 men per year, with a median age of diagnosis at 67 years and a median age of death at 80 years. As a lifetime risk, 16.15% of men born today will be diagnosed with cancer of the prostate at some time during their lifetime, but only approximately 3% of them will die from prostate cancer. The age-adjusted death rate is 23.6 per 100,000 men per year.²

Since the late 1980s, screening for prostate-specific antigen (PSA) in blood, along with digital rectal exam, was the FDA-approved method to detect prostate cancer in men age 50 and older. PSA is a protein produced by cells of the prostate gland. The higher a man's PSA level in serum, the more likely it is that cancer is present. However, the sensitivity and specificity of this biomarker was unsatisfying and resulted in a significant overtreatment rate. First, PSA levels can be elevated not only by prostate cancer but by other factors as well. Second, individual PSA levels vary widely in men, exacerbating the difficulty of defining a cut-off value. In fact, in 2011 the U.S. Preventive Service Task Force (USPSTF) advised against a general PSA screening, since two large studies in Europe (ERSCP) and the U.S. (PLCO), including more than 250,000 men, showed no or only minor benefits.

In recent years, several attempts have been made to improve the significance of the PSA test in particular and of prostate cancer screening in general. Besides PSA alone, the ratio between free and bound PSA and the velocity of PSA increase over time (doubling time) were introduced. New serum and urine molecular biomarkers have been identified, of which PCA3 (prostate cancer gene 3) has already been introduced clinically. In addition, the identification of prostate cancer specific genomic aberrations like the TMPRSS2:ERG gene fusion (transmembrane protease serine 2: Erythroblastosis virus E26 oncogene-related gene) might improve diagnosis and affect prostate cancer treatment.³ Furthermore, genome-wide association studies identified mutations or single nucleotide polymorphisms (SNPs) in the BRCA1 and BRCA2 gene (as for breast cancer) as independent risk factors that increase the risk for getting prostate cancer before age of 60 by a factor of five to 23 times.⁴ The combination of several novel biomarkers and PSA may now allow clinicians to triage men with borderline PSA values to further increase specificity and avoid overtreatment.

In summary, cancer diagnostics in blood will strongly profit from molecular signatures of the “omics” technologies. Combinations of novel and established biomarkers will not only be used for diagnosis, but will be an integral part of patient stratification, treatment, and surveillance.

References

1. FDA: Pharmacogenomic biomarkers in drug labels. http://www.fda.gov/Drugs/Science Research/ResearchAreas/Pharmacogenetics/ucm083378.htm. Accessed May 4, 2012.
2. Howlader N, et al. SEER Cancer Statistics Review, 1975-2009 (Vintage 2009 Populations), National Cancer Institute. Bethesda, MD.
3. Salagierski M, Schalken JA. J Urol. 2012;187(3):795-801.
4. Goh CL, et al. J Intern Med. 2012;271(4): 353-365.