Elusion or inspired scientific reality? Personalized medicine could revolutionize cancer management, if robust prognostic and predictive tools can be developed. However, there is debate whether it is now the right time to focus primarily on personalized medicine Citation[101] or/if comparative-effectiveness research (CER) Citation[1] should continue to attract major funding and interest by public health and private sector bodies and society?
The American Reinvestment and Recovery Act will give CER a large boost in funding over the next 2 years. Despite a consensus that better information about the relative effectiveness of different medical interventions is needed to improve the quality and value of care, some view CER with scepticism. Recently, the Partnership to Improve Patient Care Citation[2], a coalition of 36 industry, patient-advocacy and clinician organizations, raised concerns that CER will not take adequate account of individual patient differences, and may impede the development and adoption of improvements in medical care and ‘stymie progress in personalized medicine’ Citation[2,101].
Yet, effective primary prevention of a common disease, such as cardiovascular disease and cancer, or cancer recurrence-risk prediction and prevention is a major challenge. Despite research efforts, little progress has been made over the last few decades. Although the development of markers to predict a person’s disease risk or predictive markers of response to a therapy has been a first priority for the public health and private sectors, the ultimate translation of new genetic or molecular tools into wide clinical practice is modest. Most research articles provide very promising results but the proportion achieving translation from bench to clinical testing in large-scale randomized trials is small. Finally, the rate of incorporation of basic and translational research findings into a day-to-day clinical practice is even smaller. Therefore, valid concerns call for strict methodological criteria and valid reporting of biomarkers research. It is now recommended by the editors of high-impact journals, such as Nature, that scientists straddling the boundary between bench and bedside must conduct and report their research with the rigor that each individual community expects Citation[3].
Biomarkers in both prevention and treatment settings of common disease are not a new concept. It is well known that cholesterol or prostate-specific antigen (PSA) levels indicate a risk for cardiovascular disease and prostate cancer, respectively. However, both are complex multifactorial diseases, so that the clincal use of such measurements has limitations for accurate individual risks predictions and general population screening Citation[4]. In cancer treatment, the tumor–node–metastasis (TNM) staging system and clinicopathologic features have been the standard tools currently used for surgery, radiotherapy and systemic therapeutic decisions Citation[5], but its modest ability to discriminate between responder and nonresponder patients has been widely recognized.
Single-protein-coding genes research has been the standard approach for discovering biomarkers. Substantial progress has been made over the past decade. In primary prevention, the discovery of heritable mutations in several single genes allows life-saving interventions. In the treatment setting, the molecular status of some tumors may guide the new era of specific targeted therapy in specific subsets of patients. However, this is isolated success for a small proportion of patients and a few cancer types. Advances, limitations and perspectives of translational single-gene research can be summarized.
& Box 1 summarizes the achievements and weaknesses of single-gene-based development of genetic tools for cancer primary prevention. However, personalized preventive intervention at the right time in high-risk people within the general population is elusive. Genetic testing allows identification of heritable breast cancer Citation[6,7], colorectal cancer Citation[8] and gastric cancer Citation[9]. However, these mutated genes are very rare in the general population – less than 0.3%. Inherited syndromes account for 3–10% of these cancer types, the predictive accuracy at 50 years of age is less than 50% and, ultimately, 25% of mutation carriers in BRCA1/2, mismatch-repair genes or CDH1 genes will never develop breast cancer, colorectal cancer or gastric cancer, respectively. This uncertainty challenges a tailored surgical or medical preventive intervention in these high-penetrating mutations carriers. Yet, no robust tool has been developed to identify the vast majority of patients (90–97%) before they are diagnosed with these cancers.
& Box 2 delineates the landscape of biomarkers to select patients for treatment with biologic agents. The advent of targeted therapy has revolutionized systemic treatment of cancer patients. But now, a decade later, we know the limitations of this targeted treatment without molecular marker development. It was only recently, approximately 50 years after the discovery of the EGF and its receptor (EGFR), that the concept of biologic agents has been incorporated into pharmaceutical industry. Targeting cancer-only cells but not health cells, by biologics could dramatically improve poor outcomes of cancer patients with a very low adverse effects profile. However, as data from large-scale randomized controlled trials have become available, valid scientific scepticism has replaced initial overenthusiasm.
The list of biologic agents approved by the US FDA for many common solid cancers is rapidly growing. But questions and uncertainty for a day-to-day practice are simultaneously growing. Most biologics have been approved for the metastatic setting but a few have only been approved for clinical use in the adjuvant setting. In stage IV disease, FDA approval is primarily based on progression-free survival (PFS). However, without an overall survival (OS) benefit, such as trastuzumab for metastatic breast cancer and gastric cancer Citation[10], a true efficacy of these biologics is questionable. For example, inhibiting EGFR signaling pathway with the monoclonal antibodies cetuximab or panitumab only have therapeutic effects for KRAS wild-type metastatic colorectal cancer Citation[11]. In the absence of molecular markers to select patients for anti-angiogenic therapy, such as bevacizumab, hope for effective treatment with anti-angiogenic agents without biomarker-based selection is minimal Citation[12,13].
However, isolated clinical success in reducing mortality with adjuvant treatment in the solid-cancers adjuvant setting with potential life-saving effects have been reported only for the anti-HER2 antibody trastuzumab for patients selected on the basis of HER2-positive early breast cancer Citation[14]. However, even in this case, the net response rate of this agent added to chemotherapy is 10%, although the overall benefit of chemotherapy plus trastuzumab is 40% Citation[15]. Given the concerns that this generation of targeted agents either provides no therapeutic benefit or, in a few cases, may simply delay the occurrence of recurrence Citation[16], longer follow-up results are anticipated to enable robust conclusions about whether trastuzumab or other HER2 signaling pathway inhibitors can have long-term effects, resulting in a true cure.
Micrometastases or isolated tumor cells (ITCs) in sentinel and nonsetinel lymph nodes have been included as prognostic factors for breast cancer in the TNM staging systems by the American Joint Committee on Cancer (AJCC) Citation[5]. A large-scale study from The Netherlands concluded that micrometastases or ITCs could be used as tools to guide adjuvant chemotherapy decisions in early, node-negative breast cancer Citation[17]. However, a series of weaknesses arising from the retrospective nature of this study limit its power, which suggests the need for results of a randomized controlled trial for definitive conclusions Citation[18].
Novel biomarkers are urgently needed for personalized medicine progression, but which strategy is more appropriate to reach faster and more efficient primary prevention and treatment? Is it the current standard single-gene-based molecular research or the more exotic but exciting perspectives molecular-network-based option?
The results of both large-scale clinical trials testing single-gene-based development of biomarkers and cancer genetics and genomics studies confirm the highly complex and heterogeneity of the nature of cancer and reveal that a much more sophisticated approach is required to develop novel robust biomarkers.
Indeed, extensive genetic studies Citation[19,20] and next-generation DNA-sequencing technology-based whole cancer genome sequencing studies Citation[21] for lung cancer, breast cancer and melanoma Citation[22–25] confirm the complexity of molecular mechanisms underlying cancer development.
The advent of high-throughput screening over the past decade has revolutionized basic and translational research Citation[26]. Using massively parallel-sequencing technology, many gene-expression profiling studies have resulted in the development of gene signatures. The 70-gene signature (MammaPrint) and the 21-gene assay have already commercialized adjuvant chemotherapy decisions in early, node-negative breast cancer Citation[27]. However, given the heterogeneity and complexity of cancer, and some methodological weaknesses in the development of these multigene assays, no robust conclusions can be drawn before the results of two ongoing, large-scale, randomized controlled trials testing the efficacy of these two multigene assays in the USA and Europe become available Citation[28].
In the postgenomic era, next-generation DNA-sequencing technology-based results reveal, in an unpreceded level, that cancer initiation, growth and metastasis are driven by molecular networks rather than one mutated gene or single deregulated signaling pathway. Molecular system approaches may allow efforts towards understanding how intracellular signlling pathways networks operates Citation[29] and how oncological outcomes are governed by interactions among cancer cells, with different response to therapeutics within an individual primary tumor and its associated metastases. This comprehensive understanding of how a solid tumor functions as a whole biological system and its relationships with multiple host variables, including heritable causal mutations and environmental exposures and lifestyle, reveal the fundamental importance of complex biological-systems approaches Citation[30–32]. The repertoire of somatic and heritable of point mutations, rearrangements and copy number changes is extremely large Citation[19–21] and, yet, the reliability of functional cancer genomes data collected is uncertain. Therefore, sophisticated network-based approaches to identify crucial cancer genome targets are of paramount importance towards the development of novel biomarkers. The way towards personalized medicine over the next 10 years is approaching Citation[33] but may pass through the development of robust complex predictive networks models Citation[30–32].
Instead of the debate regarding personalized medicine and CER, efforts should be focused on how to harmonize both research strategies. History has demonstrated that, for effective and safe evidence-based day-to-day clinical practice, the principles and rules of CER should be considered at a very early preclinical development stage in designing molecular system-based new-generation biomarkers. Particular emphasis should be given to the integration of clinical data, apart from genetics and genomics data. Novel, network-based targets should prove their potential superiority over the current standard cancer diagnostics and prognostic tools Citation[34].
Table 1. Primary prevention of some common solid cancers.
Table 2. Current and next-generation biomarkers for targeted therapy of solid cancers.
Box 1. Molecular network-based development of biomarkers for general population screening and future clinical goals.
Multiple gene–gene and gene–environment interactions
Given the complexity of multiple interactions responsible for the vast majority of remaining cancers cases, these systems approaches are needed for network-based predictive models to identify robust biomarkers
Box 2. Molecular network-based development of next-generation biomarkers for selecting biologic agents.
• Intracellular signaling pathways network
• Different cancer cell subpopulations within an individual tumor with various treatment response
• Cancer cells interactions and interactions between them and microenvironment.
Given this complexity and heterogeneity systems approaches are needed for network-based predictive models to identify cancer targets allowing robust biomarkers development.
Financial & competing interests disclosure
The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.
References
- Mushlin AI, Ghomrawi H. Health care reform and the need for comparative-effectiveness research. N. Engl. J. Med.362(3), e6 (2010).
- Garber AM, Tunis SR. Does comparative-effectiveness research threaten personalized medicine? N. Engl. J. Med.360(19), 1925–1927 (2009).
- Valid concerns. Nature463(7280), 401–402 (2010).
- Lin K, Lipsitz R, Miller T, Janakiraman S; US Preventive Services Task Force. Benefits and harms of prostate-specific antigen screening for prostate cancer: an evidence update for the U.S. Preventive Services Task Force. Ann. Intern. Med.149(3), 192–199 (2008).
- AJCC Cancer Staging Manual (7th Edition). Edge SE, Byrd DR, Carducci MA, Compton CA (Eds). Springer, NY, USA (2009).
- Miki Y, Swensen J, Shattuck-Eidens D et al. A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science266, 66–71 (1994).
- Wooster R Bignell G, Lancaster J et al. Identification of the breast cancer susceptibility gene BRCA2. Nature378, 789–792 (1995).
- Lynch HT, Lynch PM, Lanspa SJ, Snyder CL, Lynch JF, Boland CR. Review of the Lynch syndrome: history, molecular genetics, screening, differential diagnosis, and medicolegal ramifications. Clin. Genet.76(1), 1–18 (2009).
- Guilford P, Hopkins J, Harraway J et al. E-cadherin germline mutations in familial gastric cancer. Nature392(6674), 402–405 (1998).
- Roukos DH. Targeting gastric cancer with trastuzumab: new clinical practice and innovative developments to overcome resistance. Ann. Surg. Oncol.17, 14–17 (2010).
- Van Cutsem E, Köhne CH, Hitre E et al. Cetuximab and chemotherapy as initial treatment for metastatic colorectal cancer. N. Engl. J. Med.360(14), 1408–1417 (2009).
- Ellis LM, Reardon DA. Cancer: the nuances of therapy. Nature458(7236), 290–292 (2009).
- Roukos DH. Tzakos A, Zografos G. Current concerns and challenges regarding tailored anti-angiogenic therapy in cancer. Expert Rev. Anticancer Ther.9(10), 1413–1416 (2009).
- Hortobagyi GN. Trastuzumab in the treatment of breast cancer. N. Engl. J. Med.353(16), 1734–1736 (2005).
- Roukos DH, Ziogas D. From tumor size and HER2 status to systems oncology for very early breast cancer treatment. Expert Rev. Anticancer Ther.10(2), 123–128 (2010).
- Roukos DH. Mea kulpa with cancer-targeted therapy: new thinking and new agents design for novel, causal networks-based, personalized biomedicine. Expert Rev. Mol. Diagn.9(3), 217–221 (2009).
- de Boer M, van Deurzen CH, van Dijck JA et al. Micrometastases or isolated tumor cells and the outcome of breast cancer. N. Engl. J. Med.361(7), 653–63 (2009).
- Roukos DH. Isolated tumor cells in breast cancer. N. Engl. J. Med.361, 1994–1995 (2009).
- Wood LD, Parsons DW, Jones S. The genomic landscapes of human breast and colorectal cancers. Science318(5853), 1108–1113 (2007).
- Jones S, Zhang X, Parsons DW. Core signaling pathways in human pancreatic cancers revealed by global genomic analyses. Science321(5897), 1801–1806 (2006).
- Stratton MR, Campbell PJ, Futreal PA. The cancer genome. Nature458(7239), 719–724 (2009).
- Pleasance ED, Stephens PJ, O’Meara S et al. A small-cell lung cancer genome with complex signatures of tobacco exposure. Nature463, 184–190 (2010).
- Pleasance ED, Cheetham RK, Stephens PJ et al. A comprehensive catalogue of somatic mutations from a human cancer genome. Nature463, 191–196 (2010).
- Shah SP, Morin RD, Khattra J et al. Mutational evolution in a lobular breast tumour profiled at single nucleotide resolution. Nature461(7265), 809–813 (2009).
- Stephens PJ, McBride DJ, Lin ML et al. Complex landscapes of somatic rearrangement in human breast cancer genomes. Nature462(7276), 1005–1010 (2009).
- Shendure J, Ji H. Next-generation DNA sequencing. Nat. Biotechnol.26, 1135–1145 (2008).
- Sotiriou C, Pusztai L. Gene-expression signatures in breast cancer. N. Engl. J. Med.360(8), 790–800 (2009).
- Roukos DH. Twenty-one-gene assay: challenges and promises in translating personal genomics and whole-genome scans into personalized treatment of breast cancer. J. Clin. Oncol.27(8), 1337–1338 (2009).
- Hahn WC, Weinberg RA. Modelling the molecular circuitry of cancer. Nat. Rev. Cancer2(5), 331–341 (2002).
- Schadt EE. Molecular networks as sensors and drivers of common human diseases. Nature461(7261), 218–223 (2009).
- Roukos DH. Novel clinico-genome network modeling for revolutionizing genotype–phenotype-based personalized cancer care. Expert Rev. Mol. Diagn.10(1), 33–48 (2010).
- Deisboeck TS. Personalizing medicine: a systems biology perspective. Mol. Syst. Biol.5, 249 (2009).
- Goldstein DB. 2020 visions. Personalized medicine. Nature463, 27 (2010).
- Roukos DH. Systems medicine: a real approach for future personalized oncology? Pharmacogenomics11(3), 283–287 (2010).
Website
- Partnership to Improve Patient Care. Individualized patients, personalized care. Washington, DC, USA, PIPC, 2009 www.improvepatientcare.org/individualized-patients-personalized-care (Accessed 17 April 2009)