Apr 22, 2013
By Al B. Benson III, MD, FACP, FASCO
Feinberg School of Medicine, Robert H. Lurie Comprehensive Cancer Center, Northwestern University
Personalized medicine embodies the concept of delivering the right care at the right time. An individualized approach to provide value-based, high-quality cancer care incorporates team medicine requiring patient access to comprehensive services offered by nurses, multispecialty oncology clinicians, psychologists, social workers, nutritionists, financial counselors, genetic counselors, and pathologists, as important examples. Many now prefer the term “precision medicine,” with the additional emphasis on choosing the tests that are most likely to benefit patients, whether for diagnostic, toxicity assessment, efficacy, or surveillance purposes.
Advances in developmental therapeutics have introduced an evolving portfolio of targeted agents across a spectrum of diseases, which, by definition, challenges the clinician to precisely select those individuals most likely to benefit from a molecularly driven strategy. The quest to identify particular markers to predict which patients are most likely to obtain disease control from a given targeted therapy has been elusive for many of the targeted therapies now available. Tumor heterogeneity and the complexity of molecular pathways exemplify the many obstacles that have hindered the development of precise predictive markers linked to treatment efficacy. To further address these challenges, it will be increasingly important to include pathologists and other laboratory scientists, as well as imaging biomarker experts, as critical members of the medical team with routine access to patients’ tissues, particularly for inclusion in tumor banks for future research purposes and for correlative studies in developmental therapeutic clinical trials.
In this issue of ASCO Connection’s Current Controversies in Oncology series, Dr. Lillian Siu discusses the acceptance of research biopsies by patients, the importance of obtaining these biopsies, and the limitations of developing less-invasive techniques to develop biomarkers. Dr. Luis Diaz offers insights as to the drawbacks of obtaining human biopsies and the limitations imposed on the analyses of these tissues, including sample characteristics and tumor heterogeneity, as well as novel approaches that may preclude the need for actual tissue biopsies to obtain individual biologic information. As science evolves, the goal is to fuse the bench and the bedside to more precisely offer predictably effective therapeutics for the individual patient over time by members of a truly comprehensive medical team.
Dr. Benson is a Professor of Medicine at Feinberg School of Medicine and Associate Director for Clinical Investigations at the Robert H. Lurie Comprehensive Cancer Center. He serves on the Editorial Boards of ASCO Connection and ASCO University®.
Tissue Acquisition for Personalized Cancer Therapy: Current Standard and Future Outlook
By Lillian L. Siu, MD
Princess Margaret Cancer Centre, University of Toronto, Canada
Personalized cancer medicine: “No Tissue, No Marker, No Trial”
We have entered the personalized cancer medicine era whereby we are increasingly utilizing information gathered from molecular characteristics of tumor tissue to complement histopathology, in order to enhance diagnostic, prognostic, or predictive evaluations.1 The quote, “no tissue, no marker, no trial,” by Dr. James Doroshow, Director of the Division of Cancer Treatment and Diagnosis of the National Cancer Institute, bears a clear message.2 The acquisition of tumor tissue to perform validated biomarker assays that help distinguish responders from nonresponders, or provide proof of mechanism, has become a fundamental component of high-impact clinical trials using molecularly targeted agents.
Acceptance of research biopsies by patients
Tumor biopsies, typically using a needle with or without image guidance, enable a direct assessment of biomarkers at the tissue and individual tumor cell levels to confirm the presence or absence of a genetic aberration, to assess target engagement or inhibition, or to identify mechanisms of resistance. As with all invasive procedures, tumor biopsies carry risks, but the frequencies are relatively low, ranging from 5% to 7% overall and less than 1% for major complications in the hands of experienced drug development programs.3,4
The ethics of mandatory research tumor biopsies for correlative studies in clinical trials have been the subject of many deliberations,5,6 especially when the results do not personally affect the clinical care of the trial participants. Most groups have emphasized the importance of providing trial participants a clear understanding of the rationale, risks and benefits of such procedures, so patients can make an informed and voluntary decision.5,6 We have previously surveyed 325 clinic patients who had prior diagnostic but not research biopsies; 48% of the respondents would not be deterred from enrolling into a clinical trial that mandates tumor biopsies for research purposes only.7
Post-diagnostic tumor re-biopsies
The scenario in which tumor biopsies are performed to procure tissue to measure a biomarker that can be used to guide clinical management is distinct from that in which biopsies are performed solely for exploratory research. An obvious example for the first scenario is tissue procurement to detect KRAS mutation in advanced colorectal cancer to avoid ineffective therapy with anti–epidermal growth factor receptor monoclonal antibodies. A less clear-cut example would involve tumor biopsy to select patients whose tumors harbor an unvalidated predictive biomarker of promise, based on preclinical data, for matching to an investigational agent. Although such tumor biopsies seem to fall somewhere between research and diagnostic procedures in their risk-benefit ratio, physicians should exert equal rigor in obtaining informed consent from their patients.
Current status of tumor biopsies compared with other alternatives
There are emerging interests to develop less-invasive techniques, such as circulating free nucleic acids (cfNA), to replace the need for tumor biopsies for molecular profiling.8 Until the samples offered by cfNA are validated to be as sensitive as tissue samples and representative of the global status of the cancer genomic landscape in patients, there are several key reasons why tumor biopsies should remain a gold standard at the present time:
- Tumor biopsies provide tissue for histopathologic confirmation of the diagnosis, to evaluate basic morphologic features such as tumor grade, and to perform relevant immunostains for prognostic and predictive purposes (e.g., estrogen receptor in breast cancer).
- Examination of tumor tissue enables the evaluation of heterogeneity in gene copy or gene/protein expression at the cellular level, especially in the context of histologic heterogeneity, which is not possible by cfNA.
- cfNA are not always detectable, and even when detectable, the concordance between genomic aberrations present in tumor specimens and in blood-based samples must be demonstrated for the latter to be an acceptable substitute. In a recent study of 84 patients with advanced solid tumors, a comparison of somatic mutations detected using the Sequenom OncoCarta genotyping panel (v1.0) between matched plasma and formalin-fixed paraffin-embedded archival tumor tissue of primary and/or metastatic sites revealed an overall concordance of only 60% (25 of 42 detected mutations).9
- Assays that evaluate cfNA are limited by their sensitivity of detection. Ease of detection is related to the frequency of the somatic alterations in the analyzed regions of cancer-related genes. Although sensitivity can be increased by novel techniques that enable amplification and deep sequencing of selected genomic regions, these methods need further optimization before transitioning from research to diagnostic laboratories for routine clinical application.10
- Likewise, the specificity of assays to differentiate tumor from germ-line cfNA can be a challenge and may require comparative analysis of constitutional nucleic acids to minimize false-positive results.
The ability to detect, diagnose, and monitor disease using blood-based biomarkers brings promise to a new paradigm that may eventually obviate the need for invasive procedures such as tumor biopsies. Current efforts should be focused on optimizing these procedures to improve their sensitivity, specificity, reliability, and validity such that they can be implemented in a
regulatory-approved setting and applied in clinical practice.
Dr. Siu is a Professor of Medicine at the University of Toronto and the Director of the Phase I Clinical Trials Program at the Princess Margaret Cancer Centre. She is a member of the ASCO Board of Directors.
- Tran B, Dancey JE, Kamel-Reid S, et al. J Clin Oncol. 2012;30:647-60. PMID: 22271477.
- Doroshow JH. Vision of Molecular Targeted Therapies in Colon Cancer in 2012. Presented at: 2012 Gastrointestinal Cancers Symposium; January 21, 2012; San Francisco, CA.
- Dowlati A, Haaga J, Remick SC, et al. Clin Cancer Res. 2001;7:2971-6. PMID: 11595684.
- Overman MJ, Modak J, Kopetz S, et al. J Clin Oncol. 2013;31:17-22. PMID: 23129736.
- Olson EM, Lin NU, Krop IE, et al. Nat Rev Clin Oncol. 2011;8:620-5. PMID: 21808265.
- Peppercorn J, Shapira I, Collyar D, et al. J Clin Oncol. 2010;28:2635-40. PMID: 20406927.
- Agulnik M, Oza AM, Pond GR, et al. J Clin Oncol. 2006;24:4801-7. PMID: 17050865.
- Schwarzenbach H, Hoon DS, Pantel K. Nat Rev Cancer. 2011;11:426-37. PMID: 21562580.
- Perkins G, Yap TA, Pope L, et al. PLoS One. 2012;7:e47020. PMID: 23144797.
- Forshew T, Murtaza M, Parkinson C, et al. Sci Transl Med. 2012;4:136ra68. PMID: 22649089.
Alternatives to Tissue Acquisition for Personalized Therapy: Viable Options or Scientific Parlor Trick?
By Luis A. Diaz, Jr., MD
The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins
Many of the major advances in targeted therapy in oncology have relied on the acquisition of tumor tissue prior to initiating therapy or following the onset of resistance. The availability of tissue for molecular analyses has been instrumental in understanding the primary mechanism of action of agents like trastuzumab, imatinib, panitumumab, erlotinib, crizotinib, and vemurafenib, where the molecular partner for predicting response has been defined (i.e., HER2/neu amplification, mutant KIT, mutant KRAS and mutant EGFR, ALK rearrangements, and mutant BRAF, respectively). Likewise, access to tumor tissue following clinical resistance has helped define mechanisms of secondary resistance to these targeted agents, often in the very pathway or gene that defined their responsiveness. So it seems obvious, if not mandatory, that all drug development should require pre- and post-treatment biopsies to ascertain the molecular genotype to guide the therapeutic phenotype.
Drawbacks of the biopsy approach
As with many things in translational medicine, several barriers exist to bridging the bench to the bedside. In terms of tissue acquisition, there are four major barriers that require real consideration. The first two relate to clinical topics, physician and patient opposition, and procedural toxicity. When samples are available from a cancer surgery, the availability of tissue is not an issue. However, biopsies as part of a clinical trial are complicated by physician and patient reluctance, especially since these biopsies will often not influence the outcome of the patient at hand. Biopsies are an inconvenience from a scheduling perspective, increase the cost of patient care, and are another uncomfortable, invasive procedure for patients. Furthermore, biopsies are not without complications. A recent review of the investigative biopsy experience at the MD Anderson Cancer Center, by Overman et al., reported a complication rate for thoracic biopsies of 17.1% and 1.6% for abdominal/pelvic sampling.1
There are also technical barriers to tissue acquisition that require discussion: sample characteristics and tumor heterogeneity. Following a biopsy, the majority of tumor tissue is preserved in formalin-fixed paraffin-embedded blocks (FFPE), which crosslinks DNA to the point that a large fraction of archived FFPE samples have been reported to be inadequate for molecular analysis,2 which is why freezing is the ideal choice for preserving tumor tissue. In each of these blocks of tumor tissue, the amount of tumor is dependent on the tumor cellularity (% tumor) and the size of the section of tumor. Some tumors have a high percentage of tumor cellularity (colon cancer, sarcomas, renal cell carcinomas) while other tumors have poor tumor cellularity because of necrotic tissue or stromal contamination (pancreatic cancer, glioblastoma). This is further compounded by low tissue amounts present from fine-needle aspirates and core needle biopsies, which provide a very small amount of tumor tissue for analysis in comparison to surgically resected tumors.
Tumor heterogeneity also proves to be problematic. Tumors themselves are heterogeneous, with different areas of the same tumor showing different genetic profiles (intratumoral heterogeneity); likewise, heterogeneity exists between metastases within the same patient (intermetastatic heterogeneity). A biopsy or tissue section from one part of a solitary tumor will miss the molecular intratumoral as well as intermetastatic heterogeneity. Taken together, the quality of the molecular information derived from any biopsy depends on how well the sample accounts for tumor cellularity, method of preservation, molecular and tissue heterogeneity, and quantity of tissue available for analysis.
Despite these limitations, tissue remains the “gold standard” for molecular analyses largely because of the abundance of tumor material in a sample and the clonal nature of cancer, where the key driver mutations are consistent even in multiple subclones throughout tumors and their metastases.
Recent developments in digital genomics have resulted in the emergence of novel approaches that may be able to circumvent the issues with tissue biopsies for personalized therapy. Circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA) can provide the same genetic information available in a tissue biopsy necessary to interrogate key companion diagnostics, and accessing the blood stream has clear advantages. For one, both of these are sources of fresh DNA, unhampered by preservatives. Sampling the blood from a needle stick is virtually noninvasive and therefore avoids the dangers of biopsies. Furthermore, blood can be drawn at any time during the course of therapy and allow for dynamic monitoring of molecular changes in the tumor rather than relying on a static time point.
What molecular information can be derived from CTCs and ctDNA? CTCs are shed from the tumor(s) into the circulation, and single-cell analysis of CTCs has demonstrated the potential of evaluating DNA and RNA for mutations, rearrangements, and expression changes. These analyses are still investigational tools but will likely be available for clinical use in the near future.
ctDNA are small fragments of DNA released into the circulation during tumor cell turnover. This is a normal homeostatic process for most tumors and virtually all advanced cancers shed detectible levels of ctDNA in the blood.3 These fragments are identified by the mutations present within them. In other words, these fragments are components of the cancer’s genome and therefore contain the same genetic defects as the tumor or tumors themselves. These defects span the types of genomic alterations identified in tumor and include point mutations (KRAS),4 rearrangements,5 amplifications (HER2), and even aneuploidy.6 Moreover, interrogating plasma from patients can account for molecular heterogeneity since the circulation is a collecting pool for ctDNA fragments from all of the tumors in a patient’s body.7
The evaluation of CTCs and ctDNA for genetic alterations present in the tumor tissue are, in reality, a liquid biopsy. The purpose of this test will be to identify the molecular profile of a tumor through a simple blood collection. While much of the current data for CTCs and ctDNA are investigational, the sensitivity of these liquid biopsies for patients with stage IV disease, at least for ctDNA, appears to be approaching > 99%.3,7 Future studies will determine the viability of these tests for clinical use, but at this point they do provide an alternative to tissue biopsies and are an option, especially in clinical trials, for serial collections for determinations of molecular resistance or genotyping for eligibility in a clinical trial.
Dr. Diaz is an Associate Professor of Oncology at Johns Hopkins. He is a member of ASCO’s Scientific Program Committee in the Developmental Therapeutics—Experimental Therapeutics Track.
- Overman MJ, Modak J, Kopetz S, et al. J Clin Oncol. 2013;31:17-22. PMID: 23129736.
- Tol J, Koopman M, Cats A, et al. N Engl J Med. 2009;360:563-72. PMID: 19196673.
- Diehl F, Schmidt K, Choti MA, et al. Nat Med. 2008;14:985-90. PMID: 18670422.
- Holdhoff M, Schmidt K, Donehower R, et al. J Natl Cancer Inst. 2009;101:1284-5. PMID: 19641175.
- Leary RJ, Kinde I, Diehl F, et al. Sci Transl Med. 2010;2:20ra14. PMID: 20371490.
- Leary RJ, Sausen M, Kinde I, et al. Sci Transl Med. 2012;4:162ra154. PMID: 23197571.
- Diaz LA Jr, Williams RT, Wu J, et al. Nature. 2012;486:537-40. PMID: 22722843.
The views and opinions expressed in Current Controversies in Oncology are those of the authors alone. They do not necessarily reflect the views or positions of the Editor or of the American Society of Clinical Oncology.