Christine M. Walko and Howard L. McLeod
INTRODUCTION
The evolution of understanding cancer biology has yielded many advances that have been translated into cancer treatment. Application of this knowledge has allowed for a shift in chemotherapeutics from traditional cytotoxic agents that worked by killing both healthy and malignant fast growing cells to chemical and biologic therapies aimed at targeting a specific gene or pathway critical to the particular cancer being treated.1 This age of pathway-directed therapy has been made possible by the increased availability and feasibility of high throughput technology able to provide comprehensive and clinically useful molecular characterization of tumors. Translation of these efforts have resulted in improved degree to disease control for many common cancers including breast, colorectal, lung, and melanoma as well as long-term survival benefits for chronic myelogenous leukemia (CML), gastrointestinal stromal tumors (GIST), and childhood acute lymphoblastic leukemia (ALL).2
Pharmacogenomic-guided therapy aims the use information on DNA and RNA integrity to optimize not only the treatment choice for an individual patient, but also the dose and schedule of that treatment. The assessment of both somatic and germ-line mutations contribute to the overall individualization of cancer treatment. Somatic mutations are genetic variations found within the tumor DNA, but not DNA from the normal (germ-line) tissues, which also have functional consequences that influence disease outcomes and/or response to certain therapies. These types of mutations or biomarkers can be classified as either prognostic or predictive. Prognostic biomarkers identify subpopulations of patients with different disease courses or outcomes, independent of treatment. Predictive biomarkers identify subpopulations of patients most likely to have a response to a given therapy.3 Germ-line mutations are heritable variations found within the individual and, in practical terms, are focused on DNA markers predictive for toxicity or therapeutic outcomes of a particular therapy as well as inheritable risk of certain cancers.4Pharmacogenomic mutations in the germ line provide some explanation for the interindividual and interracial variability in drug response and toxicity. For cancer chemotherapy, where cytotoxic agents are administered at doses close to their maximal tolerable dose, and therapeutic windows are relatively narrow, minor differences in individual drug handling may lead to severe toxicities. Therefore, an understanding of the sources of this variability would lead to the possibility of individualizing dosages or influencing clinical decisions that can improve patient care. Pharmacogenomics has putative utility in therapy selection, clinical study design, and as a tool to improve understanding of the pharmacology of a medication.
The term pharmacogenetics was initially used to define inherited differences in drug effects and typically focused on individual candidate genes. The field of pharmacogenomics now includes genomewide association studies and is used to describe genetic variations in all aspects of drug absorption, distribution, metabolism, and excretion in addition to drug targets and their downstream pathways.5 Table 16.1illustrates some current clinical examples of genotype-guided cancer chemotherapy. Variations in the DNA sequences encoding these proteins may take the form of deletions, insertions, repeats, frameshift mutations, nonsense mutations, and missense mutations, resulting in an inactive, truncated, unstable, or otherwise dysfunctional protein. The most common change involves single nucleotide substitutions, called single-nucleotide polymorphisms (SNP), which occur at approximately 1 per 1,000 base pairs on the human genome. Variability in toxicity or activity can also be mediated by postgenomic events, at the level of RNA, protein, or functional activity.
PHARMACOGENOMICS OF TUMOR RESPONSE
Tumor response to chemotherapy is regulated by a complex, multigenic network of genes that encompasses inherent characteristics of the tumor, differentially activated pathways of cell signaling, proliferation and DNA repair, factors that control drug delivery to the tumor cells (e.g., metabolism, transport), and cell death. These may in turn be modulated by previously administered treatment or drug exposure, which may upregulate target proteins or activate alternative pathways of drug resistance. The polygenic nature of drug response implies that a better understanding of genotype–phenotype associations would require more than the usual single-gene pharmacogenetic strategies employed to date. However, there are instances where the genomic context of a single gene within a cancer will be of high impact for specific therapeutic agents (see Table 16.1).
Pathway Directed Anticancer Therapy
One of the earliest success stories illustrating pathway-driven therapeutics is with CML. The hallmark chromosomal abnormality of this disease is the translocation of chromosomes 9 and 22 that ultimately produces the fusion gene BCR-ABL. This discovery in 1960 eventually led to the development of the targeted tyrosine-kinase inhibitor (TKI) imatinib and its subsequent Food and Drug Administration (FDA) approval for treatment of CML in 2001.6 The International Randomized Study of Interferon and STI571 (IRIS) trial began enrollment in 2000 and compared imatinib with interferon and low-dose cytarabine, which was the previous standard of care for newly diagnosed patients with chronic-phase CML. All efficacy endpoints favored imatinib, including complete cytogenetic response of 76.2% with imatinib compared with 14.5% with interferon (p <0.001).7 Overall survival (OS) after 60 months of follow-up was 89% with imatinib.8 This example is just one of many where a once fatal disease can now be considered more akin to a chronic disease, requiring a daily medication and regular physician follow-up, similar to hypertension or diabetes. Drug development has also kept pace with these advances and now several other agents, including dasatinib, nilotinib, bosutinib, and ponatinib, have joined imatinib as treatment options for CML.
The idea of changing treatment focus from a disease-based model to a pathway-driven model is also evolving. Human epidermal growth factor receptor 2 (HER2) is a transmembrane receptor tyrosine kinase that is overexpressed or amplified in up to 25% of breast cancers. Trastuzumab is a humanized monoclonal antibody directed against HER2 and demonstrated improved response rates (RR) and time to disease progression in patients with metastatic HER2 positive breast cancer and improved disease-free survival (DFS) and OS in HER2-positive breast cancer patients treated with adjuvant trastuzumab.9 Several additional agents are now available to target the HER2 pathway and vary in their pharmacology and mechanism of action. Lapatinib is an oral TKI directed against HER2 and the epidermal growth factor receptor (EGFR), pertuzumab is a humanized monoclonal antibody that binds at a different location than trastuzumab and inhibits the dimerization and subsequent activation of HER2 signaling, and ado-trastuzumab emtansine is an antibody-drug conjugate that targets HER2-positive cells and then releases the cytotoxic antimitotic agent emtansine through liposomal degradation of the linking compound. All of these agents illustrate the progress and pharmacologic diversity of pathway-directed therapy and remain as standard of care options for HER2-positive breast cancer in either the adjuvant and/or metastatic settings.10 HER2 expression is not limited to breast cancer, however. Though less common, HER2 expression is seen in numerous solid tumors including bladder, gastric, prostate and non–small-cell lung cancer with varying degrees of incidence depending on the method of detection. Based on results from a large, open-label phase III randomized, international trial of 594 patients with gastric or gastroesophageal junction cancer expressing HER2 by either immunohistochemistry or gene amplification by fluorescence in situ hybridization, trastuzumab is also approved for treatment of metastatic gastric or gastroesophageal junction adenocarcinoma that expresses HER2. Patients randomized to chemotherapy in combination with trastuzumab had a median OS of 13.8 months compared with 11.1 months in the patients receiving chemotherapy alone (hazard ratio [HR], 0.74; 0.60 to 0.91, p = 0.0046).11 Numerous examples also support that pathway-directed therapy will cross the boundaries of disease sites and that tumor genetics will become one of the biggest determining factors for treatment.
Simple expression of the drug target does not always translate into desired clinical outcomes though. Cetuximab and panitumumab are monoclonal antibodies directed against EGFR; however, it was found that colorectal cancer (CRC) patients who did not have detectable EGFR still experienced responses to these agents similar in extent to EGFR-positive patients. Kirsten rat sarcoma viral oncogene (KRAS) is a downstream effector of the EGFR pathway. Ligand binding to EGFR on the cell surface activates pathway signaling through the KRAS-RAF-mitogen-activated protein kinase (MAPK) pathway, which is thought to control cell growth, differentiation, and apoptosis.12Eventually it was found that CRC patients with a KRAS mutation did not derive benefit from cetuximab or panitumumab. The RR in CRC receiving either cetuximab or panitumumab who were KRAS wild type was 10% to 40% compared with near zero percent in those with KRAS mutations.13 This finding was the result of a retrospective analysis of small group of patients and was confirmed in large, prospective trials. Additionally, it underscores the importance of tissue collection for biomarker assessment in trials with novel therapeutics. A recent clinical trial genomic analysis suggests that mutations in NRAS may also have value in predicting the utility of EGFR antibody therapy in colorectal cancer. Although the predictive value of KRAS mutation status in colorectal cancer has been well established in clinical trials, the role of KRAS in lung cancer and other malignancies is less well elucidated. Lung cancers harboring KRAS mutations have been shown to have less clinical benefit from the EGFR-targeted erlotinib in some trials, although this has not consistently been the case across all trials. Additionally, lung cancer KRAS mutation status does not appear to reproducibly predict clinical benefit from the EGFR-targeted monoclonal antibodies, as is the case in colorectal cancer.14 Unlike the HER2 example discussed previously, the clinical application of some genetic mutations will differ between tissue of origin.
Deeper investigations and understandings of mutations driving oncogenic pathways can also elucidate mechanisms of resistance and practical therapeutic strategies for treatment and prevention. Approximately half of all cutaneous melanomas carry mutations in BRAF, with the most common being the V600E mutation. Vemurafenib is a TKI directed against mutated BRAF that demonstrated improvements in both progression-free survival (PFS) and OS when compared with the cytotoxic agent dacarbazine in previously untreated patients with metastatic melanoma carrying the BRAF V600E mutation. Vemurafenib demonstrated a 63% relative reduction in the risk of death compared with dacarbazine (p <0.001) along with a higher response rate (48% compared with 5% for dacarbazine).15 Based on these results, vemurafenib was the first BRAF targeted TKI approved by the FDA and was soon joined by dabrafenib. Although dramatic responses to these agents have been observed, relapse almost universally occurs after a median of 6 to 8 months. Activating BRAF mutations, like V600E, result in uncontrolled activity of the MAPK pathway through activation of the downstream kinase MEK, which when phosphorylated, subsequently activates extracellular signal-regulated kinase (ERK), which ultimately translocates to the cell nucleus, resulting in cell proliferation and survival (Fig. 16.1).16 An assessment of serial biopsies from patients treated with vemurafenib suggested numerous mechanisms for acquired resistance, including the appearance of secondary mutations in MEK.17This finding supports the clinical rationale for using combination therapy with a BRAF and a MEK inhibitor. The combination of dabrafenib (BRAF inhibitor) and trametinib (MEK inhibitor) was assessed in 247 metastatic melanoma patients with BRAF V600 mutations compared with dabrafenib alone. Median PFS was 9.4 months in the combination group compared with 5.8 months in the patients who received single agent therapy (HR, 0.39; 0.25 to 0.62, p <0.001). A complete or partial response was also higher in the combination therapy group (76% compared with 54%, p = 0.03). The occurrence of cutaneous squamous cell carcinoma, a known side effect of single-agent BRAF inhibitor therapy due to paradoxical activation of RAF in nonmutated cells, was also decreased in the combination therapy group (7% compared with 19%, p = 0.09), further supporting the evidence of downstream inhibition.18Although combination therapy does prolong the time to disease progression, resistance still occurs in patients through a variety of mechanisms. Utilization of sequential biopsies and a genetic assessment will help to inform rationale combination and sequential pathway-driven therapy trials that will ultimately aid in better understanding and mitigation of common mechanism of resistance.
Although advances in basic science and drug development have translated many oncogenic driver mutations across tumor types into pathway-directed therapy, this is not the case for the majority. There are numerous examples of functionally relevant recurrent driver mutations that affect protein targets that are not currently druggable. Regardless of malignancy, one of the most commonly mutated tumor suppressors is the protein p53. Mutations can result in p53 acquiring oncogenic functions that enable proliferation, invasion, metastasis, and cell survival as well as coordinating with different proteins, such as EGFR, to enhance or inhibit its effects. However, a clinical application of p53 mutation data or directly targeting p53 has been limited, to date.19 PIK3CA encodes a catalytic subunit of phophoinositol-3 kinase (PI3K), which includes four distinct subfamily kinases involved in regulating cell growth, motility, proliferation, and survival. Direct inhibitors of the kinase, as well as downstream targets, including AKT (protein kinase B [PKB]) and mammalian target of rapamycin (mTOR), are being assessed to target these mutations. Therapeutic challenges include understanding the complex signaling network germane to each cancer and the role of kinases in each subfamily.20 Both the examples of p53 and PI3K illustrate the challenge of translating the multitude of somatic mutations into applications of available therapeutic agents.
Application of Genomewide Gene Expression Profiling to Guide Therapy
Single gene approaches may not reflect the overall complexity of genetic regulation of chemotherapy responses. Genomic strategies using global gene expression data are able to provide a more complete picture of the tumor through disease classification.21 These strategies may identify subgroups of patients with early disease that need adjuvant chemotherapy, those who will not benefit from standard therapy, or help with the selection of chemotherapy from a menu of potentially active agents. Oncotype Dx is a 21-gene assay with 16 tumor-associated genes and 5 reference genes used to predict the risk of distant local recurrence in estrogen receptor (ER)-positive, HER2-negative patients with node-negative or select node-positive breast cancer. Additionally, the test also provides predictive information on which patients may benefit from the addition of chemotherapy to hormonal therapy alone. The test ultimately reports a recurrence score (RS) on a continuous scale from zero to 100. Patients with an RS <18 are considered low risk, with a 10-year distant recurrence rate (DRR) of 6.8% (95% confidence interval [CI], 4 to 9.6); RS scores of 18 to 30 are at intermediate risk, with a 10-year DRR of 14.3% (CI, 8.3 to 20.3); and RS scores ≥31 are at high risk, with a 10-year DRR of 30.5% (CI, 23.6 to 37.4).22 Additionally, high-risk patients have the largest benefit from the addition of chemotherapy to hormonal therapy (HR, 0.26; 0.13 to 0.53), whereas low-risk patients have little benefit from the addition of chemotherapy and could consider hormonal treatment alone (HR, 1.31; 0.46 to 3.78). Intermediate risk patients are harder to classify, and clinical trials are underway to further address treatment recommendations for this group of patients.23 These type of assays are also in development and in clinical trials for a variety of other solid tumor and hematologic malignancies.
Genetic-Guided Therapy Practical Issues in Somatic Analysis
Currently, targeted DNA capture is the most common type of somatic genetic screening and involves focusing on a few relevant candidate genes followed by deeper sequencing. These types of techniques can reveal common genes associated with a particular malignancy but also may uncover a signaling pathway that would not be obviously associated with a particular histology or tumor site. Application of a next-generation sequencing assay in 40 CRC and 24 non–small-cell lung cancer (NSCLC) tissue samples that assessed 145 cancer-relevant genes demonstrated that somatic mutations were seen in 98% of the CRC tumors and 83% of the NSCLCs (Fig. 16.2).24 The evolution of sequencing strategies and decreasing costs has made whole genome sequencing more available in the clinical setting, and several companies offer commercially available tumor profiling services. Several limitations exist that currently restrict the broad clinical implementation of these assays, however. Although germ-line genetic assessments can be done on a peripheral blood sample or buccal swab, somatic assessments typically require biopsy tissue, which is often in limited supply and of varying quality or may not be feasible depending on the site of the cancer. Ongoing studies are assessing the value of liquid biopsies of circulating tumor DNA.25 Optimizing and creating uniformity in quality control of gene panel or whole-genome assessment is also needed to decrease the reporting of uncertain or erroneous identification of mutations. Once sequencing is completed, a predictive analysis is needed for the 25% to 80% of instances where variants of unknown significance are identified in genes of interest. Translation of genomic sequencing into clinical practice will require a diverse team, including pathologists, medical oncologists, surgical oncologists, information technologists, geneticists, and pharmacologists.
PHARMACOGENOMICS OF CHEMOTHERAPY DRUG TOXICITY
A drug’s disposition and pharmacodynamic effects can be influenced by a number of variables, including patient age, diet, concomitant medications, and underlying disease processes. However, an individual’s genetic constitution is an important regulator of variability in drug effect. Differences in drug effects are more pronounced between individuals compared to within an individual. Indeed, studies in monozygotic and dizygotic twins identified that 20% to 80% of the variation in drug disposition is mediated by inheritance.26 Drug-metabolizing enzymes, cellular transporters, and tissue receptors are governed by genetic variation.
Advances in the treatment of most common malignancies have resulted in the availability of multiple distinct combination chemotherapy regimens with similar or equal anticancer efficacy. Therefore, differences in systemic toxicity have become a major determinant in the selection of therapy. The majority of pharmacogenomic examples affecting adverse events or efficacy from cytotoxic drugs involve hepatic metabolizing enzymes that detoxify or biotransform xenobiotics.27,28
Thiopurine Methyltransferase
One of the best-studied pharmacogenetic syndrome involves the metabolism of the thiopurine drugs—6-mercaptopurine (6MP), 6-thioguanine, and azathioprine—which have wide applications, including maintenance therapy for childhood ALL and adult leukemias. These prodrugs must be activated to thioguanine nucleotides in order to have antiproliferative effects. However, most of the variability in the formation of active metabolites is mediated by methylation via thiopurine methyltransferase (TPMT).29 TPMT is a cytosolic enzyme that catalyzes S-methylation of thiopurine agents, resulting in an inactive metabolite. Erythrocyte TPMT activity has a trimodal distribution, with 90% of patients having high activity, 10% intermediate activity, and 0.3% with very low or no detectable activity. TPMT deficiency results in higher intracellular activation of 6MP to form thioguanine nucleotides, resulting in severe or fatal hematologic toxicity from standard doses of therapy.30 The variable activity results from polymorphism in the TPMT gene, located on chromosome locus 6p22.3. Genetic variants at codon 238 (TPMT*2), codon 719 (TPMT*3C), or both codons 460 and 719 (TPMT*3A) are the most clinically significant, accounting for 95% of the patients with reduced TPMT activity.31 Heterozygotes (one wild type and one variant allele) are common (10% of patients), and have elevated levels of active metabolites (twofold more than homozygous wild type), and required more cumulative dose reductions of 6MP for maintenance ALL chemotherapy compared to homozygous wild-type patients (Fig. 16.3).32 Patients with a homozygous variant TPMT genotype are at a fourfold risk of severe toxicity, compared with wild-type patients.31 TPMT genotype tests are now available commercially in a Clinical Laboratory Improvement Amendments (CLIA)-certified environment. To date, patients homozygous for TPMT variant alleles appear to tolerate 10%, and heterozygotes appear to tolerate 65% of the recommended doses of 6MP, with no apparent decrease in clinical efficacy (Fig. 16.3).32 This has formed the basis for prospective, TPMT genotype-guided dosing of 6MP to avoid severe toxicity. Clinical Pharmacogenomics Implementation Consortium (CPIC) Guidelines recommend that homozygous wild-type patients be started at the full standard dose. Heterozygous patients should start with reduced doses at 30% to 70% of the full dose with adjustments made after 2 to 4 weeks based on myelosuppression and disease-specific guidelines. Homozygous variant patients should start with 10% of the full dose due to the extremely high levels of the active metabolite and potential for fatal toxicity at standard doses. Adjustments should be made after 4 to 6 weeks based on myelosuppression and disease-specific guidelines.33
Dihydropyrimidine Dehydrogenase (DPD)
Although 5-fluorouracil (5FU) has been available for over 40 years, it remains the cornerstone of colorectal cancer chemotherapy, both in the adjuvant and metastatic settings. Additionally, the oral prodrug capecitabine ultimately undergoes activation to 5FU and is commonly used in gastrointestinal and breast malignancies. 5FU is a prodrug that is activated intracellularly to 5-fluoro-2′-deoxyuridine monophosphate (5FdUMP), which inhibits thymidylate synthase (TS), among other mechanisms of action. TS inhibition results in impaired de novo pyrimidine synthesis and suppression of DNA synthesis. Approximately 85% of a 5FU dose is catabolized by dihydropyrimidine dehydrogenase (DPD) to inactive metabolites. Therefore, DPD is a primary regulator of 5FU activity. DPD deficiency has been described, resulting in higher 5FU blood levels, greater formation of active metabolites, and severe or fatal clinical toxicity, predominately myelosuppression, mucositis, and cerebellar toxicity.34 In theory, this toxicity could be reduced or avoided by screening for DPD activity in surrogate tissues, such as peripheral mononuclear cells. However, the technical requirements for preparation of these samples make it impractical for many practice sites. Understanding the molecular basis for DPD deficiency will provide an approach for prospective identification of patients at high risk for severe 5FU toxicity. The gene encoding DPD is composed of 23 exons, and at least 23 SNPs have been found.35 Studies in DPD-deficient patients have identified several distinct molecular variants associated with low enzyme activity. Many of these are rare, and base substitutions, splicing defects, and frame shift mutations, have been described. The prevalent variation is the splice recognition site in intron 14 (DPYD*2A), where a G to A substitution results in the skipping of exon 14, resulting in an inactive enzyme.36–38 This polymorphism has been associated with severe DPD deficiency in heterozygous patients, with a homozygous genotype associated with a mental retardation syndrome. Patients with severe 5FU toxicity may harbor one or more variant alleles of DPD, and a recent study showed that 61% of cancer patients experiencing severe 5FU toxicities had decreased DPD activity in peripheral mononuclear cells, and DPYD*2A was commonly found.39 In the patients with grade 4 neutropenia, 50% harbored at least one DPYD*2A. It is estimated that in the Caucasian population, homozygotes for the variant alleles have an incidence of 0.1% and heterozygotes occur at an incidence of 0.5% to 2%. There are additional DPD mutations that have been associated with impaired enzyme activity, including DPYD *3 and DPYD*13. CPIC guidelines recommend standard dosing for homozygous wild-type patients. Reducing the dose by at least 50% in heterozygous patients (*1/*2A) is recommended, followed by dose adjustment based on toxicity and/or pharmacokinetic testing. The use of an alternative agent is recommended in homozygous-variant patients (*2A/*2A).34 There are many patients with severe 5FU toxicity that have normal DPD activity. This highlights that many factors, including multiple genes, are potential causes of 5FU toxicity, and there will not be one simple test to avoid this important clinical problem.
Cytochrome P450 2D6
Tamoxifen is a selective estrogen-receptor modulator used in ER-positive breast cancer in both the localized and metastatic settings. It is the drug of choice for premenopausal women and is a treatment option, along with aromatase inhibitors, for postmenopausal women. The low cost of tamoxifen also makes it a preferred therapy regardless of menopausal status in numerous countries. Tamoxifen metabolism is complex, with extensive metabolism through numerous phase I and II enzymes that produce several primary and secondary metabolites and their corresponding isomers, each possessing different antiestrogen effects.40 The primary active metabolite is believed to be endoxifen, which is produced by the CYP3A4/5 mediated-conversion of tamoxifen to N-desmethyltamoxifen, which is then further converted to endoxifen (4-hydroxy-N-desmethyltamoxifen) via cytochrome P450 2D6 (CYP2D6). A direct relationship between endoxifen concentration and its antiestrogen effects has been demonstrated, potentially suggesting that a threshold concentration may be needed for optimal clinical effect.41 CYP2D6 is highly polymorphic, with more than 80 allelic CYP2D6 variants described. These alleles vary in enzyme activity and prevalence with respect to race and ethnicity.42 Based on genotype, patients can be classified by phenotype into ultrarapid metabolizers (UM; approximately 1% to 2% of patients [common alleles include *1xN, *2xN]) who carry more than two functional allele copies, extensive metabolizers (EM; 77% to 92% [e.g., *1, *2]), intermediate metabolizers (IM; 2% to 11% [e.g., *10, *17, *41]), or poor metabolizers (PM; 5% to 10% [*3, *4, *5]).43 UM patients have the highest concentrations of endoxifen, followed by EM patients, then IM patients, and finally, PM patients have the lowest concentration. Up to a sixfold variation in endoxifen levels may be seen between homozygous PM and homozygous EM patients.40
The relationship between CYP2D6 genotype, endoxifen concentrations, and disease outcomes has been investigated in numerous clinical trials. One of the largest retrospective trials assessed this relationship in 1,325 women treated with adjuvant tamoxifen 20 mg daily. Approximately 46% of the patients were classified as EM, 48% were IM, and 5.9% were PM. A statistically significant increased risk of disease recurrence was seen in the IM and PM patients compared with the EM patients (HR, 1.40; 95% CI, 1.04 to 1.90 for IM; and HR 1.90, 95% CI, 1.10 to 3.28 for PM).44 A large meta-analysis of 4,973 tamoxifen-treated patients across 12 international studies conducted by the International Tamoxifen Pharmacogenomics Consortium also supported this relationship. CYP2D6 PM phenotypes were associated with decreased DFS (HR 1.25, 95% CI, 1.06 to 1.47, p = 0.009) when only considering the data from trials with postmenopausal women with ER-positive breast cancer who received tamoxifen 20 mg daily for 5 years.45
Not all trial results have been consistent, however, and dosing guidelines for genotype-guided therapy do not yet exist. Clinical trials do support the potential for genotype-guided therapy. IM patients who received an increased dose of 40 mg daily instead of the standard 20 mg were shown to have endoxifen concentrations similar to that of EM patients (p = 0.25).46 This suggests that genotype-guided therapy with increased dose recommendations may be feasible, but additional prospective trials are needed to determine the clinical efficacy of this intervention.
CONCLUSIONS AND FUTURE DIRECTIONS
Genomic-driven cancer medicine is being translated into clinical practice through increased understanding of somatic mutations in a specific tumor that can be translated to pathway-directed therapeutics as well as germ-line mutations that affect the pharmacokinetics and pharmacodynamics of individual medications. For the practicing oncologist, knowledge of pharmacogenomics is necessary because therapeutic decisions of drug selection and dosage are being based on more molecularly and genetically defined variables than the current phenotypic information of tumor type, immunohistochemistry, and body surface area. Health-care policy changes preferring the bundling of care and reimbursement based on diagnosis coding may further drive individualized therapy where the goal is to optimize both treatment responses while minimizing toxicity. However, with advances always come challenges. Reimbursement for multiplex genomic testing is not universal, so deciding who and when to initiate testing is a consideration. Optimizing turnaround time, especially for referral patients who have had biopsies performed elsewhere, will require requesting this archived tissue prior to or during the initial patient visit to facilitate minimizing treatment delays. Although some variants have strong evidence supporting treatment recommendations, many currently do not yet. Multidisciplinary committees charged with reviewing the level of evidence for each genetic result and providing clinically actionable recommendations will be essential for translating these multigene tumor assay results into routine clinical practice. Decision tools and development of treatment guidelines will further assist with routine integration of this technology, especially for oncologists at smaller practice sites. Oncology fellowship training programs will also need to be expanded to ensure competence of new practitioners in the area of genomic-guided therapies.
Regardless of these challenges, the treatment paradigm of genomic-driven medicine and individualizing therapy has permitted the field of oncology to move beyond the limitations of nonselective cytotoxic therapy and toward the more optimal selection and dosing of oncology agents.
REFERENCES
1. McLeod HL. Cancer pharmacogenomics: early promise, but concerted effort needed. Science 2013;339:1563–1566.
2. Garraway LA. Genomics-driven oncology: framework for an emerging paradigm. J Clin Oncol 2013;31:1806–1814.
3. Mandrekar SJ, Sargent DJ. Predictive biomarker validation in practice: lessons from real trials. Clin Trials 2010;7:567–573.
4. Evans WE, Relling MV. Pharmacogenomics: translating functional genomics into rational therapeutics. Science 1999;286:487–491.
5. Wang L, McLeod HL, Weinshilboum RM. Genomics and drug response. N Engl J Med 2011;364:1144–1153.
6. Druker BJ. Translation of the Philadelphia chromosome into therapy for CML. Blood 2008;112:4808–4817.
7. O’Brien SG, Guilhot F, Larson RA, et al. Imatinib compared with interferon and low-dose cytarabine for newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med 2003;348:994–1004.
8. Druker BJ, Guilhot F, O’Brien SG, et al. Five-year follow-up of patients receiving imatinib for chronic myeloid leukemia. N Engl J Med 2006;355:2408–2417.
9. Hudis CA. Trastuzumab—mechanism of action and use in clinical practice. N Engl J Med 2007;357:39–51.
10. Figueroa-Magalhaes MC, Jelovac D, Connolly RM, et al. Treatment of HER2-positive breast cancer. Breast 2014;23:128–136.
11. Bang YJ, Van Cutsem E, Feyereislova A, et al. Trastuzumab in combination with chemotherapy versus chemotherapy alone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label, randomised controlled trial. Lancet 2010;76:687–697.
12. Bardelli A, Siena S. Molecular mechanisms of resistance to cetuximab and panitumumab in colorectal cancer. J Clin Oncol 2010;28:1254–1261.
13. Jimeno A, Messersmith WA, Hirsch FR, et al. KRAS mutations and sensitivity to epidermal growth factor receptor inhibitors in colorectal cancer: practical application of patient selection. J Clin Oncol 2009;27:1130–1136.
14. Roberts PJ, Stinchcombe TE, Der CJ, et al. Personalized medicine in non-small-cell lung cancer: is KRAS a useful marker in selecting patients for epidermal growth factor receptor-targeted therapy? J Clin Oncol 2010;28:4769–4777.
15. Chapman PB, Hauschild A, Robert C, et al. Improved survival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med 2011;364:2507–2516.
16. Dhillon AS, Hagan S, Rath O, et al. MAP kinase signalling pathways in cancer. Oncogene 2007;26:3279–3290.
17. Trunzer K, Pavlick AC, Schuchter L, et al. Pharmacodynamic effects and mechanisms of resistance to vemurafenib in patients with metastatic melanoma. J Clin Oncol 2013;31:1767–1774.
18. Flaherty KT, Infante JR, Daud A, et al. Combined BRAF and MEK inhibition in melanoma with BRAF V600 mutations. N Engl J Med 2012;367:1694–1703.
19. Muller PA, Vousden KH. p53 mutations in cancer. Nat Cell Biol 2013;15:2–8.
20. Clarke PA, Workman P. Phosphatidylinositide-3-kinase inhibitors: addressing questions of isoform selectivity and pharmacodynamic/predictive biomarkers in early clinical trials. J Clin Oncol 2012;30:331–333.
21. Ramaswamy S, Golub T. DNA microarrays in clinical oncology. J Clin Oncol 2002;20:1932–1941.
22. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med 2004;351:2817–2826.
23. Paik S, Tang G, Shak S, et al. Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer. J Clin Oncol 2006;24:3726–3734.
24. Lipson D, Capelletti M, Yelensky R, et al. Identification of new ALK and RET gene fusions from colorectal and lung cancer biopsies. Nat Med 2012;18:382–384.
25. Diaz LA Jr, Bardelli A. Liquid biopsies: genotyping circulating tumor DNA. J Clin Oncol 2014;32:579–586.
26. Watters J, McLeod H. Cancer pharmacogenomics: current and future applications. Biochim Biophys Acta 2003;1603:99–111.
27. Evans W, Relling M. Pharmacogenomics: translating functional genomics into rational therapeutics. Science 1999;286:487–491.
28. Deenen MJ, Cats A, Beijnen H, et al. Part 2: pharmacogenetic variability in drug transport and phase I anticancer drug metabolism. Oncologist 2011;16:820–834.
29. Krynetski E, Evans W. Drug methylation in cancer therapy: lessons from the TPMT polymorphism. Oncogene 2003;22:7403–7413.
30. McLeod H, Krynetski EY, Relling MV, et al. Genetic polymorphism of thiopurine methyltransferase and its clinical relevance for childhood acute lymphoblastic leukemia. Leukemia 2000;14:567–572.
31. Evans W, Hon YY, Bomgaars L, et al. Preponderance of thiopurine S-methyltransferase deficiency and heterozygosity among patients intolerant to mercaptopurine or azathioprine. J Clin Oncol 2001;19:2293–2301.
32. Relling M, Hancock ML, Rivera GK, et al. Mercaptopurine therapy intolerance and heterozygosity at the thiopurine S-methyltransferase gene locus. J Natl Cancer Inst 1999;91:2001–2008.
33. Relling MV, Gardner EE, Sandborn WJ, et al. Clinical pharmacogenetics implementation consortium guidelines for thiopurine methyltransferase genotype and thiopurine dosing: 2013 update. Clin Pharmacol Ther 2013;93:324–325.
34. Caudle KE, Thorn CF, Klein TE, et al. Clinical Pharmacogenetics Implementation Consortium guidelines for dihydropyrimidine dehydrogenase genotype and fluoropyrimidine dosing. Clin Pharmacol Ther 2013;94:640–645.
35. McLeod H, Collie-Duguid ES, Vreken P, et al. Nomenclature for human DPYD alleles. Pharmacogenetics 1998;8:455–459.
36. Wei X, Elizondo G, Sapone A, et al. Characterization of the human dihydropyrimidine dehydrogenase gene. Genomics 1998;51:391–400.
37. Ridge S, Sludden J, Wei X, et al. Dihydropyrimidine dehydrogenase pharmacogenetics in patients with colorectal cancer. Br J Cancer 1998;77:497–500.
38. Johnson M, Wang K, Diasio R. Profound dihydropyrimidine dehydrogenase deficiency resulting from a novel compound heterozygote genotype. Clin Cancer Res 2002;8:768–774.
39. Van Kuilenburg A, Meinsma R, Zoetekouw L, et al. Increased risk of grade IV neutropenia after administration of 5-fluorouracil due to a dihydropyrimidine dehydrogenase deficiency: high prevalence of the IVS14+1g>a mutation. Int J Cancer 2002;101:253–258.
40. Mürdter TE, Schroth W, Bacchus-Gerybadze L, et al. Activity levels of tamoxifen metabolites at the estrogen receptor and the impact of genetic polymorphisms of phase I and II enzymes on their concentration levels in plasma. Clin Pharmacol Ther 2011;89:708–717.
41. Desta Z, Ward BA, Soukhova NV, et al. Comprehensive evaluation of tamoxifen sequential biotransformation by the human cytochrome P450 system in vitro: prominent roles for CYP3A and CYP2D6. J Pharmacol Exp Ther 2004;310:1062–1075.
42. Bradford LD. CYP2D6 allele frequency in European Caucasians, Asians, Africans and their descendants. Pharmacogenomics 2002;3:229–243.
43. Crews KR, Gaedigk A, Dunnenberger HM, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines for codeine therapy in the context of cytochrome P450 2D6 (CYP2D6) genotype. Clin Pharmacol Ther 2012;91:321–326.
44. Schroth W, Goetz MP, Hamann U, et al. Association between CYP2D6 polymorphisms and outcomes among women with early stage breast cancer treated with tamoxifen. JAMA 2009;302:1429–1436.
45. Province MA, Goetz MP, Brauch H, et al. CYP2D6 genotype and adjuvant tamoxifen: meta-analysis of heterogeneous study populations. Clin Pharmacol Ther 2014;95:216–227.
46. Irvin WJ Jr, Walko CM, Weck KE, et al. Genotype-guided tamoxifen dosing increases active metabolite exposure in women with reduced CYP2D6 metabolism: a multicenter study. J Clin Oncol 2011;29:3232–3239.