The Basic Principals of Pharmacogenetics Testing in Cancer Treatment A

The Basic Principals of Pharmacogenetics Testing in Cancer Treatment A Bojana M. Cikota-Aleksić1, Nemanja K. Rančić1, Nenad G. Ratković2, Viktorija M. Dragojević-Simić1 A 1 Center for clinical pharmacology; Military Medical Academy; Faculty of Medicine of Military Medical Academy, University of Defence, Belgrade, Serbia 2 Sector for Treatment; Military Medical Academy; Faculty of Medicine of Military Medical Academy, University of Defence, Belgrade, Serbia


INTRODUCTION
Th e term "precision medicine" is relatively new, but the concept that takes into account genetic, environmental and lifestyle factors, has long been a part of therapeutic and preventive procedures [1]. Although precision medicine may be applied in diff erent fi elds of medicine, oncology is of particular interest due to the increasing incidence of malignant diseases worldwide, high mortality rates and common usage of toxic and disfi guring therapies [2]. Increased knowledge about molecular pathways underlying cancer has begun to infl uence risk assessment, diagnostic classification' and therapeutic strategies years before Precision Medicine Initiative was announced.
Among successful examples is the usage of imatinib, the tyrosine kinase inhibitor that revolutionized the treatment of patients with chronic myeloid leukemia (CML) (the inventors were awarded for "converting fatal cancer into a manageable condition" in 2009) [3]. It should be emphasized that CML patients are mandatorily tested for the presence of t(9;22) (Philadelphia chromosome) prior to imatinib administration. Th e introduction of polymerase chain reaction (PCR) -based methods into routine diagnostic practice enabled detection of BCR-ABL1 fusion gene transcript, quantifi cation of residual disease during/aft er the therapy course, and detection of mutations related to imatinib resistance. Th e number of therapies tailored to specifi c genetic features has been continuously increasing and some of these therapies showed signifi cant benefi ts. According to the last update, the US Food and Drug Administration (FDA) listed more than 150 drugs for the cancer treatment that contains information on genomic biomarkers in the labeling [4]. We are not so close to the moment when "matching a cancer cure to our genetics code is just as easy", but pharmacogenetics testing became the part of routine management in patients with breast, lung and colorectal cancers, as well as melanomas and hematologic malignancies. Th ese tests play an important role in identifying potential responders, avoiding adverse events and optimizing drug dose.

METHODS
Th is article provides an overview of currently available pharmacogeneticss methods that are driving precision cancer treatment. Th e article addresses only biomarkers in drug labels analyzed by methods of cytogenetics or molecular genetic. Since the paper mainly discusses changes in individual genes, we prefer to use the term pharmacogeneticss over pharmacogenomics.
We searched PubMed, Google Scholar, SCIndex, Dimension, Scopus and Google for English and Serbian language abstracts, using the searching terms "pharmacogeneticss", "polymerase chain reaction", "next-generation sequencing", "comprehensive genome profi ling", "precision medicine", "hotspots", "carcinoma", "cancer" and "cancer treatment". Based on expert selection review, we chose both open and blinded studies, reviews and metaanalysis, and available comments and editorials, related to the MESH terms.

Samples for pharmacogenetics testing
Pharmacogenetics biomarkers in the drug labels concern variations in the germline genome (germline variants) or somatic mutations.
Th e majority of cancer patients are treated with cytotoxic chemotherapy which is not targeted to specifi c mutations in tumor genome. However, variations in the toxicity and effi cacy have been recorded between the patients, indicating that chemotherapy-induced phenotype is related to both somatic and hereditary (germline) variations. Pharmacogenetics analyses of germline variants commonly include single nucleotide polymorphisms (SNPs) in genes encoding drug metabolizing enzymes or drug transporters [5,6,7,8,9]. Well-known example includes thiopurine Smethyltransferase defi ciency due to missense mutations in thiopurine S-methyltransferase (TPMT) gene. Treatment of acute lymphoblastic leukemia (ALL) with standard doses of mercaptopurine results in severe toxicity in patients with mutated TPMT. Both FDA and European Medicine Agency (EMA) recommend TPMT genotyping since patients with inactive alleles may be successfully treated with reduced doses of mercaptopurine [6]. Similarly, the usage of 5-fl uorouracil (FU)/capecitabine leads to moderate or severe toxicity in up to 40% of patients as a result of low dihydropyrimidine dehydrogenase (DPD) activity. Th e majority of DPD defi cient patients are carriers of mutations in DPYD, corresponding genes. Even though the association between DPYD genotypes and 5-FU/capecitabine effi cacy has not been completely characterized, FDA and also Th e Netherlands National guideline for colorectal carcinoma recommends DPD and DPYD testing prior to treatment [10]. TPMT, DPYD, and other germline variants are commonly analyzed in peripheral blood samples. Since tumor samples represent a mixture of malignant and normal cells, they are unreliable in refl ecting the germline genotype and should be avoided in germline analyses. In patients with hematological malignancies, saliva may be considered as a sample of choice [6]. Table 1 presents an overview of germline variants relevant to cancer treatment.
Diff erently to germline variants that commonly aff ect genes encoding drug-metabolizing enzymes and drug transporters, somatic mutations aff ect tumor suppressor genes, oncogenes or genes involved in DNA reparation, and usually represent "cancer drivers". Pharmacogenetics testing of somatic mutations is used to identify candidates for diff erent targeted therapies and includes detection of base substitutions, nucleotide insertions/ deletions, a variable number of (nucleotides) tandem repeats, (gene) copy number variations, chromosomal translocations, and altered gene expression. Cancer drivers may be targets for the development of both safer and more effi cient therapies. So far, examples include the success of imatinib in Philadelphia chromosome-positive CML, trastuzumab in breast cancer with human epidermal growth factor receptor 2 (HER2) amplifi cation, vemurafenib in B-Raf proto-oncogene (BRAF) -mutated melanoma, erlotinib in epidermal growth factor receptor (EGFR)-mutated non-small cell lung cancer (NSCLC) and other [11,12,13]. Pharmacogenetics testing related to targeted therapy is performed on tumor tissue, commonly formalin-fi xed paraffi n-embedded (FFPE); in patients with leukemia, bone marrow aspirates/blood samples are analyzed. However, in some patients tumors are inaccessible or biopsy is contraindicated due to unfavorable clinical conditions. In recent years numerous research projects have been focused on circulating DNA derived from necrotic or apoptotic tumor cells (ctDNA). Analysis of ctDNA from serum/plasma ("liquid biopsy") may be a powerful tool in pharmacogenetics because it refl ects molecular heterogeneity and evolution over time of both, tumor and metastasis. EGFR testing demonstrated high concordance between tumor and liquid biopsies of NSCLC. Th e diagnostic value of ctDNA was also confi rmed in colorectal cancer [14]. So far, FDA approved the use of gefi tinib in NSCLC patients with EGFR mutations detected in ctDNA, but only when it's impossible to perform a tissue biopsy [15].

Methods in pharmacogenetics testing
Pharmacogenetics biomarkers can be determined by cytogenetics, fl uorescent in situ hibridization (FISH), a variety of polymerase chain reaction (PCR)-based methods, or by DNA sequencing. Germline variants are commonly analyzed by Real-Time (RQ) PCR, PCR-RFLP (restriction fragment length polymorphism) or PCR-ARMS (allele refractory mutation system) [16,17]. Th e majority of laboratories use in-house methods validated on regional/country population.
Detection of genetic variations in tumor cells employs a broader spectrum of methods. Regarding the number of analyzed markers, these methods can be divided into three distinct categories. Th e fi rst category includes methods that analyze/detect one target. For example, RQ-PCR is used to detect BRAF V600E mutation in melanoma; reverse transcriptase (RT) PCR is used for detection and quantifi cation of breakpoint cluster region-proto-oncogene tyrosine-protein kinase (BCR-ABL1) fusion gene transcript in CML; FISH is a gold standard for detection of Echinoderm microtubule-associated protein-like 4-anaplastic lymphoma kinase (EML4-ALK) gene rearrangement in NSCLC, etc. Detection of somatic mutations in pharmacogenomic testing can be performed by in-house methods. However, the usage of CE marking IVD (in vitro diagnostic device) tests (devices complied with the European in vitro Diagnostic Devices Directive 98/79/EC) or FDA recommended tests is preferred [18].
Th e second category of methods includes next-generation sequencing (NGS) to detect the most frequent variations ("hotspots") that occur in limited areas of genes of interest. Optimized NGS panels for a wide range of cancers (breast cancer, colorectal cancer, melanoma, lung cancer, glioma, prostate cancer, ovarian cancer, sarcoma) and hematological malignancies (myeloid malignancies, chronic lymphocytic leukemia) are off ered by companies, or laboratories can contact companies to create NGS panels that meet their requirements. Th e second category of methods also includes EndoPredict® (Myriad Genetics), a multigene test that combines the expression of cancer-related genes with con-898 Volume 7 • Number 1 • April 2020 • HOPH ventional prognostic factors (nodal status and tumor size) to predict the likelihood of disease progression/ metastases and thus guide treatment in ER+/Her2-early-stage breast cancer patients. Th e clinical value of EndoPredict® and similar tests (Oncotype DX™, MammaP-rint®, Genomic Grade Index®, PAM50™, Breast Cancer Index™) was confi rmed in practice (FDA approvеd) [19,20].
Conventional hotspot methods that detect one or more variants in particular area(s) of the gene(s) are (and will be) important part of diagnostic and treatment procedures in patients with malignant diseases. However, these methods cannot address the increasing complexity of therapeutically relevant genomic and clinical information [21]. Th e third category of methods includes comprehensive genomic profi ling (CGP), which is testing of all the known clinically relevant cancer genes for the most common classes of alterations. Th e CGP has already become an important procedure in the management of patients with cancer, as discussed in the text below. Th e list of FDA approved pharmacogenetics tests is available at http://www.fda.gov/ companiondiagnostics.

Comprehensive genomic profi ling
Th e main advantage of CGP over the hotspot testing is an analysis of the most common variants (base substitutions, insertions/deletions of nucleotides, gene rearrangements and copy number alterations) in all clinically relevant cancer genes in one run. Considering potential complications, scarce biopsy and frequent inability to repeat the procedure, the possibility of CGP in one assay became a valuable tool in oncology practice [22]. For that reason, guidelines for advanced/metastatic NSCLC recommend that molecular testing such as EGFR, anaplastic lymphoma kinase (ALK), C-ROS Proto-Oncogene 1 (ROS1), B-Raf proto-oncogene (BRAF) should be conducted as a part of broader molecular profi ling. In addition, published data clearly imply lower sensitivity of conventional hotspot methods compared to CGP (e.g. 17% of EGFR exon 9 deletions are missed by hotspot testing, 35% of ALK rearrangements are missed by FISH, etc.) [23]. Th us, CGP detects gene alterations missed by hotspot testing and potentially converts "noncandidates" to "candidates" for targeted therapy. Of note, CGP detects both inherited and Chronic myeloid leukemia with Philadelphia chromosome and/or bcr-abl rearrangement acquired mutations in the tumor. However, the report of pharmacogenetics testing is intended for making a decision on therapy, not for counseling on hereditary cancer [24]. Th e well-known example of CGP test is FondationOne® Companion diagnostic (F1CDx) (Foundation Medicine Inc) that analyses the most common genetic aberrations in a total of 324 cancer-related genes in DNA from FFPE tumor tissue. Th e test also provides information on microsatellite instability (MSI) and tumor mutation burden (TMB), both relevant for the introduction of immunotherapy. To support clinical decision making, a report of F1CDx testing includes the genetic profi le of tumors in association with approved therapies or options for clinical trials [25].
It should be emphasized that CGP improved our understanding of cancer as a complex disease of the genome and opened a possibility to classify (and treat) tumors ac-cording to biomarkers (mutations) across tissue/organ types.

Oncology therapy in Serbia that depends on genomic testing
Based on the list of medicines "B" and "C" prescribed by the Republic Health Insurance Fund of Serbia [26], prescribing of the particular drugs depends on genetic testing (Table 3). In Serbia, genetic testing is necessary for 13 drugs during the therapy in the patients with colorectal cancer, breast cancer, non-small cell lung cancer, and skin melanoma, chronic myeloid leukemia, as well as serous epithelial carcinoma of the ovary, fallopian tube, or primary peritoneal carcinoma.
routine clinical practice allowing delivery of safe and effi cient therapy.
It's obvious that cancer treatment is undergoing a fundamental change moving toward the usage of targeted therapies for subsets of patients with certain molecular characteristics, across multiple tumor types. Diagnostic methods for comprehensive genetic profi ling of tumors are essential for the successful delivery of personalized therapy. Th ere is no doubt that decreasing costs will allow the broader application of CGP in clinical practice (initial evaluation in majority/all patients, assessment of clonal evolution during the therapy and at relapse). Th e concept of precision medicine is so appealing, but the usage of therapies directed to genetic markers across tumor types needs evaluation through well-designed research projects and clinical trials.

ACKNOWLEDGMENTS
Th is work was supported in part by the Ministry of Education, Science and Technological Development of the Republic of Serbia (Grants No 175014 and 175093).