Cancer pharmacogenomics, adoption by ... - Future Medicine

Research Article For reprint orders, please contact: [email protected]

Cancer pharmacogenomics, adoption by oncologists and patient benefit

Aim: The study aim was to understand physician experience and factors influencing the adoption of cancer pharmacogenomic (caPGx) testing by oncologists practicing in academic and nonacademic settings. Method: Anonymous paper surveys were distributed to oncologists practicing in North Carolina (USA). Results: Although 98% of oncologists see promise in utilizing PGx tests in their practice, few were comfortable with their knowledge (33%) or interpreting test results (37%). At one site, the survey was not distributed due to clinician unfamiliarity with the term ‘pharmacogenomics’. Compared with oncologists in academia, community oncologists were more likely to order the new Oncotype Dx™ test for colon cancer (33% vs 0; p = 0.0071), more likely to indicate future use of caPGx tests (94 vs 75%; p = 0.012) and less likely to have never ordered a caPGx test (2 vs 35%; p < 0.001). Nearly every oncologist was interested in additional PGx education. Conclusion: A critical need exists to disseminate accurate and updated caPGx information to oncologists practicing in both academic and nonacademic settings. Keywords: cancer pharmacogenomics • diffusion theory • personalized medicine • pharmacogenetics

Pharmacogenomic (PGx) testing is an innovative approach to predict drug safety and effectiveness, select drug dose and exclude undue therapy [1] . In oncology, PGx utilizes the genomic information of the patient’s tumor, as well as the genomic information the patient has inherited from their parents and from generation to generation to predict response to anticancer therapy. In this way, cancer PGx testing offers an approach to cancer treatment where individualized, genomic-guided decision-making, rather than trial and error, provides the potential to select the treatment or dose that may be most effective, with the least side effects at the time a patient is diagnosed with cancer. Cancer pharmacogenomics is the low-hanging fruit for implementing genomic medicine into clinical practice and has significantly changed the face of cancer care and pharmacological management of numerous types of cancer [2] . Within the last 5 years, druggable targets in lung and colo

10.2217/PME.14.1 © 2014 Future Medicine Ltd

rectal cancer and melanoma have revolutionized the management of these diseases, providing dramatically improved outcomes for many patients [3–5] . Although clinical application of cancer pharmacogenomics currently focuses on tumor characteristics to predict response to therapy, the patient’s germline characteristics – variations that the patient has inherited – promises to provide an additional powerful resource to optimize cancer patient care [6] . For cancer patients to benefit from these tests, however, their oncologist must first adopt the tests, be able to interpret results and find them useful in managing the care of their patient. From past experience with diffusion and uptake of innovative health technology, some clinicians will adopt these tests early, others later and some not at all [7–10] . According to diffusion theory, the decision process is influenced by a variety of factors, including clinical knowledge, familiarity and complexity of the test, attitudes about relative advan-

Pers. Med. (2014) 11(2), 143–153

Lynn G Dressler*,1, Allison M Deal2, Jai Patel3, Janell Markey2, Marcia Van Riper4 & Howard L McLeod5 Mission Health, Personalized Medicine Program, Fullerton Genetics Center, Asheville, NC 28803, USA 2 University of North Carolina, School of Medicine, Lineberger Comprehensive Cancer Center, Chapel Hill, NC 27599, USA 3 Carolinas Medical Center, Levine Cancer Institute, Carolinas HealthCare System, Charlotte, NC 28203, USA 4 University of North Carolina, School of Nursing, Chapel Hill, NC 27599, USA 5 Moffit Cancer Center, 12902 Magnolia Drive, Tampa, FL 33612, USA *Author for correspondence: Tel.: +1 828 213 0025 Fax: +1 828 213 0039 lynn.dressler@ msj.org 1

part of

ISSN 1741-0541

143

Research Article Dressler, Deal, Patel, Markey, Van Riper & McLeod tages of the test information, peer use, and channels of communication [7,11] . A previous survey of more than 10,000 physicians from a variety of disciplines and specialties reported that early adopters of PGx tests were more likely to be in an urban setting, at an intermediate stage in their careers (15–29 years in practice), and in those practicing oncology or a surgical specialty [12] . Since approximately 85% of US cancer patients are diagnosed and treated in the nonacademic setting, analyzing oncologists’ perspectives and use of PGx testing in both the academic and community setting is essential [13] . In addition, exploring these issues with oncologists practicing in a largely rural state, which includes several major academic institutions, can inform a state-wide effort to optimize cancer patient management. The purpose of this paper is to describe the results of an empirical survey study conducted with oncologists practicing in the academic and nonacademic settings in North Carolina (USA) to understand and analyze the factors that influence oncologists’ adoption of cancer PGx testing to manage their cancer patients. The study is a first step in assessing whether a need exists to promote efficient dissemination of accurate cancer PGx information. Although the study was conducted with oncologists in North Carolina, findings may be applicable to other regions across the USA, especially regions with similar rurality, disparate access to cancer care and/or the presence of cancer health disparities (in particular among minority and underserved populations). This paper will also present suggestions to optimize the dissemination of cancer PGx information. To our knowledge, no other study has focused on perspectives of oncologists in the academic and nonacademic setting practicing in a largely rural state. Materials & methods Population

Our population consisted of medical, surgical or hematologic oncologists practicing in North Carolina in academic and nonacademic settings (community hospital based or community private practice with hospital affiliations). North Carolina is a southeastern state of 100 counties, with nearly 10 million residents residing in four major areas (mostly rural areas in western and eastern North Carolina; mostly urban areas in central North Carolina, including the Piedmont and Triangle regions) [41] . In 2010, the overall incidence and death (per 100,000 residents) from cancer in North Carolina was 477 and 178, respectively, with rates higher in minority populations [42,43] . Recruitment approaches

A multipronged recruitment approach was developed to optimize survey response. The principal investiga-

144

Pers. Med. (2014) 11(2)

tor (LG Dressler) received approval to briefly present the study, distribute and collect surveys at state-wide oncology conferences, including oncology-oriented Area Health Education Center (AHEC) meetings, and at faculty/staff meetings of cancer divisions in academic settings. The paper survey was also distributed in a routine monthly mailing from the Southeast Community Cancer Oncology Program (SECCOP) [44] with a self addressed stamped envelope for return. Oncologists were asked to fill out the survey only once and return to the principal investigator. There were no incentives for oncologists to participate; participation was completely voluntary and anonymous. There was no indication that these same individuals responded to other similar surveys (i.e., no comments were made; no show of hands when the general questions was asked about participation in other similar surveys); however, we did not specifically ask this question on the survey. Survey instruments

An anonymous paper survey, consisting of 23 questions, was adapted and revised from a measure developed by Van Riper and colleagues [14,15] . Following pretesting with a variety of healthcare providers, including oncologists outside North Carolina, revisions were made to wording of questions for clarification and the paper survey was distributed and collected in person (at meetings) or through the mail. The average time for completing the survey was 5 min (range: 3–8 min). The survey consisted of four sections: 1. General perspectives and knowledge about cancer pharmaco genomics; 2. Factors influencing use of cancer pharmacogenetic tests; 3. Desire and format for additional education; and 4. Demographic information (including type of practice: academic vs nonacademic; years in practice; gender; county of practice and type of oncologist [medical, surgical, heme-onc, other]). The survey was distributed from April 2010 to August 2010. Analysis & statistical methods

Comparisons between groups (gender, specialty: medical oncologist vs others, academic vs nonacademic; years in practice: ≤10 years vs >10 years) were made using Fisher’s exact test. For analysis, items that were reported on a Likert scale (‘strongly agree’, ‘agree’, ‘neutral’, ‘disagree’, ‘strongly disagree’; coded 1–5) were often collapsed into two categories (agree/strongly agree vs others). All analyses were conducted using SAS statistical software version 9.3 (NC, USA). Analyses were restricted to only oncologists practicing in North Carolina. Not all respondents answered all questions, an observation common in all survey studies. The survey did not ask about frequency of use for each available cancer pharmacogenomic test available

future science group

Cancer pharmacogenomics, adoption by oncologists & patient benefit

at the time; instead, the survey specifically asked about the frequency of using tests in breast cancer (HER2, CYP2D6, Oncotype Dx™ for breast cancer [Genomic Health, CA, USA]) and a new test in colorectal cancer (Oncotype Dx for colorectal cancer). Those individuals who indicated that they did not treat breast or colorectal cancer were not included in the analysis for the relevant questions (see survey: see Supplementary Material online at www.future-medicine.com/doi/full/10.2217/ pme.14.1). We also asked which individuals had never used a PGx test (see survey document). Unfortunately, not every responder answered every survey question; therefore, the analysis is based on the subset of individuals who responded to the question. Since our cohort was small and consisted of more than 70% medical oncologists, subgroup analysis was not possible. The intent of this survey was to inform a larger survey that we hope to distribute to and validate in a larger cohort of oncologists (e.g., those participating in NCI Cooperative Group trials), which will provide sufficient statistical power to conduct more informative subgroup analysis. This study was reviewed and approved by the University of North Carolina Institutional Review Board (IRB) at Chapel Hill. Results Demographics & survey response rates

A total of 94 surveys were returned and evaluable. Since this was a convenience sample and an anonymous survey, a response rate could not be calculated. However, based on the number of oncologists licensed to practice in North Carolina at the time of the survey (300 according to the North Carolina Medical Board), we estimate that approximately 30% of oncologists responded to the survey. The majority of respondents self-identified as medical oncologists (72%); followed by medical oncologists/hematologists (20%), surgical oncologists (4%) and hematologists (3%). A total of 59% of respondents practiced in nonacademic institutions (i.e., community-based hospital practice or private practice) and 41% practiced at an academic institution. The majority of respondents were male (73%). A total of 57% of all respondents indicated having been in practice for more than 10 years. Rates of response varied by recruitment method: highest when surveys were distributed and collected in person either at state medical conferences (98%) or faculty meetings (90%) and lowest for mailings (45%). At a rural AHEC site, the survey was not distributed due to lack of familiarity with the term ‘pharmacogenomics’. What do oncologists consider to be a PGx test?

When asked whether the respondent considered PGxs tests to include both somatic and germline tests or


Research Article

only germline tests, 90% agreed that PGx tests include both germline and somatic tests. Respondents were asked to select from a list of 14 tests, which they considered to be a PGx test. No definition of PGx or PGx testing was given to respondents. Figure 1 illustrates the responses for each test. Test methodology – for example, IHC versus FISH – influenced this consideration. Respondents were more likely to select the HER2 marker when it was a FISH test (66%; HER2 FISH) compared with when HER2 was measured by immunohistochemistry (IHC; 43%). Similarly, although well established as predicting response to antihormonal therapy, only 40% of respondents considered estrogen/progesterone receptors by IHC to be a PGx test. For tumor array testing, Oncotype Dx was more commonly chosen to be a PGx test compared with Mammoprint™ (Agendia Inc., CA, USA; 67 vs 36%), likely due to the marketing strategies of the two test manufacturers (Oncotype Dx is highly marketed in the USA; Mammaprint is marketed more in Europe). The top two responses for PGx tests were CYP2D6 (74%) and KRAS (76%) – two markers that, at the time of the survey, were frequently discussed in the literature [16,17] . Of concern, however infrequent, was that some respondents selected complete blood count (CBC; 3%) and mammogram (4%) as being considered a PGx test. Perspectives on pharmacogenomic testing

The majority of respondents strongly agreed (SA) or agreed (A) that “Pharmacogenomics holds great promise in guiding treatment decisions in cancer” (98%); however, only 33% reported being “comfortable with their knowledge of PGx in relation to cancer PGx tests” and 36% reported being “comfortable interpreting the results of cancer PGx tests.” Figure 2 illustrates the relative Likert scale responses for each of these questions. Less than one fourth of respondents (24%) reported being comfortable with BOTH their knowledge of PGx and interpreting the results of cancer PGx tests. Regardless of comfort level, overall 87% of respondents indicated they expect to use more PGx tests in the next 5 years. What influences the decision to utilize PGx tests in practice?

Overall, the top three factors influencing respondents’ decision to utilize PGx tests in their practice was “prospective clinical trials confirming association with outcome” (81%), “evidence-based studies demonstrating safety and efficacy of the test” (78%) and “endorsement of the PGx test by American Society of Clinical Oncology (ASCO) professional guidelines” (61%). The “use of test(s) by colleagues” (10%) or having

www.futuremedicine.com

145

Research Article Dressler, Deal, Patel, Markey, Van Riper & McLeod

Figure 1. Responses to “Which of the following do you consider to be a PGx test?”. Respondents were asked to select which of the tests listed were considered to be pharmacogenomic tests. ER/PR: Estrogen receptor/progesterone receptor; IHC: Immunohistochemistry; HER2: Human epidermal growth factor receptor 2; EGFR: EGF receptor; CBC: Complete blood count; TS/TP/DPD: Thymidylate synthase/thymidine phosphorylase/dihydropyrimidine dehydrogenase.

“a patient request the test” (4%) were least likely to influence the decision. Fifteen respondents indicated never having ordered a PGx test before. All 15 cited the reason as “there are no tests currently available that are helpful to me in my practice.” Interestingly, none cited a “lack of convincing evidence of clinical utility” as a reason for nonuse. All 15 nonusers, however, still indicated that “PGx holds great promise” and most (12/15) were interested in learning more about cancer PGx tests. Use of cancer PGx testing

Overall, 82% of respondents treated breast cancer patients and 78% treated colorectal cancer patients. Of those treating breast patients, 100% indicated using HER2 testing to predict response to HERCEPTIN™; 97% indicated use of the Breast Oncotype Dx test for chemotherapy treatment decisions. CYP2D6 testing was used by 32% of those treating breast cancers. Only 26% of respondents who treated colorectal cancer patients utilized Oncotype Dx for colon cancer. The most important factor influencing use of Oncotype Dx (breast or colon) to manage patients was the scientific literature, selected by 91% of respondents. Of note, 26% of respondents also indicated that US FDA approval of the test influenced their decision to use the test. However, the Oncotype Dx test is not FDA approved,

146

Pers. Med. (2014) 11(2)

nor does it need to be as a laboratory-developed test (LDT) [18] . Responses stratified by demographic characteristics:

When responses were stratified by demographic characteristics, we observed no differences according to gender or specific oncology discipline (medical oncology vs surgical oncology/hematology/hematology and oncology). When responses were stratified by type of practice (academic vs nonacademic) and years in practice (≤10 vs >10 years), however, significant differences in perspectives and use of PGx tests were observed (note: “Years in practice” and “type of oncologist” was similarly distributed in academic and nonacademic groups). Nonacademic vs academic oncologists

Although the majority of respondents indicated that in the next 5 years they “expect to use more PGx tests in their oncology practice”, oncologists practicing in the nonacademic setting were more likely to agree or strongly agree with this statement compared with oncologists in the academic setting (94 vs 75%, respectively; p = 0.012) (Table 1) . Of the 15 respondents indicating they had never used a cancer PGx test in their practice, all but one was from academia (p < 0.001).



Utilization of a newly introduced test, Oncotype Dx for colon cancer, also varied between academic and nonacademic oncologists. Of the oncologists who treat colon cancer, only nonacademic oncologists reported using the newly marketed (2010) Oncotype Dx test for colon cancer (33% nonacademic vs 0% academic; p = 0.0071). For the Oncotype Dx test for breast cancer, a test commercially available since 2004 and endorsed by ASCO and National Comprehensive Cancer Network (NCCN), no differences were observed between academic and nonacademic oncologists (Table 1) . Oncologists were also asked to respond to the statement: “I would only use PGx testing in my practice if I saw the following” (Figure 3) . According to responses, “prospective clinical trials” and “evidence-based studies” were the strongest factors for use by both groups. Compared with academic oncologists, nonacademic oncologists were more likely to consider factors such as “endorsement by professional groups”, “reimbursement by third-party payors”, and “use of test by colleagues” important; however, these differences were not statistically significant. Comparison of responses by years in practice

Respondents with ≤10 years of practice reported being comfortable with their knowledge of PGx more fre-

quently than those with >10 years of practice (48 vs 21%; p = 0.01). However, those with >10 years of practice were significantly more likely to identify CYP2D6 as a PGx test and to use the marker to predict response to tamoxifen compared with those with 10 years of practice (EGFR: 75 vs 49%, p = 0.017; KRAS: 88 vs 67%, p = 0.027; TS, TP, DPD mRNA expression: 78 vs 57%, p = 0.047, respectively). Perspectives on PGx education

Regardless of practice site or number of years in practice, the majority of respondents (92%) were interested in learning more about cancer PGx testing. When asked about the preferred method of learning, nonacademic oncologists were more likely to select a 1-day local symposium with CME credits (71 vs 49%; p = 0.06), whereas academic oncologists were more likely to select a symposium at a national meeting with CME credit (52 vs 29%; p = 0.06). Internet-based learning for CME credits was selected by only 32% of respondents (no difference between practice types). Only 5% of respondents indicated they did not want to learn more.

58

60

Research Article

PGx holds great promise Comfortable with knowledge Comfortable interpreting results

Percentage

50

40

40 33.70

31.5 30

33

30 27.2

27

20

10

6

5.4 2

0 SA

A

4 0

N Survey response (Likert scale)

0 D

2.2

SD

Figure 2. Oncologists’ perspectives about pharmacogenomic testing. Oncologists were asked about their perspective regarding the promise of pharmacogenomic tests to improve management of their cancer patients and the level of comfort with their knowledge about pharmacogenomics and interpreting results from cancer pharmacogenomic tests. The graph illustrates the frequency of Likert scale responses. SA: Strongly agree; A: Agree; N: Neutral; D: Disagree; SD: Strongly disagree.



147

Research Article Dressler, Deal, Patel, Markey, Van Riper & McLeod

Table 1. Perspectives and current use of cancer pharmacogenomics testing: academic versus nonacademic oncologists. Statement

Both, % (n†)

Academic, % (n†)

Nonacademic, % (n†) p-value

PGx holds great promise

98 (88/90)

97 (36/37)

98 (52/53)

NS

Use of PGx test in the next 5 years

87 (77/89)

75 (27/36)

94 (50/53)

p = 0.012

Comfort interpreting result

36 (32/88)

40 (14/35)

34 (18/53)

NS

Use of Oncotype Dx™ Breast‡

97 (69/71)

95 (18/19)

98 (51/52)

NS

Use of Oncotype Dx™ Colon

25 (17/68)

0 (0/16)

33 (17/52)

p = 0.007

Never used PGx test

16 (14/90)

35 (13/37)

2 (1/53)

p < 0.001

Desire for caPGx education

94 (85/90)

95 (35/37)

94 (50/53)

NS

‡

Note: 98% of nonacademics treated breast cancer vs 58% of academics (p < 0.0001); and 98% of nonacademics treated colon cancer vs 49% of academics (p < 0.0001). † Strongly agree/agree response. For this analysis, Likert responses were collapsed into two categories: strongly agree/agree vs neutral/disagree/strongly disagree. ‡ Based on number of respondents indicating treating breast or colon cancer patients in their practice. caPGx: Cancer pharmacogenomics; NS: Not statistically significant; PGx: Pharmacogenomics.

Discussion Results from our study continue to reinforce common themes in the adoption of PGx tests, particularly the need to prepare clinicians, regardless of specialty, with appropriate education and tools to access and utilize PGx information in their practice [10,12,15,19–36] . Our study provides an understanding of the factors influenc-

ing PGx test adoption among a specialized group of physicians–oncologists who manage cancer patients. Few studies have focused on this physician group [21,27,36] . Furthermore, our study includes a comparison of the use of PGx testing by oncologists practicing in two different settings: academic and nonacademic. Only one other study that we are aware of has offered this

90 Academic

80

Nonacademic

Percentage

70 60 50 40 30 20 10 Patient request

Use of test by colleagues

US FDA approval

Reimbursement

Endorsement by NCCN

Endorsement by ASCO

Evidence-based studies

Prospective clinical trials

0

Survey response Figure 3. Pharmacogenomics testing requirements: academic versus nonacademic oncologists†. Respondents were asked to identify the factors that would promote use of pharmacogenomic testing in their practice. Respondents could select multiple responses. The figure graphically depicts responses for each factor stratified by type of practice (academic versus nonacademic). † We observed no statistical differences between academic and nonacademic oncologist responses associated with any factor. ASCO: American Society of Clinical Oncology; NCCN: National Comprehensive Cancer Network.

148

Pers. Med. (2014) 11(2)



comparison [36] , and although that study was limited to one test (CYP2D6 for tamoxifen response), similar patterns of adoption behavior were observed, suggesting that oncologists practicing in the community may be more willing to adopt new tests compared with their academic counterparts. We also observed that, regardless of practice setting, oncologists in practice for a longer period of time may be more likely to adopt a test compared with clinicians who are earlier in their career. Finally, results from this study further document the variable dissemination of accurate information to clinicians practicing in a largely rural state. There is a need to optimize local and regional educational initiatives tailored to the needs and preferences of area oncologists. Each of these issues is discussed in more detail below. Lack of standard nomenclature

At the time of our survey, and even today, there are different views about what “PGx testing” means: some oncologists believe that the term refers to germline testing only, while others consider this term to include both somatic (tumor) and germline testing. The survey deliberately avoided defining the terms “pharmaco genomics” or “pharmacogenetics” because we wanted to learn what the term meant to our respondents and what clinical tests they considered PGx. Although most respondents in our study considered PGx to include both testing of tumor tissue as well as germline, 10% considered these tests to be only germline. It is important that communication about cancer PGx testing be precise, regardless of whether clinicians are being educated, interpreting results or communicating these results to their patients. Education, dissemination and marketing approaches rely on this precision and understanding. Another illustration of the challenge of communicating information about PGx tests is the observation that test methodology influenced the selection of a test as PGx. Methods associated with protein analysis (e.g., HER2 by IHC) were less likely to be considered PGx compared with those associated with molecular evaluation (e.g., HER2 by FISH), even though both tests predict response to the same therapy and the protein method evaluates the product of gene translation. Whether this observation is due to lack of education or individual perspective, the message that these tests can help predict response to anticancer therapy, regardless of what they are called, is important and can help inform educational approaches and development of communication tools. The group of oncologists who were not familiar with the term (and to whom we did not distribute the survey) may still have utilized cancer PGx tests (e.g., HER2 IHC, FISH; Oncotype Dx for breast cancer), but did not refer to them as “PGx tests”.


Research Article

Enthusiasm but lack of preparedness

Similar to other studies that have included oncology physicians [12,21,27,36] , we observed that oncologists see promise in PGx tests; however, most are not comfortable interpreting or utilizing the results for patient care. These data echo results of previous studies with other physicians and healthcare professionals (e.g., nurses and pharmacists) demonstrating that clinicians’ enthusiasm about the potential to predict a patient’s response to drug therapy are often offset by their sense of being unprepared or inadequately informed about the availability and appropriate utilization of PGx testing [10,12,15,19–36] . Similar to other studies, the most common factors influencing oncologists to utilize a cancer PGx test were prospective clinical trials and other evidencebased clinical information demonstrating safety and efficacy of the test result [12,15,22,36] . In contrast to our study, however, Stanek and colleagues found that “FDA approval” and “use of tests by colleagues” were major factors influencing adoption by a variety of physicians (including some oncologists) [12] . In our study we found that FDA approval did not seem to be a major promoter of use. But we did observe that some oncologists’ misunderstood which tests require FDA approval (e.g., identifying “FDA approval” as a factor influencing their use of the Oncotype Dx test). Nonacademic oncologists as potential early adopters

In our study we observed that nonacademic oncologists may be more willing to adopt cancer PGx tests early compared with those in academia. Nonacademic oncologists reported a higher likelihood of using PGx tests in general in the next 5 years and were the only oncologists utilizing the new (at the time) Oncotype Dx test for colorectal cancer. This early adoption behavior was observed in a previous study evaluating the use of CYP2D6 to predict response to tamoxifen therapy [36] . Although overall frequency of use was low, twice as many nonacademic oncologists reported using the test compared with academic oncologists who responded to the survey [36] . At the time of our study, the Oncotype Dx test for colon cancer was on the market only a short time (months), and although today CYP2D6 testing for tamoxifen response is controversial [16] , the similar observations from two different studies raise intriguing questions related to the factors that may be promoting different adoption patterns among academic and nonacademic oncologists. Addressing these questions would provide an important understanding of how cultural, political and social factors, in addition to channels of communication, may influence oncologists’ adoption behavior [7,10] .


149

Research Article Dressler, Deal, Patel, Markey, Van Riper & McLeod Years in practice paradox

Another component of our study evaluated how years in practice might influence response to questions relating to knowledge and use of cancer PGx tests. Although oncologists in practice ≤10 years indicated that they were more comfortable with their knowledge and interpretation of PGx test results, those in practice more than 10 years more frequently identified CYP2D6 as a PGx test and reported a higher frequency of using CYP2D6 for predicting tamoxifen response (at the time of our survey, there was considerable support for CYP2D6 use). This difference between experience and PGx use is similar to the observation of Stanek and colleagues surveying more than 10,000 physicians, where “intermediate or longer practicing” physicians were more likely to adopt some PGx tests compared with more recent graduates [12] . The relationship between experience (years in practice) and the timing of when new information is learned and when it is applied in practice, is another intriguing nuance of diffusion theory to explore in a future study. Dissemination of accurate PGx information

We observed a need to optimize dissemination of accurate PGx information. Not being able to distribute a survey due to lack of familiarity with terminology is a reflection of inefficient dissemination of information. The rurality of a region should not be a barrier to access this information. In addition, every oncologist should understand that neither a mammogram nor a CBC is a PGx test – fundamentally, neither test predicts response to anticancer therapy. And for oncologists who consider FDA approval of a test as a condition for ordering that test, it is important to understand regulatory oversight of genetic testing and know that the test they are using (e.g., Oncotype Dx) does not have nor needs to have FDA approval for clinical use. Inefficient dissemination of accurate information into clinical practice can prevent some patients from having access to these tests, as well as expose cancer patients to harmful or ineffective therapies. It can also obstruct one of the goals of cancer PGx to deliver more cost-effective patient care. Conclusion This study, combined with others in the literature, throws into stark relief the continuing need for education. Nearly all respondents wanted additional education, even those who did not think they would use PGx in the next 5 years. PGx testing will not be integrated into daily clinical practice if clinicians are not equipped with relevant content knowledge, see value in the clinical utility that testing provides for managing their patients, and understand when and how to apply

150

Pers. Med. (2014) 11(2)

the results in clinical practice [12] . The need for this continuing education will only grow as genomic information becomes more integrated into the practice of medicine. However, to maximize opportunity for equitable access to these tests and to benefit patients as well as clinicians, optimal communication channels, appropriate resources and high-quality education programs must exist at local, regional, state and national levels. At the national and international levels, a number of efforts have already been initiated or are planned to promote education about PGx testing. PGx education has been incorporated into the core curriculum of many medical, nursing and pharmacy programs [37–40] . The provision of genetic and genomic education programs for practicing clinicians is also growing and varied: the NIH supports a variety of programs to train and educate clinician scientists in pharmacogenomics [45] ; professional organizations, including ASCO [46] as well as individual institutions [47] , also provide educational opportunities. The National Coalition for Health Care Providers Education in Genetics (NCHPEG) [48] developed a curriculum for training primary care and other clinicians in genetics and PGx. However numerous these efforts may be, most occur independent of each other and many are not accessible to every oncologist. Future perspective: next steps The following next steps are suggested to expedite and coordinate the process of more efficient oncology education: Integrate efforts to develop high-quality, consistent training & education programs

There is a need for an overarching group to champion the process of tracking, coordinating and consolidating these educational initiatives, approaches and methods to develop consistent and effective high-quality education programs. A first step could include the development of a website for curriculum sharing, sponsored by the National Cancer Institute or National Institutes of General Medical Sciences in partnership with the American Society of Clinical Oncology, the National Comprehensive Cancer Network, and the College of American Pathology. Development of a coordinated infrastructure, especially at the local level, to disseminate accurate & updated PGx information

Curation and consolidation of educational programs by itself is insufficient to promote efficient dissemination of this information at the local level. Grant support from the NIH, AHRQ and/or local public and private foundations is needed to incentivize the develop ment of a local infrastructure, building on



Research Article

existing networks, including regional AHECs, Community Cancer Oncology Programs (CCOP)s, and the NCI-designated Comprehensive Cancer Centers. This ‘local hub’ for information dissemination could also help tailor innovative education approaches to local oncologists’ preferences for learning.

effectively disseminated. The utilization of this information as ‘best practices’ can then become incorporated into quality measures for cancer patient management, as assessed by oncology performance indicators (e.g., ASCO’s Quality Oncology Practice Initiative), further incentivizing their use and dissemination.

Facilitate use of PGx information

Financial & competing interests disclosure

To facilitate use of PGx information, access to a central hub of vetted information, at the local or national level, could also provide ongoing services and clinical decision-support tools. These services could include a consultation service to aid in the interpretation, utilization and communication of PGx test results to optimize cancer patient management. The central hub could also assist in the dissemination of evidence-based clinical decision-support tools, endorsed by professional organizations and interoperable with electronic medical record systems or accessible through the oncologists’ personal e-device.

The authors have 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.

Evaluation & quality metrics

Educational programs will require an evaluation component to ensure that accurate information is being

Ethical conduct of research The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.

Executive summary Perspectives on pharmacogenomic testing • Although oncologists in North Carolina (USA) are very enthusiastic about the use of pharmacogenetic/omic (PGx) tests to guide management of their cancer patients, most are not comfortable with their knowledge about PGx nor their ability to interpret results from these tests.

Nonacademic vs academic oncologists • Nonacademic oncologists (hospital-affiliated or private practice community oncologists) may be more likely to adopt new PGx testing earlier than oncologists in academia.

Perspectives on pharmacogenomic education • Oncologists want additional education about PGx, tailored to their preferences for learning.

Dissemination of accurate pharmacogenomic information: • A critical need exists to efficiently disseminate accurate PGx information to all oncologists, regardless of the type of practice (e.g., academic vs nonacademic) or practice location (rural vs urban).

Next steps • Potential next steps to facilitate this dissemination include: –– An overarching effort to track, coordinate and integrate the disparate educational initiatives to develop high quality education and training programs, accessible to all oncologists; –– Development of a coordinated infrastructure at the local level to disseminate accurate and updated PGx information and facilitate PGx test use, including the provision of clinical decision-support tools to aid in the interpretation, application and communication of pharmacogenomic test results; –– Using/leveraging existing oncology infrastructures and quality indicators of performance to optimize channels of communication and support appropriate test use. more fully integrate these tests into the management of cancer patients.

References Papers of special note have been highlighted as: • of interest 1

McLeod HL. Cancer pharmacogenomics: early promise, but concerted effort needed. Science 339, 1563–1566 (2013).

2

Green ED, Guyer MS; National Human Genome Research I. Charting a course for genomic medicine from base pairs to bedside. Nature 470, 204–213 (2011).

•

Review of somatic and germline cancer pharmacogenomic opportunities, challenges and what needs to be done to

•

Describes the National Human Genome Research I vision for genomic medicine to optimize care, including managing cancer patients.



151

Research Article Dressler, Deal, Patel, Markey, Van Riper & McLeod 3

Chapman PB, Hauschild A, Robert C et al. Improved survival with vemurafenib in melanoma with BRAF V600E mutation. N. Engl. J. Med. 364, 2507–2516 (2011).

4

Van Cutsem E, Kohne CH, Lang I et al. Cetuximab plus irinotecan, fluorouracil, and leucovorin as first-line treatment for metastatic colorectal cancer: updated analysis of overall survival according to tumor KRAS and BRAF mutation status. J. Clin. Oncol. 29, 2011–2011 (2019).

5

19

Zachry WM 3rd, Armstrong EP. Health care professionals’ perceptions of the role of pharmacogenomic data. J. Manag. Care Pharm. 8, 278–284 (2002).

20

Gardiner SJ, Begg EJ. Pharmacogenetic testing for drug metabolizing enzymes: is it happening in practice? Pharmacogenet. Genomics 15, 365–369 (2005).

21

Woelderink A, Ibarreta D, Hopkins MM et al. The current clinical practice of pharmacogenetic testing in Europe: TPMT and HER2 as case studies. Pharmacogenomics J. 6, 3–7 (2006).

22

Fargher EA, Eddy C, Newman W et al. Patients’ and healthcare professionals’ views on pharmacogenetic testing and its future delivery in the NHS. Pharmacogenomics 8, 1511–1519 (2007).

23

Mrazek M, Koenig B, Skime M et al. Assessing attitudes about genetic testing as a component of continuing medical education. Acad. Psychiatry 31, 447–451 (2007).

6

Deenen MJ, Cats A, Beijnen JH et al. Part 1: background, methodology, and clinical adoption of pharmacogenetics. Oncologist 16, 811–819 (2011).

•

Part 1 of a four-part series targeting the implementation of pharmacogenomics into cancer clinical practice.

7

Rogers EM. Diffusion of Innovations (4th Edition). The Free Press, NY, USA, 10–35 (1995).

24

8

Gurwitz D, Zika E, Hopkins MM et al. Pharmacogenetics in Europe: barriers and opportunities. Public Health Genomics 12, 134–141 (2009).

Trinidad SB, Fryer-Edwards K, Crest A et al. Educational needs in genetic medicine: primary care perspectives. Community Genet. 11, 160–165 (2008).

25

Feero WG, Guttmacher AE, Collins FS. Genomic medicine – an updated primer. N. Engl. J. Med. 362, 2001–2010 (2011).

Kadafour M, Haugh R, Posin M et al. Survey on warfarin pharmacogenetic testing among anticoagulation providers. Pharmacogenomics 10, 1853–2009 (1860).

26

Hoop JG, Lapid MI, Paulson RM et al. Clinical and ethical considerations in pharmacogenetic testing: views of physicians in 3 ‘early adopting’ departments of psychiatry. J. Clin. Psychiatry 71, 745–753 (2010).

27

Bonter K, Desjardins C, Currier N et al. Personalised medicine in Canada: a survey of adoption and practice in oncology, cardiology and family medicine. BMJ Open 1(1), e000110 (2011).

28

Haga SB, Tindall G, O’Daniel JM. Professional perspectives about pharmacogenetic testing and managing ancillary findings. Genet. Test. Mol. Biomarkers 16(1), 21–24 (2011).

29

McCullough KB, Formea CM, Berg KD et al. Assessment of the pharmacogenomics educational needs of pharmacists. Am. J. Pharm. Educ. 75, 51 (2011).

30

Benzeroual K, Shah B, Shinde S. Pharmacogenomics: assessing educational exposure, confidence in knowledge and training elements of pharmacists. Pers. Med. 9, 387–393 (2012).

9

10

Hedgecoe AM. Context, ethics and pharmacogenetics. Stud. Hist. Philos. Biol. Biomed. Sci. 37, 566–582 (2006).

11

Sanson-Fisher RW. Diffusion of innovation theory for clinical change. Med. J. Australia 180, S55–S56 (2004).

12

Stanek EJ, Sanders CL, Taber KA et al. Adoption of pharmacogenomic testing by US physicians: results of a nationwide survey. Clin. Pharmacol. Ther. 91, 450–458 (2012).

•

Seminal study by MEDCO of physician adoption of pharmacogenomic testing and the need for physician education.

13

Zaren HA, Nair S, Go RS et al. Early-phase clinical trials in the community: results from the national cancer institute community cancer centers program early-phase working group baseline assessment. J. Oncol. Pract. 9, e55–e61 (2013).

14

Roederer MW, Van Riper M, Valgus JM et al. Knowledge, attitudes and education of pharmacists regarding pharmacogenetic testing. Pers. Med. 9, 19–27 (2012).

31

Haga S, Burke W, Ginsburg G et al. Primary care physicians’ knowledge of and experience with pharmacogenetic testing. Clin. Genetics 82(4), 388–394 (2012).

15

Dodson C, Van Riper M. Analysis of clinicians’ attitudes towards pharmacogenomics. Pers. Med. 8, 533–540 (2011).

32

16

Hertz DL, McLeod HL, Irvin WJ Jr. Tamoxifen and CYP2D6: a contradiction of data. Oncologist 17, 620–630 (2012).

de Denus S, Letarte N, Hurlimann T et al. An evaluation of pharmacists’ expectations towards pharmacogenomics. Pharmacogenomics 14, 165–175 (2013).

33

17

Bardelli A, Siena S. Molecular mechanisms of resistance to cetuximab and panitumumab in colorectal cancer. J. Clin. Oncol. 28, 1254–1261 (2010).

Formea CM, Nicholson WT, McCullough KB et al. Development and evaluation of a pharmacogenomics educational program for pharmacists. Am. J. Pharm. Educ. 77, 10 (2013).

34

Selkirk CG, Weissman SM, Anderson A et al. Physicians’ preparedness for integration of genomic and pharmacogenetic testing into practice within a major healthcare system. Genet. Test. Mol. Biomarkers 17, 219–225 (2013).

18

152

Zhou C, Wu YL, Chen G et al. Erlotinib versus chemotherapy as first-line treatment for patients with advanced EGFR mutation-positive non-small-cell lung cancer (OPTIMAL, CTONG-0802): a multicentre, openlabel, randomised, Phase 3 study. Lancet Oncol. 12, 735–742 (2011).

use of tumor markers in breast cancer. J. Clin. Oncol. 25, 5287–5312 (2007).

Harris L, Fritsche H, Mennel R et al. American Society of Clinical Oncology 2007 update of recommendations for the

Pers. Med. (2014) 11(2)



Research Article

•

Survey study highlighting how physicians spanning multiple disciplines are underprepared to incorporate genomics into medical practice.

42

State Center for Health Statistics. North Carolina Cancer Incidence Rates (2010). www.schs.state.nc.us/schs/CCR/reports.html

35

Rogausch A, Prause D, Schallenberg A et al. Patients’ and physicians’ perspectives on pharmacogenetic testing. Pharmacogenomics 7, 49–59 (2006).

43

State Center for Health Statistics. North Carolina Cancer Mortality by Race (2011). www.schs.state.nc.us/schs/CCR/mort2011r.pdf

36

Peppercorn J, Hamilton E, Marcom PK et al. Pharmacogenetic testing in the face of unclear clinical efficacy: lessons from cytochrome P450 2D6 for tamoxifen. Cancer 119(20), 3703–3709 (2013).

44

Southeast Cancer Control Consortium. www.southeastcancercontrol.org

45

NIH Career Awards. K23 Translational Scholar Career Awards in Pharmacogenomics and Personalized Medicine. www.nigms.nih.gov/Research/FeaturedPrograms/PGRN/ CTSAscholars

46

American Society for Clinical Oncology (ASCO Education Programs). Building Bridges to Conquer Cancer. May (2013). http://am.asco.org/sites/chicago2013.asco.org/files/Ed%20 Book%202013%20lores%20cov%20to%20cov%20 TOC%20order%20no%20index.pdf

47

Vanderbilt-Ingram Cancer Center. My Cancer Genome. www.mycancergenome.org

48

NCHPEG. National Coalition for Health Care Provider Education in Genetics. Core Competencies for all health professionals (2007). www.nchpeg.org/index.php?option=com_content&view=art icle&id=237&Itemid=84

37

McKinnon RA, Ward MB, Sorich MJ. A critical analysis of barriers to the clinical implementation of pharmacogenomics. Ther. Clin. Risk Manag. 3, 751–759 (2007).

38

Green JS, O’Brien TJ, Chiappinelli VA et al. Pharmacogenomics instruction in US and Canadian medical schools: implications for personalized medicine. Pharmacogenomics 11, 1331–1340 (2010).

39

Murphy JE, Green JS, Adams LA et al. Pharmacogenomics in the curricula of colleges and schools of pharmacy in the United States. Am. J. Pharm. Educ. 74, 7 (2010).

40

41

Dodson CH, Lewallen LP. Nursing students’ perceived knowledge and attitude towards genetics. Nurse Education Today 31, 333–339 (2011). United States Census Bureau. State and County Quick Facts. Population Estimates. North Carolina, USA (2012). http://quickfacts.census.gov/qfd/states/37000.html



153