ReCAP: Oncologists' Selection of Genetic and Molecular Testing in the Evolving Landscape of Stage II Colorectal Cancer.

Parikh AR1, Keating NL1, Liu PH1, Gray SW1, Klabunde CN1, Kahn KL1, Haggstrom DA1, Syngal S1, Kim B2, Parikh AR1, Keating NL1, Liu PH1, Gray SW1,Klabunde CN1, Kahn KL1, Haggstrom DA1, Syngal S1, Kim B2.

Author information

Abstract

CONTEXT AND QUESTION ASKED:

Genetic testing can be used in the diagnosis of Lynch syndrome, formerly known as hereditary nonpolyposis colorectal cancer (CRC), the most common inherited disorder that increases the risk for CRC; however, test results related to Lynch syndrome screening may also be used for predictive and prognostic purposes in patients with stage II CRC. Although national guidelines recommend the use of several genetic and molecular tests for patients with CRC, little is known about how guidelines, particularly the complex testing recommendations for Lynch syndrome, are translated into clinical practice. In this study, we asked: how does the family history of patients with stage II CRC influence medical oncologists' selection of genetic and molecular testing, both related and unrelated to Lynch syndrome?

SUMMARY ANSWER:

We found that oncologists' self-reported ordering of Lynch syndrome-related tests was strongly associated with the strength of CRC family history, but even so, not all oncologists would order germline testing for mismatch repair (MMR) genes, much less screen for Lynch syndrome by ordering microsatellite instability and/or immunohistochemistry for MMR proteins, in a patient scenario with the strongest family history of CRC (Table 2). We also found overtesting of KRAS and Oncotype DX for stage II CRC associated with certain practice and provider characteristics, with graduates of non-US or non-Canadian medical schools and physicians compensated under fee-for-service or by productivity-based salaries being more likely to choose KRAS testing. Fee-for-service and productivity-based salaries were also associated with increased Oncotype DX testing.jop;12/3/259/T02T1T02Table 2.Percentages of Oncologists Who Reported They Would Order Genetic and Molecular Testing for a Patient Newly Diagnosed With Stage II CRC, UnadjustedTestOverall, No. (%)Clinical Scenario*No Family History, % (n = 109)Weak Family History,% (n = 103)Strong Family History, % (n = 109)PIndividual Germline143 (45)163287< .001 MSI and/or IHC for MMR proteins205 (64)535882< .001 BRAF44 (14)81024.001 Oncotype DX120 (38)373839.97 KRAS75 (24)202427.51Combinations BRAF and MSI and/or IHC for MMR proteins32 (10)4720< .001 Germline and MSI and/or IHC for MMR proteins124 (39)152677< .001 Germline and BRAF30 (10)1524< .001 Oncotype DX and MSI and/or IHC for MMR proteins87 (28)222635.11 KRAS and BRAF35 (11)7918.04ABBREVIATIONS: IHC, immunohistochemistry; MMR, mismatch repair; MSI, microsatellite instability.*Maximum number of respondents for each scenario.

METHODS:

In 2012 and 2013, we surveyed medical oncologists in the Cancer Care Outcomes Research and Surveillance Consortium (CanCORS) and evaluated their selection of microsatellite instability and/or immunohistorchemistry for MMR proteins, germline testing for MMR genes, BRAF and KRAS mutation analysis, and Oncotype DX in stage II CRC. Physicians were randomly assigned to receive one of three vignettes, varying by strength of CRC family history. We compared differences in testing by family history and provider and practice characteristics, and we used multivariate logistic regression to identify provider and practice characteristics associated with testing.

BIAS, CONFOUNDING FACTORS, DRAWBACKS:

Although we surveyed a large cohort of oncologists from diverse geographic and practice settings, there were several limitations to this study. Whereas CanCORS patients are representative of the national patient population, participants were mostly oncologists who care for patients enrolled in CanCORS and who may be slightly older than US oncologists as a whole. Furthermore, our measures of testing relied on physician self-reporting rather than direct measures of use. In addition, we did not ask oncologists to report on the sequence in which they would order the various tests, and we were unable to determine whether such respondents would have ordered simultaneous or sequential testing. Finally, our study focused on patients with stage II CRC and may not be further generalizable.

REAL-LIFE IMPLICATIONS:

The high lifetime risk of CRC and other cancers among affected individuals and family members and low detection rates led the Centers for Disease Control and Prevention to recommend universal Lynch syndrome screening of all patients newly diagnosed with CRC. Previous efforts to increase the identification of patients and family members with Lynch syndrome have unfortunately achieved limited success. It remains to be seen whether the recapitulation by the National Comprehensive Cancer Network of the Centers for Disease Control and Prevention recommendation to screen all incident CRC specimens for Lynch syndrome can increase diagnoses. Undertesting related to Lynch syndrome screening and overtesting involving molecular tests among surveyed oncologists highlight the need for improved implementation, targeted education, and evaluation of organizational and financial arrangements to promote the appropriate use of genetic and molecular tests.