Leonard Saltz, MD
The joint American Society for Clinical Pathology (ASCP), College of American Pathologists (CAP), Association for Molecular Pathology (AMP), and ASCO guideline reported by Sepulveda et al, and reviewed in this issue of The ASCO Post, represents a collaboration of three pathology societies and ASCO to codify standards for use of biomarkers in colorectal cancer.1 Much of the 30-page document is focused on methodology, quality control, and standardization from a pathology perspective. I will restrict my comments to the first 7 of the 21 statements—those that directly impact medical oncology practice.
RAS Mutational Testing
Statement 1 asserts that all patients being considered for antiepidermal growth factor receptor (EGFR) therapy must receive extended RAS mutational testing. I agree. This is standard practice and has been mandated in the National Comprehensive Cancer Network® (NCCN®) Guidelines since 2014. Although some earlier data had suggested that G13D KRAS mutations might still be responsive to anti-EGFR agents, those data have not been corroborated, and any KRAS or NRAS mutation should now be regarded as a strict contraindication to anti-EGFR therapy. Not only are anti-EGFR agents (cetuximab [Erbitux] and panitumumab [Vectibix] for the purposes of this discussion) ineffective in patients whose tumors harbor RAS mutations, but they actually appear to have the ability to accelerate tumor growth in these patients.
The wording in Statement 1 that says “Colorectal cancer patients being considered for anti-EGFR agents” is less precise than I would have preferred, and I would note that only patients with metastatic disease would be appropriate for such consideration. As such, the guideline fails to comment on the inappropriateness of expending resources on RAS testing for patients with stages I, II, and III disease. Such testing provides no actionable information and therefore is not, at this time, warranted.
BRAF Mutational Testing
Statement 2a states: “BRAF V600 mutational analysis should be performed … for prognostic stratification.” Although this is certainly true of any patient going on a clinical trial, I would respectfully disagree with the appropriateness of prognostic stratification in standard practice. The distinction between “predictive” and “prognostic” warrants discussion here.
A predictive biomarker tells us whether or not an intervention is or is not likely to benefit the patient. By definition, predictive markers provide actionable information and so can—and should—guide therapeutic decisions. A prognostic marker tells us whether the patient is likely to have a favorable outcome or not, but it does not provide information on whether or not a particular intervention will be helpful toward achieving that outcome.
ADDITIONAL INFORMATION
For additional information about the guidelines on molecular biomarkers for the evaluation of colorectal cancer, visit www.asco.org/CRC-markers-guideline and www.asco.org/guidelineswiki.In short, prognostic information may give you a warm, fuzzy feeling or a cold chill down your spine, but there is nothing you can do about it; it is not actionable. Predictive biomarkers, whether they are inclusionary markers such as ERBB2 overexpression in breast or esophagogastric cancer or exclusionary markers such as RAS mutations in colorectal cancer, are highly useful and are value-added to the patient. In my opinion, prognostic biomarkers provide no real benefit to patients and constitute an added expense without added value. As such, I do not believe they should be used, or reimbursed, in standard practice.
Statement 2b states: “BRAF V600 mutational analysis should be performed in deficient mismatch repair tumors with loss of MLH1….” I agree. However, it further states: “The presence of a BRAF mutation strongly favors a sporadic pathogenesis.” I again agree but would prefer the guideline offer stronger guidance here. Barring young age and/or a compelling family history, a BRAF V600E mutation in the setting of MLH1 deficiency is virtually diagnostic of somatic MLH1 loss, thus excluding Lynch syndrome, and further genetic evaluation would therefore not be indicated.2-4
Statement 3 states: “Clinicians should order mismatch repair status testing in patients with colorectal cancer for identification of patients at high risk for Lynch syndrome and/or prognostic stratification.” I completely agree with the importance of Lynch screening. In fact, current NCCN guidelines now recommend that all colorectal cancer patients be screened for Lynch. However, as previously stated, I respectfully disagree that any expenditure of resources for prognostic stratification is appropriate in standard practice. Mismatch repair status needs to be identified in patients with stage II disease, not because it is prognostic (it is), but because it is predictive of the absence of benefit of fluoropyrimidine chemotherapy.5 In addition, mismatch repair status now needs to be determined on all colorectal cancer patients with metastatic disease, because although only 2% to 4% of them will have mismatch repair–deficient disease, the patients in that small subgroup may benefit from treatment with immune checkpoint inhibitors.6
“No patient should undergo a biopsy of a metastasis if sufficient primary tumor tissue is available.”— Leonard Saltz, MD
Tweet this quote
Statement 4 states: “There is insufficient evidence to recommend BRAF V600 mutational status as a predictive molecular biomarker for response to anti-EGFR inhibitors.” Here I respectfully disagree. Current NCCN guidelines state that “BRAF V600E mutation makes response to panitumumab or cetuximab highly unlikely.” A review of 10 randomized first-line and non–first-line trials analyzed 463 patients with BRAF-mutated colorectal cancer.7 The authors found that the addition of an anti-EGFR monoclonal antibody did not significantly improve overall survival (P = .63), progression-free survival (P = .33), or response rate (P = .25) compared with control regimens.
I believe the preponderance of available data does support the use of a BRAF V600E mutation as a compelling reason not to expose a patient to the toxicity of an anti-EGFR agent. BRAF V600E–mutant colorectal cancer does not respond to single-agent BRAF tyrosine kinase inhibitors in the way that melanomas with such mutations do.8 Recent studies have suggested more activity with the combination of BRAF inhibitors and EGFR inhibitors,9 but such approaches remain investigational at this time.
Statements 5 and 6 note that data do not support the use of PIK3CA or PTEN analysis for selection (or rejection) of therapy outside of a clinical trial. I agree at this time, although I do believe that PIK3CA mutations likely confer relative resistance to EGFR inhibitors.
Preferred Specimens for Testing
Statement 7 states: “Metastatic or recurrent colorectal cancer tissues are the preferred specimens for treatment-predictive biomarker testing and should be used if such specimens are available and adequate. In their absence, primary tumor tissue is an acceptable alternative and should be used.” They note that the strength of the evidence for this is “inadequate/insufficient.” Data have suggested a very high concordance rate between primary and metastasis for RAS and BRAF mutations.10-12 As such, no patient should undergo a biopsy of a metastasis, specifically for RAS, BRAF, or mismatch repair testing, if sufficient primary tumor tissue is available. It is worth noting that a number of different technologies for obtaining tumor mutational status through “liquid biopsies,” testing blood or urine, may make the invasiveness of a biopsy a non-issue. ■
DISCLOSURE: Dr. Saltz reported no conflicts of interest.
References
4. National Comprehensive Cancer Network: NCCN Guidelines for Genetic/Familial High-Risk Assessment: Colorectal, version 2.2016 (p 19). Available at https://www.nccn.org/professionals/physician_gls/pdf/genetics_colon.pdf. Accessed June 1, 2017.
