Extended RAS Mutation: Molecular Oncology Tumor Boards

ASCO University
Jan 14, 2015 9:31 AM

User Instructions: Welcome to the Molecular Oncology Tumor Board Series! This educational initiative is a collaboration between the American Society of Clinical Oncology (ASCO), College of American Pathologists (CAP), and Association for Molecular Pathology (AMP).

A new case will be presented each month with discussions moderated by an expert pathologist and medical oncologist. This month’s topic is led by Drs. Stanley Hamilton (MD Anderson Cancer Center) and Leonard Saltz (Memorial Sloan Kettering Cancer Center).

Users are encouraged to leave comments and post questions about the case in order to generate a wide discussion among the cancer care community. You can also receive email notifications when new comments are posted by clicking the “Email me when this thread has replies.” option located at the bottom of this page.

When posting, users are required to abide by the terms and conditions of this website.

Please click here for supplemental resources related to the case.



ASCO University
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 14, 2015 9:43 AM

Patient Case

:  62 year-old male

Medical History: 10 year history of Type II diabetes, on oral hypoglycemic, without complications.  Maternal grandmother had colon cancer at age 75, stage II; resected and died age 86. No other significant medical problems and on no other medications.

Date of Diagnosis: 14 months ago (November, 2013) he developed fatigue and was found to be mildly anemic. Colonoscopy identified a fungating mass at the hepatic flexure. Multiple biopsies were obtained which revealed moderately differentiated adenocarcinoma.  CT scan of the chest and abdomen demonstrated multiple unresectable liver and lung metastases. CEA found to be elevated to 176 ng/dl (normal < 5.0).  He underwent placement of a venous access catheter.

Co-morbidities: Uncomplicated type II diabetes mellitus.

Pathology specimen from multiple biopsies of mass at hepatic flexure:

Clinical history: Liver and lung masses, colon mass
Source of specimen:  Bx hepatic flexure mass
Gross description:  Six fragments of tan tissue received in formalin and measuring 0.5 centimeter in aggregate. Entirely embedded in block A1.
No non-neoplastic colonic mucosa included in specimen.
Molecular testing order: KRAS and BRAF mutation, microsatellite instability
Pre-analytics:  Review of histopathology; marking of slide for high-cellularity tumor; recuts of block for unstained sections for microdissection for tumor enrichment and immunohistochemistry;  DNA extraction and quality control checks
Analysis: Molecular diagnostics laboratory for KRAS exon 2 (codons 12 and 13) with FDA-cleared assays for seven most common KRAS mutations and for BRAF (V600E). Immunohistochemistry for MSH2, MLH1, MSH6, and PMS2 mismatch repair gene products

Type of Tumor:  Moderately differentiated adenocarcinoma. KRAS exon 2 testing showed a wild-type tumor, BRAF testing also wild type. Mismatch repair proteins are intact.

Relevant Markers:  CEA 176 ng/ml; alkaline phosphatase mildly elevated; bilirubin WNL.


ASCO University
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 14, 2015 9:44 AM

Discussion Questions:

  1. Based off of the information above, which cytotoxic chemotherapy combination would you choose? (ex. FOLFOX, FOLFIRI, CapeOx, Cape Iri) Why?

  2. Which biologic would you add, if any, and why?

  3. Would you consider it necessary to wait for the results of tumor genotyping (KRAS, etc.) before making your decision on biologics?


Anis Toumeh, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 14, 2015 11:57 AM

1- Randomized controlled trials comparing doublets did not show major advantage of one over the other. He has uncomplicated DM and no peripheral neuropathy. I would choose FOLFOX or CAPOX 
2- At this point, he has good PS and no contraindications for anti-angiogenic therapy. At the same time, I would like to do extended RAS testing to determine his eligibility for Anti-EGFR therapy. I would choose Bev as a biologic 
3- I would recommend extended KRAS testing.


Ryan H. Devine, DO
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 14, 2015 12:53 PM

1. FOLFOX, CapeOx or FOLFOXIRI. FOLFOX and CapeOx are similar in efficacy. In good PS patient younger than 70, could consider FOLFOXIRI with improved response rates and updated analysis when compared to FOLFOX with avastin demonstrates improved survival by 4 months, doubling in 5 yr survival rates.
2. prior to tumor genotyping would choose biologic avastin to go with cytotoxic chemotherapy. If tumor testing back and KRAS WT, would choose cetuximab or panitumumab.
3. It is not necessary to wait for tumor genotyping. I would use avastin up front, but if tumor type back and KRAS WT would use anti-EGFR therapy.


Leonard B. Saltz, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 19, 2015 1:15 PM

Course Faculty Response

The patient presented has extensive multifocal metastatic colorectal cancer, and clearly is never going to be a candidate for resection. Thus, treatment is non-curative, but has substantial potential to prolong life, and we need to consider toxicities in terms of how they might impact the quality of that life.  A number of reasonable options could be used.  An argument for fluoropyrimidine (5FU/LV or capecitabine) plus bevacizumab could be made, and likely would be favored by many European colorectal experts.  Cytotoxic doublets are more likely to be preferred by most American oncologists, myself included.  While oxaliplatin and irinotecan-based doublets have been shown numerous times to be equivalent in terms of efficacy, their side effect profiles are different.  In a patient such as this, with a 10 year history of diabetes, I would favor irinotecan over oxaliplatin, as neuropathies from diabetes and oxaliplatin have a high likelihood of exacerbating one another. Even though we are told this man does not have complications of his diabetes, I would be concerned that with a 10 year history, there is likely to be at least subclinical neuropathy, and therefore he carries a greater than average risk of having neuropathy with oxaliplatin become a clinically significant problem for him.  For this reason, I would favor FOLFIRI. Note that if one chooses to use FOLFOX, then CapeOx is an acceptable alternative.  In U.S. studies, however, Cape-Iri has not been well tolerated and has not been shown to be as effective as FOLFIRI (in marked contradistinction to studies done in Europe, where Cape-Iri appears to be a perfectly acceptable alternative). So in my practice in New York, I would choose FOLFIRI, and Cape-Iri would not be an acceptable consideration.

In the absence of a contraindication, I usually use bevacizumab with first line chemotherapy.  History of clinically significant cardiovascular disease, incomplete wound healing, or risk of impending obstruction which might require urgent surgery are the most common relative contraindications to first line bevacizumab.  My decision is not influenced by RAS mutation status, and I do not feel that it is necessary to wait for RAS genotyping results before starting first line therapy, although all patients should have this genotyping done at the first identification of metastatic disease. Note that it is not appropriate to obtain RAS genotyping for stage I, II, or III patients, since the information will have no impact on management unless the patient develops metastatic disease, in which case the information can then be obtained off of the patient’s original surgical paraffin block.  

I virtually never use cetuximab or panitumumab in first line, as I feel that the impact that the skin rash has on a patient’s quality of life is too often extremely negative, and I prefer to reserve anti EGFR therapy for later in the course of treatment, usually third line.  It should be remembered that only patients who experience a substantial skin rash benefit from anti EGFR therapy, so the thought that a patient might derive benefit without a substantial skin rash is unrealistic.  Treatments for the skin rash are only marginally effective at best.  

Note that the largest trial which addresses the question of anti VEGF versus anti EGFR, the CALGB/SWOG 80405 trial, showed no difference in efficacy outcome with cetuximab versus bevacizumab in KRAS wild type and all-RAS wild type patients. The FIRE-3 study, a trial half the size of the 80405 trial, was negative for the pre-specified primary endpoint of response rate, showing no significant difference in response between FOLFIRI-cetuximab and FOLFIRI-bevacizumab.  Thus, any other conclusions drawn from this negative trial are necessarily hypothesis-generating at best.  PFS was identical between the two arms. Overall survival was statistically significantly longer in the arm that received first line cetuximab; however the curves do not separate until around two years out, well over a year after progression on the first line regimen. It is unclear how the survival difference in this negative trial, beginning over a year after the first line treatment is completed, should be interpreted, but it certainly must be interpreted with caution.


ASCO University
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 19, 2015 1:21 PM

Patient Case Update

Treatment History/Response: After a discussion of treatment options, the patient began treatment with FOLFIRI plus bevacizumab.  A CT scan after 8 weeks of treatment showed a favorable response with approximately a 30% decrease in the size of all lesions. CEA decreased to 81 ng/dl.   He continued therapy with CT scans at 12 week intervals, with continued response, and CEA continuing a decline to 26 ng/nl. However, a CT scan approximately 10 months into therapy showed substantial growth compared with the prior film, with an increase in CEA to 95 ng/ml.  He has continued to work full time through treatment and continues to feel well with an ECOG 1 performance status.

You are now is considering switching to either FOLFOX plus continued bevacizumab, or continued irinotecan with the addition of an anti EGFR monoclonal antibody.


ASCO University
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 19, 2015 1:23 PM

Discussion Question:

Is exon 2 KRAS testing sufficient in the United States in 2015, or is additional genotype testing necessary?


Chanjuan Shi
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 20, 2015 10:20 AM

Only a small percentage of CRCs (<10% of KRAS exon 2 wild type tumors) have mutations outside of exon 2. Testing for this on every mCRC would likely be an added cost and some of the current commercial kits do not include these. However, if you are doing a multi mutation platform like NGS or single base extension assays (e.g. SNapshot platform) which has some of these recurrent mutation included, you can test for them at no extra expense or TAT. The question is whether there are sufficient data supporting that KRAS mutations in exon 3 and exon 4 confer resistance to anti-EGFR MoAb. If so, we may want to extend RAS mutation panel.


Stanley R. Hamilton, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 20, 2015 4:14 PM

Course Faculty Response

The topic of expanded/extended RAS testing has been addressed thoroughly in a current review in the Journal of Clinical Oncology:  Atreya CE, Corcoran RB, Kopitz S.  Expanded RAS: Refining the patient population. Published ahead of print on January 12, 2015.

The initial evidence for the importance of tumor RAS gene mutation as an indicator that anti-epidermal growth factor receptor antibodies should not be used was based on evaluation of the common KRAS exon 2 (codons 12 and 13) mutations by Sanger sequencing of tumor specimens from patients on clinical trials.  With additional studies and the availability of advancing technology for broader evaluation of the RAS family, it is now apparent that expanded/extended RAS mutation testing is crucial. The mutations in KRAS exons 3 and 4 and in NRAS exons 2, 3 and 4 also have biochemical effects similar to those of KRAS exon 2, providing biological plausibility. As presented in detail in the review cited above, retrospective evaluation of specimens from completed clinical trials of anti-EGFR antibodies has now provided clear evidence that the low-frequency mutations in KRAS and NRAS also prevent effectiveness of the antibodies, although the number of analyzable cases is relatively small at present.

The use of the term “wild type” is unfortunate in conveying a false sense of security about the knowledge of mutation status in the RAS family.  “No mutation identified”, or something similar, with specification of what was tested is far more satisfactory in providing information to oncologists on the completeness of RAS family coverage by the assay. The evaluation of five trials of first-line therapy for metastatic colorectal cancer has shown a prevalence in “wild-type” tumors of expanded/extended mutations in KRAS and NRAS of 15% to 27% of patients. When the thousands of patients with metastatic colorectal cancer worldwide are considered, treating about one sixth to one fourth of patients without prior identification of contraindication to anti-EGFR therapy would translate to unacceptable numbers of patients who would be exposed to  toxicities and the costs of the expensive drugs with little chance of benefit.  There are other plausible indicators, e.g. BRAF mutation, that need to be studied more thoroughly, as well as subclones and tumor heterogeneity.

The course faculty would appreciate input from readers to see how their use of expanded/extended RAS testing in this particular case and in their practice.


Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 20, 2015 6:04 PM

HI Dr. Saltz - Does the fact that all his mismatch repair genes are all intact play a role in the choice of his chemotherapy? I thought that if the tumor doesn't have mutation in the mismatch repair genes then that would make them microsatelite stable and therefore would not response to 5-FU.  Could you comment on that part? 


Chanjuan Shi
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 20, 2015 9:06 PM

Thank you for the very informative review paper, Dr. Hamilton. At Vanderbilt we use SnapShot platform to analyze mutations in KRAS exons 2, 3 and 4 and in NRAS exons 2 and 3 for all stage IV colorectal cancer. NRAS exon 4 mutations are not included.


Leonard B. Saltz, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 20, 2015 10:33 PM

Course Faculty Response

Daphne - You are correct that mismatch repair (MMR) status would not play a role in selection of therapy for a patient with metastatic disease. This patient is 62 years old, so it is appropriate to obtain MMR to screen for Lynch Syndrome. Per NCCN guidelines, all colorectal cancer patients presenting below the age of 70, or any older patients who meet Bethesda Criteria for family history, should be screened for Lynch by either immunohistochemistry for MMR or PCR for microsatellite instability.


Roberto Antonio Leon-Ferre, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 20, 2015 10:52 PM

Great discussion - thanks Dr Saltz. At Mayo, we are doing extended RAS testing.
I am curious about how much weight one should place on the history of diabetes without over diabetic neuropathy when thinking about using oxaliplatin. Is there any evidence that the presence of diabetes alone predisposes patients to develop oxaliplatin-induced peripheral neuropathy? There is one retrospective pooled analysis of MOSAIC, ECF4584 and EFC2962 that did not find an association between DM and increased risk of peripheral neuropathy in patients receiving oxaliplatin, suggesting that patients without symptoms of neuropathy can be treated with oxaliplatin with no greater risk of cumulative peripheral neuropathy than non diabetics [Ramanathan et al. Ann Oncol 2010 21 (4): 754-758]. Granted, it is a retrospective analysis and the sample of diabetic patients was not large.
Thanks again for a great discussion


Leonard B. Saltz, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 21, 2015 9:30 AM

Course Faculty Response

Roberto - You raise an interesting question regarding neuropathy and diabetes.  Several factors influence my thinking on this. The first is that I feel we really do not have great data on neuropathy from oxaliplatin, because I believe it has been very much under-appreciated, and so under-reported in studies.  Consider how both the incidence, as well as the frequency of irreversibility, changed in the Mosaic study between the initial 2004 NEJM report and the 2009 JCO follow up. Early on, investigators were not focused on the neuropathy, and likely did not interview extensively about it. This was early in the days of every other week treatment cycles, and we likely did not ask in great detail about events from two weeks ago. So consider the patient that felt neuropathy for a week from a particular dose.  I expect that by the time the patient was seen for the next treatment, the neuropathy was in the past, and unless specifically asked about it, he/she would have been unlikely to bring it up, as patients often try to take a positive, optimistic attitude, and also worry about being seen as “complainers” if they bring up problems from the past that are not active. Many also worry that if they report too many side effect their chemo dose will be decreased, and they are fearful that this will lessen the effectiveness of their treatment. Consider too that the “coasting phenomenon” that we all see, in which the neuropathy often worsens and peaks as late as three to four months after the last dose of oxaliplatin is given, has never been formally reported or quantitated, and so patients may often be “off study” when the neuropathy is at its worst.  

We know that virtually any pre-existing damage to sensory nerves is a potential vulnerability to oxaliplatin neuropathy, and it is hard to believe that a ten year history of diabetes would not have precipitated some degree of neuropathy, even if it is subclinical. As such, I worry that the potential for diabetes to impact on oxaliplatin neuropathy is always there.  Note that it is one factor that goes into the thinking.  It is not a strict contraindication.  

Finally, my thinking is influenced by the equivalence of irinotecan and oxaliplatin in first line care.  In the U.S., oxaliplatin is used more often in first line.  In many countries in Europe, first line Irinotecan is more common, confirming that what we have is regional preferences, rather than data for superiority of one over the other.  In my practice, I individualize (low tech “personalized medicine”). I consider and discuss both irinotecan and oxaliplatin with patients as equally effective options and work with each patient to decide which toxicity profile is less likely to be problematic for him/her. I wind up using about half FOLFIRI and half FOLFOX (makes sense….I’m in New York, which is somewhere half way between the U.S. and Europe in many ways). In this particular case, the potential increase in neuropathy risk leans me towards using irinotecan first. If the patient had reason to be particularly concerned about alopecia, which is substantially higher with irinotecan than with oxaliplatin, then that might sway me the other way….the patient and I would discuss the relative risks and relative importance to him/her of the side effects and then decide together which to use first. Note that when I do use oxaliplatin up front, I always use the “OPTIMOX” approach, and discuss this with the patient at the start, so that he/she knows that the plan is to hold the oxaliplatin after the first 6 doses and then reintroduce it at a later time. 


Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 21, 2015 7:20 PM

Hi, I really enjoyed the highly interactive discussions on this case.  Dr. Hamilton, thank you very much for providing the great up-to-date review article on this topic.  As you mentioned, there seem to be sufficient evidences to support routine expanded RAS testing.  In addition to the unnecessary toxicities and costs, the PRIME study also raised concern for possible harm of adding anti-EGFR therapy to patients negative for KRAS exon2 mutation but positive for other RAS mutations.  Although the latter finding was not reproducible in some other studies, the possible rationale as discussed by Dr. Jordan Berlin (n engl j med 369;11) sounds plausible: “suppression of nonmutated RAS with an EGFR inhibitor would theoretically 'activate' mutated RAS isoforms by releasing the suppressive effects of the nonmutated RAS isoforms”.  I wonder what people’s thoughts and experiences are regarding the expanded RAS testing: do you run KRAS exon 2 first followed by other less common mutations as needed, or do you run all mutations together upfront before treatment decision is made?  What factor plays a more important role: cost of test, TAT, indications, etc?  Thanks.  


Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 22, 2015 10:38 AM

Good morning,  The interactive discussions on this case along with your comments were very informative.  Do you feel that sufficient evidence to support routine expanded RAS testing exists and would you recommend it upfront or would proceeding in a more step-wise fashion like our example case?  Do you feel this testing could be beneficial when used routinely in other organ system tumors with known RAS mutations?  Thank you for this enlightened discussion.


Manoel Carlos Leonardi de Azevedo Souza, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 22, 2015 2:23 PM

1- All chemotherapy schemes proposed in the case are similar in orerall survival for the patient describred. In stage IV colon cancer, i would be more conservative in indicating a strategy regarding the use of molecular agent in first line therapy, since it is a uncurable disease. Therefore, FOLFOX or FOLFOXIRI would be my first choice for the patient. In an eventual progression of disease, i would add the biological agent with a second line  chemotherapy regimen. The main goal here is the prolonged OS, not response rates. To maintain good performance is essential to thepatient, once the use of all chemotherapic agents is the most effective way to control the metastatic disease.
2- The All-RAS molecular research is very important and should be followed in this case.
3- The use in second line of Panitumumab/Cetuximab, associated with citotoxic drug regimen, would be my choice.
4- Submitting the patient to secrennig tests to Bethesda criteria for Lynch syndrome is helpful, although not cost effective in my country.   
5- Thank you all for the comments and opinion on the case.. it is a very good learning experience. 


Leonard B. Saltz, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 23, 2015 9:42 AM

Course Faculty Response

As many of you have correctly noted, all-RAS testing is now the “coin of the realm,” and is what we should be doing on the tumors of all our patients who have colorectal cancers that have metastasized (Douillard, J.-Y., et al., Panitumumab–FOLFOX4 Treatment and RAS Mutations in Colorectal Cancer. New England Journal of Medicine, 2013. 369(11): p. 1023-1034). Patients such as the one presented here, in whom only exon 2 KRAS has been done and in whom the result has been wild type, or non-mutated, need to have the remained of the RAS genes tested. (Digression: It is best that we try to use the term “RAS” testing, and stop speaking of just “KRAS”, and to try to avoid using the terms RAS (+) or RAS(-), as these are ambiguous; does RAS (+) mean normal or mutated?….I’m not sure.)

Since there is virtually complete concordance between primary and metastases,(Vakiani, E., et al., Comparative Genomic Analysis of Primary Versus Metastatic Colorectal Carcinomas. Journal of Clinical Oncology, 2012. 30(24): p. 2956-2962.)  it is fine to perform the analysis on archived paraffin-embedded tissue from the primary in those patients whose disease has  recurred; it is not necessary nor appropriate to subject a patient to a biopsy of a metastasis for RAS genotyping if the tissue block from the primary is available.  

It should be noted that there is no reason to send stage I, II, or III tumors for RAS testing, as there is no decision that we would make with that information; the role of RAS testing is to determine whether or not anti-EGFR monoclonal antibodies are going to be a consideration in the treatment of that particular patient. Since there is no role for use of panitumumab or cetuximab in any stage I, II, or III patient, ordering RAS genotyping on such patients incurs cost without benefit; most stage I, II, and III patients will never need the information. 

We should not think that the original KRAS exon 2 mutations that we have been routinely checking for the past several years are any more or less important that the “newer” non-exon 2 KRAS mutations or the NRAS mutations. We have simply been slower to recognize the importance of these. Given their equal importance, all RAS mutations should be assayed for at the same time in a new patient.  Most labs are now doing this on a multiplexed platform of some sort, such that all the information is obtained at once, rather than in any sort of sequential fashion. BRAF mutation is also typically obtained at the same time, as available data suggest that patients with BRAF-mutated colorectal cancer are extremely unlikely to benefit from non-first line use of anti-EGFR agents.  As such, I do not use cetuximab or panitumumab in patients with BRAF-mutated colorectal cancer. 

For a patient such as the one under discussion, who has had only exon 2 KRAS assayed and who does not have an identified mutation, the non-exon 2 KRAS and NRAS genes need to be tested.  However, note that if there is an exon-2 KRAS mutation, then further RAS testing would not be needed, as the information available already tells you that cetuximab and panitumumab are off the table, and further RAS testing would provide no useful or actionable information.

Some clinicians have expressed concerns regarding insurance coverage for all RAS testing. This should not be an issue, since NCCN guidelines now clearly require all-RAS testing as part of standard practice. The exact words form the current NCCN guidelines, version 2.2015, posted October 3, 2014, are: “All patients with metastatic colorectal cancer should have tumor tissue genotyped for RAS mutations (KRAS and NRAS). Patients with any known KRAS mutation (exon 2 or non-exon 2) or NRAS mutation should not be treated with either cetuximab or panitumumab.”  The statement appears on page COL-A 4 of 5, and the words  "KRAS/NRAS WT gene only" appear next to every cetuximab or panitumumab listing in the guidelines. 


ASCO University
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 23, 2015 9:48 AM

Patient Case Update
Based on more recent data on extended RAS mutation testing, his tumor blocks were submitted for full KRAS and NRAS testing.  His current ECOG PS is 1, with mild RUQ pain.  His alkaline phosphatase is 1.5x ULN, and his bilirubin has remained normal.
Testing Platform/Molecular Profiling: Additional molecular testing order for “extended RAS”: 
Pre-analytics: As previously
Analysis: Mutation testing in Clinical Laboratory Improvement Amendments (CLIA)-accredited laboratory with laboratory-developed 50-gene “hot spot” next generation sequencing solid tumor panel (billed for KRAS and NRAS only)
See attached report.
Results for adenocarcinoma:
For ordered genes: Pathogenic mutation in KRAS codon 61
Additional findings:  Pathogenic mutations in APC and TP53
Probable germline mutation:  KIT (germline DNA not used in assay)


ASCO University
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 23, 2015 9:54 AM

Discussion Questions:


  1. Are non-exon 2 KRAS mutations or NRAS mutations as strong a contraindication to use of cetuximab or panitumumab as exon 2 KRAS mutations, and are codon 13 (G13D) mutations as strong a contraindication as other exon 2 mutations?

  2. What would be the preferred molecular testing for patients presenting with Stage IV or recurrent colorectal cancer? Is a combination of methods appropriate? Some possible approaches include:

    1. Evaluating either a biopsy specimen of the primary tumor (Stage IV patient) or the prior resection specimen of the primary tumor (recurrent disease patient) for microsatellite instability status by immunohistochemistry and for microsatellite allelic shifts
    2. Evaluating a biopsy specimen of a metastasis for microsatellite instability status by immunohistochemistry and microsatellite allelic shifts
    3. Evaluating the primary tumor for KRAS exons 2, 3 and 4; NRAS exons 2 and 3; and BRAF exons
    4. Evaluating a metastasis for KRAS exons 2, 3 and 4; NRAS exons 2 and 3; and BRAF exons
    5. Evaluating a “liquid biopsy” with a multigene panel

  3. What is the role in standard practice, if any, of next generation sequencing evaluation for mutations beyond KRAS, NRAS, and BRAF?


James A. Knost, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 26, 2015 6:43 PM

Have we learned enough from the KRAS vertical pathway suppression studies to warrant BRAF testing on all patients? Is there a clinical subset of colon cancer patients that you would consider KRAS to BRAF up front? If you knew a patient with metastatic disease had a BRAF mutation ( you might find this in a patient with IHC mismatch findings in which you did KRAS reflexing to BRAF to rule out Lynch) would you be inclined to use FOLFIRINOX up front?


Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 27, 2015 5:22 PM

Here are some of my thoughts to the questions and your comments/corrections are welcome. 1. According to the review by Atreya et al. (JCO 2015), retrospective analyses suggested that EGFR antibodies may have a modest benefit (intermediate between that of KRAS wild-type and codon 12–mutated tumors) on patients with KRAS G13D mutations.  However, the effect was commented to be not robust enough to merit practice change.  I’m not sure if there is adequate data to assess whether each individual expanded RAS mutations is as “equally” strong a contraindication as exon2 KRAS mutations, but there have been both preclinical and clinical data from multiple phase II to III clinical trials to support them as negative predictors to anti-EGFR therapy.

2. Depending whether Lynch syndrome is considered, maybe a combination of approaches 1 and 3.
3. Although mutations in other genes (such as PTEN, PIK3CA) are candidate predictors, further evidences seem to be needed to better elucidate their impacts on clinical benefit.


Leonard B. Saltz, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 29, 2015 2:54 PM

Course Faculty Response

Currently, although there are some suggestions that some RAS mutations may differ from others, we do not have data to justify use of either cetuximab or panitumumab in any patient with any known KRAS or NRAS mutation.  Some preliminary data had suggested that G13D mutations were “less of a contraindication” to anti-EGFR therapy, however this has not been consistently borne out, and use of anti-EGFR agents in patients with known G13D mutations is not recommended, and is outside of both FDA approval and outside of NCCN guidelines.  It is important to recall that there is evidence that treatment of a RAS-mutated tumor with EGFR agents not only is not beneficial, but appears to actually be associated with a modest detriment in long term outcome.

The question of use of FOLFOXIRI, with or without bevacizumab, was raised by one of you in this discussion.  It is a point worth elaborating on. I’m not yet comfortable with the use of FOLFOXIRI in the treatment of American patients. Note that the published trial was performed in Europe (Loupakis, Fotios, et al: NEJM 2014; 371: 1609-1618)  and uses a substantially higher dose of 5FU than is typically used even with 5FU/LV alone. How much this contributes to the outcome of the trial is unclear. However, American patients have been shown to tolerate fluorpyrimidines less well than the rest of the world (Haller, Daniel, et al: J Clin Oncol 2008;26: 2118-2123), and whether this regimen will be safe and tolerable at these doses in the U.S. is something I would like to see some published data on.

The question was also asked as to whether, since the BRAF-mutated tumors appeared to do well on the trial, perhaps this is an appropriate regimen for these patients. That is an interesting hypothesis, but one that will require more investigation before I would be comfortable accepting it. Recall that in the FOLFOXIRI trial 90% of patients were ECOG 0 and the rest were ECOG 1.  Because BRAF-mutated tumors often portend an aggressive growth pattern, peritoneal involvement, and a poor prognosis, many patients with BRAF mutated colorectal cancer have a compromised performance status at presentation and so would not have been candidates for the trial, and would be unlikely to tolerate FOLFOXIRI well. It may be that the subset of patients with BRAF mutated tumors who present with ECOG 0 PS might benefit from an aggressive combination such as FOLFOXIRI, but this is a question that remains unsettled. Dr. Sabine Tejpar, in her elegant Keynote Address at the 2015 GI cancer Symposium, presented intriguing data regarding subsets of BRAF-mutated colorectal cancer, and it may be that these different subsets might help explain differences in presentation, and may one day provide guidance for therapy.

An important, albeit sobering, point to consider is that thus far KRAS, NRAS and BRAF are the only mutations that we know how to utilize in standard practice, and are the only ones recommended as part of routine care by NCCN guidelines. The use of multiplexed assays to assess RAS and RAF are commonly utilized, however, other genetic information, including PiK3CA, AKT, APC, TP53, etc., may ultimately prove useful one day, but at present these are not actionable, and should not be regarded as part of standard care. In the context of investigational programs, such multigene assays, either on tumor tissue or using “liquid biopsy”, measuring circulating DNA in blood or in urine, may help select patients for specific investigational trials, however we should not lose sight of the fact that the very use of these genetic selection factors are themselves part of the unproven experimental hypothesis being tested (note, for example, that PIK3CA or AKT mutations do not, thus far, appear to predict for activity of investigational agents targeting these pathways in colorectal cancer). As such, finding a mutation does not equate with finding a therapeutic option, and use of Next Generation Sequencing assays to find mutations beyond RAS and BRAF are at this time an investigational endeavor and not part of standard practice. We all hope, of course, that new data will change this, but at present we await such data.   


Richard L. Schilsky, MD, FASCO, FACP
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Jan 31, 2015 5:04 PM

I want to commend all the participants in this case discussion for their excellent commentary and questions. In particular, I want to thank Drs. Saltz and Hamilton for their erudite and stimulating discussions and timely responses. This is exactly the kind of community discourse we hoped to stimulate with this new Molecular Oncology Tumor Board series.


Leonard B. Saltz, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Feb 05, 2015 9:21 AM

Course Faculty Summary

Thanks to everyone who participated in this discussion. In summary, these are my individual interpretations based off of the available information:

  1. Irinotecan or oxaliplatin based combinations are equally safe and efficacious for first line treatment of metastatic colorectal cancer. Toxicity patterns differ, and individual patient preferences and acceptance of different toxicity patterns, or individual patient vulnerabilities (such as pre-existing neuropathy) should be considered and discussed with the patient in deciding which to use first.
  2. Bevacizumab is reasonable with all first line regimens in the absence of a contraindication.
  3. I do not favor use of cetuximab or panitumumab in first line, regardless of RAS status, as I feel the preponderance of data do not suggest a better outcome for this, and the skin toxicity and higher cost are strongly negative factors. Either one of these agents (but not both…one does not work after progression on the other) should be used in the non-first line treatment of patients with all RAS wild type, BRAF wild type colorectal cancer. I typically utilize these agents in 3rd line, after failure of both irinotecan and oxaliplatin-based regimens.
  4. Tumor status for all RAS (KRAS exon 2 and non exon 2, and NRAS) mutations should be assayed once stage IV metastatic colorectal cancer is diagnosed. Assay may be performed on archived tissue from the primary; as per NCCN guidelines, it is not usually necessary, nor appropriate, to subject a patient to a biopsy of a metastasis simply for RAS genotyping. Currently there is no role for obtaining RAS mutation status on stage I, II, or III colorectal cancer.
  5. All patients below the age of 70, plus those who meet Bethesda Criteria, should be screened for Lynch syndrome. Some, but not all, expert panels have advocated such screening for all patients, but others have suggested that the yield in screening patients 70 and over with no family history is very low, and such universal screening may not be cost effective. (Stoffel et al, J Clin Oncol 2015;33: 209-217)
  6. At our institution, screening is performed by immunohistochemistry for mismatch repair proteins. Other centers use PCR for microsatellite instability. MD Anderson uses both.  
  7. Genotyping beyond KRAS, NRAS, and BRAF is not a recommended part of standard management of colorectal cancer at this time, as we do not yet have an actionable way of utilizing information on these non RAS, non BRAF mutations. Extensive research efforts are ongoing to try to change this.


Stanley R. Hamilton, MD
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Feb 05, 2015 9:24 AM

Course Faculty Summary

I would also like to thank everyone for participating in this discussion. In summary, these are my individual interpretations based upon the available information:

  1. “Fit-for-purpose” testing for actionable alterations is essential.
  2. Prognostic and predictive biomarkers are now in routine use.
  3. In my opinion, in patients with Stage IV or recurrent disease, a proximate biopsy specimen of a metastatic lesion(s) should be considered, especially in clinical trials, because of the possible occurrence of tumor evolution over time and effects of prior therapy on the characteristics of recurrent disease , and differences between primaries and metastases at presentation in Stage IV disease.
  4. Intra-tumoral heterogeneity with subclones, intra-patient inter-tumoral heterogeneity among multiple metastases, and genomic instability present problems for evaluation of some actionable alterations.
  5. Predictors of Lynch syndrome are moderately accurate, and more universal testing for evidence of the syndrome is now recommended in multiple guidelines.  (Stoffel et al, J Clin Oncol 2015;33: 209-217)
  6. Immunohistochemistry for loss of mismatch repair gene products is widely available, quick, directs attention to likely germline mutation, and identifies patients with germline MSH6 mutation, but does not evaluate microsatellite instability status, is most commonly abnormal due to somatic MLH1 methylation, and has false negative results for some Lynch syndrome patients.
  7. DNA-based microsatellite instability testing fails to identify some patients with germline MSH6 mutation.
  8. FDA-cleared assays are not always the most efficacious.
  9. The low frequency of many potentially actionable alterations complicates the generation of strong levels of evidence.
  10. The pace of emergence of new knowledge is rapid.
  11. Reimbursement for molecular testing is remarkably heterogeneous and problematical.


ASCO University
Re: Extended RAS Mutation: Molecular Oncology Tumor Boards
Feb 05, 2015 9:29 AM

Thank you to Drs. Hamilton and Saltz for leading the discussion of this case and also to all of those who contributed to the conversation! The forum is now closed to further comments but users have the opportunity to claim CME credit on ASCO University by accessing this link.

Please check back in mid-February for a new case in this series related to Lung Cancer.