Beyond KRAS exon 2: Research at ASCO and Important Treatment Implications for Metastatic Colorectal Cancer
By Julie Katz, MPH, M.Phil. and Arnold DuBell, Ph.D.
Biomarkers in metastatic colorectal cancer (mCRC) played a prominent role in the research presented at the 2014 American Society of Clinical Oncology (ASCO) annual meeting. New data presented at ASCO 2014 showed clinical implications of CRC biomarkers, including the RAS family of mutations. In our post-ASCO discussion, we illustrate the influence of biomarkers on testing and treatment of mCRC, of genetic changes throughout the course of disease, and treatment patterns. These issues have potentially profound implications on a tumor’s behavior and sensitivity throughout the course of the disease. The biomarker research presented represents different approaches to cancer treatment for mCRC.
In 2014, CRC has the fifth highest incidence of all cancers in the United States after breast, prostate, non-small cell lung, and melanoma. Among stage IV incidence cancer cases, CRC is the fourth most common cancer in the United States, behind only non-small cell lung, pancreas, and head and neck. CRC is also the fifth most prevalent disease in the United States (five-year prevalence) behind breast, prostate, melanoma and non-small cell lung cancers. In Western Europe, CRC is the third most frequently diagnosed malignancy and the second most common cause of cancer death, behind lung cancer. In Japan, it is the third most frequently diagnosed malignancy after gastric and lung cancer.
History of EGFR inhibitors and KRAS status
The EGFR monoclonal antibodies, Erbitux® (cetuximab, Bristol-Myers Squibb, Eli Lilly, Merck KGaA) and Vectibix® (panitumumab, Amgen), have been shown to provide differential clinical benefit based on KRAS exon 2 mutational status. Table 2 presents a synopsis of the retrospective clinical data published to date for Erbitux and Vectibix in mCRC patients treated in the first line, according to KRAS status. These data firmly established that EGFR inhibitors, such as Erbitux and Vectibix, provide significant clinical benefit in patients with wild-type KRAS, but that patients with mutant KRAS in exon 2 derive no clinical benefit from these agents.
Beyond KRAS exon 2: Research at ASCO and Important Treatment Implications for Metastatic Colorectal Cancer (cont.)
Given the lack of benefit (and some data suggest a negative effect) and the high cost – both financial and toxicity – of these agents, the National Comprehensive Cancer Network (NCCN) has amended their Guidelines for colon and rectal cancers to recommend the use of Erbitux and Vectibix only in mCRC patients with wild-type KRAS. Additionally, in Europe, Vectibix’s initial approval label was restricted to use in KRAS wild-type patients; Erbitux’s label was amended in July 2008 to restrict its use to KRAS wild-type patients in all lines of therapy, and the recent positive Committee for Medicinal Products for Human Use (CHMP) opinion for Vectibix in combination with systemic therapy was limited to patients with wild-type KRAS. The U.S. FDA followed suit in 2009 by adding a recommendation to the existing labels for both drugs that they only be used in patients with wild-type KRAS.
According to a survey by Kantar Health for Treatment Architecture, 92% of physicians reported that they are testing for the presence of KRAS mutations at some point in the treatment of their CRC patients; this nearly universal use of KRAS testing is similar to practices in Western Europe and Japan, and reflects both the widespread availability of the KRAS mutational test in these developed countries and awareness of the need to test patients.
However, mutations in KRAS exon 2 are not the only possible mutations in this gene. Moreover, although the use of KRAS exon 2 testing identifies patients who are likely to have a response to anti-EGFR therapy, a fraction of these patients do not benefit from this treatment. In order to understand this, a retrospective analysis of the PRIME study evaluated the additional benefit one might observe in the use of an increased definition of having a KRAS mutation.
PRIME randomized 1,183 previously untreated patients to receive FOLFOX or FOLFOX plus Vectibix. In the original presentation (Douillard, J Clin Oncol, 2010) and as presented in Table 2 above, the benefit of Vectibix was identified only in patients with KRAS wild-type (in exon 2). However, PRIME also prospectively planned, following patients forward, to study all RAS mutations based on the retrospective analysis done on KRAS mutations. This was based on stratification of other mutations in KRAS as well as mutations in NRAS. The inclusion of other KRAS mutations as screening has repercussions: the inclusion of requiring these other sites to also be wild-type improved the hazard ratio for progression or death with the combination therapy from 0.80 to 0.72, indicating a lower chance of an event occurring. More importantly, progression-free survival was not improved by Vectibix in patients who were KRAS exon 2 wild-type but had mutations in other parts of KRAS (HR 1.28). Data such as this helped reinvigorate the examination of biomarker in CRC (Douillard, NEJM, 2013).
The Phase II PEAK Study, which was presented at the ASCO 2013 annual meeting, evaluated Vectibix not only in mCRC patients with wild-type KRAS exon 2 but also in patients wild-type for exons 3 and 4 of KRAS and exons 2, 3 and 4 for NRAS. This study compared Vectibix plus mFOLFOX6 or Avastin plus mFOLFOX6 in 285 randomly assigned patients. In the KRAS wild-type exon 2 intent-to-treat group, progression-free survival was not significantly different between the two arms (HR 0.62, p=0.353). Median overall survival was significantly better in the Vectibix arm over the Avastin arm (34.2 months versus 24.3 months, HR 0.62, p=0.009).
In the wild-type RAS subgroup (defined as KRAS having no mutations in exons 2, 3 and 4 as well as being NRAS wild-type), progression-free survival and overall survival favored the Vectibix arm over the Avastin arm. Progression-free survival was improved by the addition of the targeted therapy (HR 0.65, p=0.029), and the benefit of targeted therapy was still apparent for overall survival (41.3 months versus 28.9 months, HR 0.63, p=0.058; Schwartzberg, J Clin Oncol, 2014). Findings such as these led to a continued push to understand how far-reaching these observations are, leading to several exploratory retrospective analyses presented at the ASCO 2014 meeting.
The Evolving Story of the RAS Family of Mutations
Using the tumor samples evaluable for RAS mutation status from the original intent-to-treat study populations of the OPUS, CRYSTAL and PRIME studies, retrospective analyses were performed and presented at the ASCO 2014 annual meeting. The study question for these retrospective analyses was, “What is the influence on EGFR-inhibitor treatment outcome of other identified activating mutations at other exons within KRAS or NRAS?” These studies hypothesized that the KRAS exon 2 mutation is not the only biomarker in the RAS family with important treatment implications for mCRC patients.
In the OPUS study, the intent-to-treat population (n=337) was randomized to FOLFOX4 and Erbitux or FOLFOX4 alone. This first-line treatment study showed that in KRAS exon 2 wild-type patients’ response and progression-free survival were significantly improved. However, in the KRAS exon 2 mutant population, the addition of Erbitux was negatively associated with the treatment outcome (Bokemeyer, Ann Oncol, 2011). In a continued analysis of these data (Bokemeyer, Abstract 3505, ASCO 2014), mutation status was evaluable in 118 of 179 (66%) patients. In those with RAS wild-type tumors, response was significantly improved by the addition of Erbitux to FOLFOX4 (57.9% versus 28.6%, p=0.008). New RAS mutations were detected in 31 of 118 (26%) patients. In KRAS exon 2 wild-type patients who had these other KRAS mutations, the addition of Erbitux was associated with a lack of benefit (mean progression-free survival: 7.5 months versus 7.4 months, HR 0.77, p=0.60).
The CRYSTAL study added further strength to the analysis of OPUS as well as PRIME. Adding Erbitux to FOLFIRI also significantly improved progression-free survival, overall survival and response in the first-line treatment of patients with KRAS exon 2 wild-type (Van Cutsem, NEJM, 2009). In a further retrospective analysis of the CRYSTAL data, mutational status was evaluated in all evaluable KRAS exon 2 wild-type tumors from (430 of 666, or 65%) patients.
Beyond KRAS exon 2: Research at ASCO and Important Treatment Implications for Metastatic Colorectal Cancer (cont.)
New RAS mutations were detected in 63 of 430 (15%) of KRAS exon 2 wild-type patients. In those with RAS wild-type tumors, a significant benefit across all endpoints was associated with the addition of Erbitux to FOLFIRI (ORR: 66.3% versus 38.6%, p<0.0001; mean progression-free survival: 11.4 months versus 8.4 months, HR 0.56, p=0.0002; mOS: 28.4 months versus 20.2 months, HR 0.69, p=0.0024).
As in the OPUS study, in patients with wild-type KRAS in exon 2 but with mutations in other locations in KRAS or NRAS, no benefit from the addition of Erbitux to FOLFIRI was apparent in progression-free survival (7.2 months versus 6.9 months, HR 0.81, p=0.56) or overall survival (18.2 months versus 20.7 months, HR 1.22, p=0.50). The conclusions of the study were that patients with RAS wild-type tumors derived a marked benefit from the addition of Erbitux to FOLFIRI, while patients with any RAS tumor mutations did not benefit (Ciardiello, Abstract 3506, ASCO 2014).
Epidemiological Implications of the Other RAS Mutants among the KRAS exon 2 Wild-Types
The findings from these numerous retrospective analyses regarding the impact of extended RAS mutants on efficacy of EGFR monoclonal antibodies has significant epidemiological implications. A systematic review of the literature in Kantar Health’s CancerMPact® Biomarker Analysis Report finds that approximately 36% of U.S. CRC patients possess KRAS exon 2 mutations. This translates to nearly 18,000 mCRC patients in 2014 ineligible for first-line treatment with an anti-EGFR monoclonal antibody. However, the retrospective analyses from the trials described above suggests that 15% to 26% of the patients who are seemingly wild-type for KRAS (exon 2) in actuality possess other RAS mutations, raising the population of first-line mCRC patients who are unlikely to benefit from an EGFR inhibitor to somewhere between 22,675 to 26,383 in the U.S. in 2014, representing 45% to 53% of the treatment-eligible population. For a tumor as common as CRC, the percentage of patients not benefiting from a cutting-edge targeted therapy is considerable.
KRAS Discordance between Primary Site and Metastasis
In most cases, current clinical practice involves assessment of biomarker status on a biopsied or surgically resected tumor sample. This is done purely for feasibility and ease – but in reality, is evaluation of biomarker status in a solitary lesion telling us the whole story in a patient with disseminated disease?
Dr. Kopetz presented an abstract at ASCO 2014 that sought to determine whether KRAS biomarker status was consistent between various lesions in mCRC patients. In this study, sequencing was successfully completed on 115 pairs of primary and metastatic tissue. The concordance rate for KRAS mutation was 89%. However, chemotherapy treatment was associated with 3.5-times higher odds of discordance compared to patients who did not receive therapy between resection of primary and metastatic tumors (P = 0.008; Kopetz, Abstract 3509, ASCO 2014).
In another presentation, a cohort of 10 patients with tissue from the primary site and one or more metastases, samples were compared for presence of mutations and variant allele frequency. Analysis for the six patients with KRAS exon 2 mutation and one patient with NRAS mutation showed mutation persistence from the primary tumor to the metastasis. However, KRAS variant allele frequency differed between primary and metastases based on interval treatment. There was shown to be a 2.4-fold elevated allele frequency between the primary tumor and the metastasis in three mCRC wild-type KRAS exon 2 patients who received Erbitux as third-line therapy (Graham, Abstract 3510, ASCO 2014).
Based on these data, we are beginning to learn that the rates of KRAS discordance between primary and metastatic mutations may vary considerably and can be influenced by prior treatment. These studies highlight the importance of considering the entirety of the disease when treating a patient with mCRC. It is likely that other RAS mutation testing will also be subject to mutation status discordance between the primary tumor and metastasis in mCRC, as well as changing mutation clonal status over the course of therapy, all of which will have important treatment and prognostic ramifications.
At the ASCO 2014 annual meeting, further data were presented highlighting that anti-EGFR monoclonal antibodies are inactive not only in patients with exon 2 KRAS mutations, as originally reported, but in patients with a mutation in any part of KRAS or other RAS genes. Data from retrospective, exploratory analyses presented at the 2014 ASCO annual meeting, building upon past studies, suggest that it is not sufficient to test mCRC patients for KRAS exon 2 mutation status to determine which patients will gain a treatment benefit from EGFR inhibitors. More prospective studies are necessary. All RAS mutations need to be diagnosed and seemingly only those patients wild-type for all RAS genes should be treated with EGFR-inhibitors. While the European Medicines Agency (EMA) has already changed the label for Erbitux and Vectibix to “exclude all RAS mutations,” the FDA still labels these agents as “excluding KRAS mutations.” There is now convincing evidence that the label in the U.S. should be restricted like the European label. Discussions to expand U.S. labelling are presumed to be under way.
About the Contributor
Julie Katz, MPH, M.Phil. is Associate Consultant, Global Oncology Epidemiology at Kantar Health
Arnold DuBell, Ph.D., is a Consultant at Kantar Health
Kantar Health is a leading global healthcare advisory firm and trusted advisor to the world’s largest pharmaceutical, biotech, and medical device and diagnostic companies. It combines evidence-based research capabilities with deep scientific, therapeutic and clinical knowledge, commercial development know-how, and marketing expertise to help clients launch products and differentiate their brands in the marketplace.
Kantar Health’s oncology-related offers include Oncology Conference Insight, client-directed oncology conference coverage that analyzes the most important research at significant oncology meetings; and
CancerMPact® Treatment Architecture, which assesses the current clinical management of cancer patients by site and stage for all treatment modalities.
If you would like us to act as catalysts for you, contact us at www.kantarhealth.com/contactus.