December 2012, Vol 1, No 6

← Back to Issue

KRAS and Colorectal Cancer: Shades of Gray

Chloe E. Atreya, MD, PhD

Colorectal Cancer

Key Points

  • Although RAS mutations at glycine-12 and glycine-13 are adjacent, identical substitutions at these positions (eg, G12S vs G13S) lead to very different levels of RAS activation
  • The central clinical question remains unanswered: will a patient with metastatic colorectal cancer harboring a KRAS G13D mutation benefit from anti-EGFR therapy?
  • If mutations in adjacent codons can have unequal effects, what level of “personalization” is required in oncology clinical research?

Retrospective subset analyses of metastatic or advanced colorectal cancer trials have not shown a treatment benefit for Erbitux or Vectibix in patients whose tumors had KRAS mutations in codon 12 or 13. Use of Erbitux or Vectibix is not recommended for treatment of colorectal cancer with these mutations.
–FDA update to indications and usage of Erbitux (cetuximab) and Vectibix (panitumumab), July 2009.

The association between Kirsten-Ras (KRAS) mutations, present in approximately 40% of colorectal cancers (CRCs), and lack of benefit from epidermal growth factor receptor (EGFR)-targeted antibodies is well known to oncologists. It is standard practice to test for KRAS codon 12 and 13 mutations in tumors of all patients with metastatic CRC (mCRC) because the results affect treatment decisions. Recently, however, 2 reports suggest that KRAS codon 12 and 13 mutations are nonequivalent, although they code for adjacent amino acids.1,2 In cetuximab-treated patients, the presence of a glycine (G) to aspartic acid (D) change at codon 13 of KRAS (c.38G>A, or p.G13D) was associated with intermediate survival, between that of patients with KRAS wild-type tumors and tumors harboring a KRAS mutation at glycine-12. While the overall impact of KRAS mutations on CRC prognosis remains unclear,3 G13D mutations may be associated with worse outcomes.1,2,4

The purpose of this review is to place differences between KRAS mutations at codon 12 and 13 in CRC into scientific and clinical contexts, and to consider the potential implications of nonequivalent mutations within the same gene for the pursuit of personalized medicine in oncology.

Scientific Context
The RAS guanine nucleotide-binding protein regulates cell survival, growth, and proliferation by acting as a molecular switch in response to extracellular signals. RAS regulates downstream pathways by cycling between an inactive guanosine diphosphate (GDP)-bound state and an active guanosine triphosphate (GTP)-bound state.5,6 Posttranslational processing targets RAS to the membrane, where it resides in the OFF state, bound to GDP.7,8 Upon stimulation by growth factors such as insulin and EGF, RAS activators called guanine nucleotide exchange factors localize to the membrane and stimulate the release of GDP from RAS, thus allowing GTP binding and activation.9,10 Once ON, RAS activates downstream signaling cascades, including the RAF-MEK-ERK pathway.11

Inactivation of RAS signaling requires hydrolysis of the RAS-bound GTP back to GDP with release of phosphate. However, RAS is extremely slow at carrying out this reaction on its own. Therefore, inactivation of RAS depends upon the action of GTPase-activating proteins (GAPs), which supply a critical arginine residue to the RAS active site to assist in catalysis of GTP hydrolysis.5,12

Dysregulation of RAS signaling is almost invariably associated with disease. Hyperactivating somatic mutations in RAS are among the most common lesions found in human malignancies. Although mutation of any of the 3 RAS isoforms (KRAS, NRAS, or HRAS) can lead to oncogenic transformation, KRAS mutations are by far the most common.13-16 Hyperactivation of the RAS pathway due to oncogenic mutations in other pathway members, including BRAF and EGFR, accounts for a significant proportion of RAS wild-type cancers.17

Three-dimensional structures of RAS isoforms have been well studied. Indeed, over 100 crystal structures of RAS in various complexes have been solved. Nucleotide-dependent signaling relies on conformational changes in 2 regions that border the nucleotide-binding pocket, switch I (residues 30-38) and switch II (residues 60-76) (Figure 1).18 Threonine-35 and glycine-60 make key hydrogen bonds with the ?-phosphate of GTP, holding switch I and switch II in their active conformations. Upon hydrolysis of GTP and release of phosphate, these regions relax into the inactive GDP-binding conformation. Only the GTP-bound conformation of the switch regions activates downstream effector proteins.

The regions surrounding the nucleotide-binding pocket contain the most common sites of RAS mutation. Along with the switch regions, the borders of the pocket include the phosphate-binding loop (P-loop, residues 10-17) and the base-binding loops (residues 116-120 and 145-147).19 The majority of oncogenic mutations occur at residues 12 or 13 in the P-loop, or residue 61 in switch II.17 Mutations activate RAS by 2 main mechanisms: decreasing GTP hydrolysis or increasing nucleotide exchange. Most oncogenic mutations follow the first mechanism. In wild-type RAS, residues 12 and 13 are both glycine, the smallest amino acid. Mutation of either of these residues to larger amino acids can occlude the critical arginine of the GAP, leading to decreased GAP binding and insensitivity to GAP-stimulated GTP hydrolysis.5,20,21 Glutamine-61 of RAS participates in catalysis, and therefore its mutation also affects GTP hydrolysis; however, GAP binding is generally normal or increased.12,21 Mutations near the base-binding region of RAS (eg, alanine-146 or phenyl­alanine-156) diminish nucleotide binding affinity and increase the intrinsic rate of exchange.22,23 The active form of RAS then predominates due to the higher concentration of GTP than GDP in the cell.

Biochemical and cellular assays demonstrate that although RAS mutations at glycine-12 and glycine-13 are adjacent, identical substitutions at these positions (eg, G12S vs G13S) lead to very different levels of RAS activation.21,24 Mutations at codon 12 are generally the most potent activators of RAS: the crystal structure of HRAS complexed with a GAP suggests that mutations at this position sterically clash with the arginine finger of the GAP, preventing both binding and catalysis.12 Indeed, mutation to any amino acid other than proline induces focus formation and anchorage-independent growth of rat fibroblasts, both measures of oncogenic transformation.24

Mutations at codon 13 are generally less activating in vitro. Three G13 mutations (G13S, G13V, and G13D) have been compared in GAP stimulation and focus formation assays. G13D completely blocks GAP activity and is the only codon 13 mutant that induces focus formation. G13S is only slightly activating, having a small effect on GAP stimulation. G13V shows intermediate oncogenicity; this mutation completely blocks GAP activity, without inducing focus formation.20,21 G12 mutations may confer greater resistance to apoptosis and increase the proportion of GTP-bound RAS well above the level achieved by G13 mutations.1,25-27

Glutamine-61 mutations decrease RAS GTPase activity but vary widely in transformation efficiency.28 Unlike G12 and G13 mutants, some Q61 mutants unexpectedly increase binding to GAP, which could hyperactivate wild-type RAS.21 The most potent mutant, Q61L, increases nucleotide exchange in addition to diminishing GTPase activity.22,29 Substitutions at alanine-146 are less activating than G12 mutations and may be comparable to G13 mutations in transforming efficiency.30 These base-binding mutations increase the intrinsic rate of nucleotide exchange,22 thus increasing RAS-GTP without affecting GAP stimulation. Despite lower oncogenicity, one would expect KRAS A146 mutants to be growth factor independent due to their capacity for autonomous nucleotide exchange and activation.31

The spectrum of RAS pathway mutations differs markedly by cancer type (Figure 2).17 Pancreatic cancer exhibits the highest frequency of KRAS mutations: 72% in the Sanger Catalogue of Somatic Mutations in Cancer and >90% in other series (Figure 2 B).17 By contrast, breast cancer (not shown) has a low frequency of RAS pathway mutations: 4% KRAS, 3% BRAF, 2% NRAS, and <1% HRAS mutations.17 Melanoma has a high frequency of RAS pathway mutations, but KRAS mutations are rare (Figure 2 D). In CRC, half of tumors exhibit a mutation in the RAS pathway (mutually exclusive), most often a missense mutation in KRAS itself (Figure 2 A). Unlike in lung cancer, EGFR mutations are rare in CRC.

Differences by cancer type are also observed in the distribution of KRAS single-base substitution mutations (Figure 2). Across cancers, mutations at codon 12 predominate: G12D and G12V are the first and second most common KRAS mutations. In lung cancer, however, the most frequent KRAS substitution is glycine to cysteine (G12C). Of all cancer types, CRC exhibits the greatest proportion of G13D mutations: ~20% of KRAS mutations and 8% overall. Of note, clinical testing for KRAS mutations is typically performed using polymerase chain reaction–based assays with probes specific for mutations in codons 12 and 13. Consequently, KRAS mutations at other codons, including Q61 and A146, may be underrepresented.32,33

Despite decades of RAS research, 5 years passed between FDA approval of cetuximab for treatment of EGFR-expressing mCRC and the label change to exclude patients with tumors harboring KRAS mutations. A major reason for the delay was that no correlation was observed between KRAS mutational status and sensitivity to cetuximab in CRC cell lines.34 This experience underscores the necessity of evaluating annotated human tumor specimens for associations between specific mutations, survival, and response to therapies, thereby enabling the clinical findings to guide further mechanistic investigations.

Clinical Context
A landmark study linking KRAS G13D mutations with clinical outcomes was a retrospective, pooled analysis of 774 patients with chemotherapy-refractory mCRC treated with cetuximab-based therapy from 3 data sets (the National Cancer Institute of Canada/Australasian Gastrointestinal Trials Group and a Leuven and an Italian data set) encompassing 7 clinical trials (CO.17, EVEREST, BOND, SAVAGE, BABEL, MABEL, and EMER202600).1

Three hundred ten patients (40%) had a KRAS mutation at codon 12 or 13, 45 of which were G13D (14.5% of KRAS mutations; 6% overall). Thirteen patients with KRAS G13D-mutated tumors received best supportive care on the CO.17 trial and experienced significantly worse overall survival (OS) compared with patients with other KRAS-mutated or KRAS wild-type tumors in a univariate analysis. After adjustment for potential prognostic factors (eg, age, sex, performance status, prior chemotherapy, and primary site) in the multivariate analysis, however, there were no survival differences by KRAS status (Table). Patients with G13D-mutated tumors who received any cetuximab-based therapy (monotherapy or cetuximab plus chemotherapy) experienced longer survival than patients with other KRAS mutations, with OS 7.6 versus 5.7 months and progression-free survival (PFS) 4.0 versus 1.9 months (Cox regression P=.005 and .004, respectively) (Table). But because G13D did not portend a worse prognosis in multivariate analysis, the significance of improved survival end points for cetuximab-treated patients with G13D-mutated tumors compared with other KRAS mutations is unclear. Moreover, all of the benefit occurred in patients treated with cetuximab plus chemotherapy, raising the possibility that outcomes could have been more affected by the chemotherapy. In the presence of a G13D mutation, cetuximab monotherapy (pooled data) did predict significantly better OS compared with no cetuximab (CO.17 trial) (P=.02 by log-rank test), and there was a trend toward significance in the CO.17 trial only. There was no significant difference in response rates between G13D and other KRAS mutants treated with cetuximab-based therapy. Overall, the results of De Roock et al1 provide greater support for G13D as a predictive rather than a prognostic marker.

Following these findings in the refractory setting, the association of KRAS G13D with outcomes in patients with mCRC treated with first-line chemotherapy with or without cetuximab was investigated by Tejpar et al in a retrospective analysis of 1378 patients from the randomized CRYSTAL and OPUS trials.2 Here, 533 patients (39%) had a KRAS mutation at codon 12 or 13, 83 of which were G13D (16% of KRAS mutations; 8% overall). Median OS was longer than in De Roock et al,1 consistent with first-line therapy, compared with the chemotherapy-refractory setting (Table). When treated with chemotherapy alone, patients with G13D mutant tumors had a worse outcome compared with patients with other KRAS mutant or KRAS wild-type tumors. When cetuximab was added to chemotherapy, significant improvements in PFS (median 7.4 vs 6.0 months, hazard ratio [HR] 0.47; P=.039) and tumor response (40.5% vs 22.0%; odds ratio 3.38; P=.042) were observed among patients with G13D mutations, although the OS benefit did not reach significance (Table). The cetuximab treatment effect in the G13D subset was comparable to that in patients with KRAS wild-type tumors, but OS and PFS were considerably lower. By contrast, survival worsened with the addition of cetuximab to chemotherapy among patients with other KRAS mutations (6.7 vs 8.1 months; HR 1.37).

To summarize, both of these large, pooled retrospective studies found that patients with KRAS G13D mutant mCRC may have a poor prognosis when treated with best supportive care or standard chemotherapy in the chemotherapy-refractory1 or first-line setting,2 respectively. In both studies, patients with G13D mutations appeared to derive modest benefit from the addition of cetuximab.

Smaller, single-arm studies have yielded complex results with respect to the prognostic significance of KRAS G13D mutations, a confounding factor that clouds interpretation of predictive impact.4,35 Evaluation of 229 Japanese CRC patients found a nonsignificant trend toward worse OS of patients with G13D mutations compared with patients with KRAS wild-type tumors (HR 1.67; 95% CI, 0.93-3.02; P=.086 on univariate analysis).4 Survival of patients with G12 mutant tumors was similar to KRAS wild-type, whereas patients with BRAF V600E mutant tumors had the worst OS (HR 3.78; 95% CI, 1.89-7.54; P?.001).4 On the other hand, 2 reports suggest a link between KRAS G13D and microsatellite instability (MSI) associated with favorable outcome.35,36 An analysis of mutation frequency in 158 hereditary nonpolyposis CRC (HNPCC) and 864 sporadic CRC cases identified a significantly higher rate of G13D mutations in HNPCC tumors (microsatellite unstable) compared with sporadic microsatellite-stable tumors (51% vs 12% of KRAS mutations).36 G13D mutations were also overrepresented in sporadic MSI-high tumors without hMLH1 methylation (27% of KRAS mutations).36 In a study of 94 hereditary and 404 sporadic CRC cases, KRAS G13D mutations were associated with a better prognosis in HNPCC and sporadic MSI-high subsets.35
A growing number of small studies are examining the predictive value of KRAS G13D with respect to cetuximab treatment.37-39 These studies, limited by size and clinical heterogeneity, highlight the importance of access to tissue from large groups of similarly treated patients for subset analyses.

Peeters et al40 presented an important counterpoint to the findings in De Roock et al1 and Tejpar et al2 at the 2012 American Society of Clinical Oncology (ASCO) Annual Meeting. Evaluation of 1083 patients treated with second-line FOLFIRI plus panitumumab (NCT00339183), 1096 patients treated with first-line FOLFOX4 plus panitumumab (PRIME, NCT00364013), and 184 patients who received panitumumab monotherapy (NCT00113776) yielded inconsistent associations between specific KRAS mutations and survival.40 Tejpar et al2 pooled data from the CRYSTAL (FOLFIRI ± cetuximab) and OPUS (FOLFOX4 ± cetuximab) studies because they found that irinotecan- and oxaliplatin-containing chemotherapy regimens produced similar treatment effects in patients stratified by KRAS mutation status. Conversely, Peeters et al40 reported a trend toward better OS among patients with G13D-mutated tumors treated with FOLFIRI plus panitumumab (N=39; P=.07), whereas G13D was significantly associated with worse OS among patients treated with FOLFOX4 plus panitumumab (N=46; P=.003), and G13D was not predictive or prognostic of outcome when the panitumumab studies were pooled. The possibility that different anti-EGFR agents, in addition to lines of therapy and chemotherapy backbones, may contribute to unequal outcomes in association with KRAS mutation status further complicates the G13D story.

More Questions Than Answers
Following these analyses of over 3000 patients, evidence to support KRAS G13D as an independent predictive and/or prognostic biomarker in CRC still has not matured. Returning to the question of biologic mechanism, is there a plausible explanation for intermediate effects of KRAS G13D with respect to both response to cetuximab and survival? Intermediate response to cetuximab is easier to explain because KRAS mutations at codon 13 are less oncogenic than mutations at codon 12.1,20,21,24 A less activating KRAS mutation may mean that the tumor is partially growth factor dependent and thereby somewhat responsive to EGFR inhibition. It is difficult to directly attribute worse survival to G13D mutations. More likely, G13D mutations could signify a different biological signature that portends worse prognosis.

Different carcinogenic triggers produce distinct patterns of mutations: depending on the chemical used, KRAS codon 12 or 61 mutant lung cancers can be induced in mice.41 In human lung cancers, KRAS mutations are common in smokers, whereas EGFR and ALK mutations are more frequent in never-smokers.42 In CRC, rather than being a driver, KRAS G13D mutations may develop secondary to another oncologic process and propagate because they confer a modest survival advantage for tumor cells. This hypothesis is supported by findings in cultured cells subjected to continuous cetuximab: KRAS G13D mutations accompanied emergence of resistance, and deep sequencing subsequently detected preexisting G13D mutations in a minority of parental cells.43 KRAS G13D is also a nontumor-initiating mutation in HNPCC, caused by defects in DNA mismatch repair genes that lead to a myriad of somatic mutations.36,44 A poor prognosis, BRAF-mutated–like gene expression signature was identified in 30% of KRAS mutant CRCs.45 Follow-up work presented at the 2012 ASCO Annual Meeting showed additional evidence for heterogeneity among KRAS mutant colon tumors but no significant division by specific KRAS mutation.46 A moderately aggressive CRC subgroup marked by KRAS G13D mutations has not been described.

If “shades of gray” exist among KRAS mutations, might the same be true for other oncogenes and tumor suppressor genes? The answer is, of course, yes. For example, phosphoinositide-3-kinase (PI3K, coded by the PIK3CA gene) is mutated in ~15% of CRCs. PIK3CA has 2 mutational hot-spot clusters located approximately 1500 nucleotides apart, corresponding to the helical domain (eg, E545K) and the kinase domain (eg, H1047R). Helical and kinase domain mutations induce gain of function by different mechanisms. Helical domain mutations are dependent on RAS binding for transformation; kinase domain activation is mediated by interaction with the p85 regulatory subunit.47

In CRC, helical domain mutations associate with KRAS mutations, methylation of the MGMT DNA repair gene, and CpG island methylator phenotype–low status, while kinase domain mutations correlate with serrated pathway features including BRAF mutation and MSI.48 As with KRAS, the distribution of PIK3CA helical and kinase domain mutations varies by tissue of origin. In CRC, the distribution is roughly two-thirds helical to one-third kinase domain mutations. Breast cancer has the highest rate of PI3K mutations (25%), and the ratio of helical to kinase domain mutations is opposite that in CRC.17 Even identical mutations can have differing effects by cancer type, as highlighted by dissimilar responses to BRAF inhibitors among patients with BRAF V600E CRC or melanoma.49

A final example of the potential for disparate impact of mutant alleles within a single gene involves the tumor suppressor protein p53. Mutation of p53 is the most common genetic lesion in human malignancies, including nearly half of CRCs. Whereas a limited number of mutations activate oncogenes, a greater variety of lesions can inactivate tumor suppressors. One study assessed effects of 2314 point mutations in the p53 DNA-binding domain.50 We ascertained p53 status in 604 tumors from the CALGB 89803 trial of adjuvant therapy for stage III colon cancer. Tumors were classified as p53 wild-type or mutation present in the zinc-binding (ZB) or the non–zinc-binding (NZB) regions of the DNA-binding domain. Overall, p53 mutational status was not associated with survival.51 However, among women, survival was superior with NZB mutations compared with wild-type p53 and worst with ZB mutations.51 A larger sample would be required to determine whether these differences could be attributed to specific p53 hot-spot mutations.

Implications for Personalized Medicine
The central clinical question remains unanswered: will a patient with mCRC harboring a KRAS G13D mutation benefit from anti-EGFR therapy? Based on the modest efficacy of anti-EGFR therapy even among patients with KRAS wild-type tumors, the sample size required for a prospective randomized study in the KRAS G13D subset would be prohibitive and perhaps inappropriate in light of the priority to identify new targets and more active agents. Additional large retrospective analyses of cohorts of CRC patients with matched clinicopathologic and treatment characteristics may provide greater precision in estimating the prognostic and predictive value of G13D, although there remains a distinct possibility that the impact is unsatisfyingly context-dependent and marginal. Are the differences meaningful, or are we trying to distinguish gray from grey?

KRAS G13D heralds an approaching onslaught of new questions as next-generation sequencing and other technologies reach the clinical arena. If mutations in adjacent codons can have unequal effects, what level of “personalization” is required in oncology clinical research? Even synonymous mutations (resulting in the same amino acid sequence) can produce differences in RNA processing that significantly alter protein levels or functions.52 We submit that validation of novel “personalized medicine” biomarkers should involve an iterative approach whereby hypothesis-generating data from clinical samples – such as the observations reviewed here suggesting a differential impact of KRAS G13D versus other KRAS mutations – should trigger a return to preclinical mechanistic studies in parallel with additional retrospective analyses of independent clinical data sets. Across these endeavors, tumor genotyping should be specific to the nucleotide level and compiled to identify broad signals of high-impact genetic events. As our understanding of CRC biology becomes more nuanced, we must steadfastly search for therapies with unequivocal benefit for molecularly defined subsets of patients.

Acknowledgments
CEA is supported in part by Postdoctoral Fellowship 11-183-01-TBG from the American Cancer Society and also acknowledges the support Millennium provided through the Alliance for Clinical Trials in Oncology Foundation. JMO is supported by the UCSF Medical Scientist Training Program NIH-NIGMS-MSTP (#GMO7618). We thank David Donner, PhD, for editorial assistance.

References

  1. De Roock W, Jonker DJ, Di Nicolantonio F, et al. Association of KRAS p.G13D mutation with outcome in patients with chemotherapy-refractory metastatic colorectal cancer treated with cetuximab. JAMA. 2010; 304:1812-1820.
  2. Tejpar S, Celik I, Schlichting M, et al. Association of KRAS G13D tumor mutations with outcomes in patients with metastatic colorectal cancer treated with first-line chemotherapy with or without cetuximab [published online ahead of print June 25, 2012]. J Clin Oncology.
  3. Yokota T. Are KRAS/BRAF mutations potent prognostic and/or predictive biomarkers in colorectal cancers? Anticancer Agents Med Chem. 2012;12:163-171.
  4. Yokota T, Ura T, Shibata N, et al. BRAF mutation is a powerful prognostic factor in advanced and recurrent colorectal cancer. Br J Cancer. 2011;104:856-862.
  5. Trahey M, McCormick F. A cytoplasmic protein stimulates normal N-ras p21 GTPase, but does not affect oncogenic mutants. Science. 1987; 238:542-545.
  6. Field J, Broek D, Kataoka T, et al. Guanine nucleotide activation of, and competition between, RAS proteins from Saccharomyces cerevisiae. Mol Cell Biol. 1987;7:2128-2133.
  7. Gutierrez L, Magee AI, Marshall CJ, et al. Post-translational processing of p21ras is two-step and involves carboxyl-methylation and carboxy-terminal proteolysis. EMBO J. 1989;8:1093-1098.
  8. Casey PJ, Solski PA, Der CJ, et al. p21ras is modified by a farnesyl isoprenoid. Proc Natl Acad Sci U S A. 1989;86:8323-8327.
  9. Medema RH, de Vries-Smits AM, van der Zon GC, et al. Ras activation by insulin and epidermal growth factor through enhanced exchange of
    guanine nucleotides on p21ras. Mol Cell Biol. 1993;13:155-162.
  10. Buday L, Downward J. Epidermal growth factor regulates p21ras through the formation of a complex of receptor, Grb2 adapter protein, and Sos nucleotide exchange factor. Cell. 1993;73:611-620.
  11. Moodie SA, Willumsen BM, Weber MJ, et al. Complexes of Ras. GTP with Raf-1 and mitogen-activated protein kinase kinase. Science. 1993;260: 1658-1661.
  12. Scheffzek K, Ahmadian MR, Kabsch W, et al. The Ras-RasGAP complex: structural basis for GTPase activation and its loss in oncogenic Ras mutants. Science. 1997;277:333-338.
  13. Parada LF, Tabin CJ, Shih C, et al. Human EJ bladder carcinoma oncogene is homologue of Harvey sarcoma virus ras gene. Nature. 1982;297:474-478.
  14. Santos E, Tronick SR, Aaronson SA, et al. T24 human bladder carcinoma oncogene is an activated form of the normal human homologue of BALB- and Harvey-MSV transforming genes. Nature. 1982;298:343-347.
  15. Der CJ, Krontiris TG, Cooper GM. Transforming genes of human bladder and lung carcinoma cell lines are homologous to the ras genes of Harvey and Kirsten sarcoma viruses. Proc Natl Acad Sci U S A. 1982;79:3637-3640.
  16. Shimizu K, Goldfarb M, Suard Y, et al. Three human transforming genes are related to the viral ras oncogenes. Proc Natl Acad Sci U S A. 1983;80:2112-2116.
  17. Wellcome Trust Sanger Institute. www.sanger.ac.uk/genetics/CGP/cosmic.
  18. Milburn MV, Tong L, deVos AM, et al. Molecular switch for signal transduction: structural differences between active and inactive forms of protooncogenic ras proteins. Science. 1990;247:939-945.
  19. Vetter IR, Wittinghofer A. The guanine nucleotide-binding switch in three dimensions. Science. 2001;294:1299-1304.
  20. Fasano O, Aldrich T, Tamanoi F, et al. Analysis of the transforming potential of the human H-ras gene by random mutagenesis. Proc Natl Acad Sci U S A. 1984;81:4008-4012.
  21. Gideon P, John J, Frech M, et al. Mutational and kinetic analyses of the GTPase-activating protein (GAP)-p21 interaction: the C-terminal domain of GAP is not sufficient for full activity. Mol Cell Biol. 1992;12:2050-2056.
  22. Feig LA, Cooper GM. Relationship among guanine nucleotide exchange, GTP hydrolysis, and transforming potential of mutated ras proteins. Mol Cell Biol. 1988;8:2472-2478.
  23. Quilliam LA, Zhong S, Rabun KM, et al. Biological and structural characterization of a Ras transforming mutation at the phenylalanine-156 residue, which is conserved in all members of the Ras superfamily. Proc Natl Acad Sci U S A. 1995;92:1272-1276.
  24. Seeburg PH, Colby WW, Capon DJ, et al. Biological properties of human c-Ha-ras1 genes mutated at codon 12. Nature. 1984;312:71-75.
  25. Recktenwald CV, Mendler S, Lichtenfels R, et al. Influence of Ki-ras-driven oncogenic transformation on the protein network of murine fibro­blasts. Proteomics. 2007;7:385-398.
  26. Horsch M, Recktenwald CV, Schadler S, et al. Overexpressed vs
    mutated Kras in murine fibroblasts: a molecular phenotyping study. Br J Cancer. 2009;100:656-662.
  27. Guerrero S, Casanova I, Farré L, et al. K-ras codon 12 mutation induces higher level of resistance to apoptosis and predisposition to anchorage-independent growth than codon 13 mutation or proto-oncogene overexpression. Cancer Res. 2000;60:6750-6756.
  28. Der CJ, Finkel T, Cooper GM. Biological and biochemical properties of human rasH genes mutated at codon 61. Cell. 1986;44:167-176.
  29. Lacal JC, Aaronson SA. Activation of ras p21 transforming properties associated with an increase in the release rate of bound guanine nucleotide. Mol Cell Biol. 1986;6:4214-4220.
  30. Sloan SR, Newcomb EW, Pellicer A. Neutron radiation can activate K-ras via a point mutation in codon 146 and induces a different spectrum of ras mutations than does gamma radiation. Mol Cell Biol. 1990;10:405-408.
  31. Feig LA, Cooper GM. Inhibition of NIH 3T3 cell proliferation by a mutant ras protein with preferential affinity for GDP. Mol Cell Biol. 1988;8:3235-3243.
  32. Smith G, Bounds R, Wolf H, et al. Activating K-Ras mutations outwith ‘hotspot’ codons in sporadic colorectal tumours – implications for personalised cancer medicine. Br J Cancer. 2010;102:693-703.
  33. Edkins S, O’Meara S, Parker A, et al. Recurrent KRAS codon 146 mutations in human colorectal cancer. Cancer Biol Ther. 2006;5:928-932.
  34. Jhawer M, Goel S, Wilson AJ, et al. PIK3CA mutation/PTEN expression status predicts response of colon cancer cells to the epidermal growth factor receptor inhibitor cetuximab. Cancer Res. 2008;68:1953-1961.
  35. Zlobec I, Kovac M, Erzberger P, et al. Combined analysis of specific KRAS mutation, BRAF and microsatellite instability identifies prognostic subgroups of sporadic and hereditary colorectal cancer. Int J Cancer. 2010;127:2569-2575.
  36. Oliveira C, Westra JL, Arango D, et al. Distinct patterns of KRAS mutations in colorectal carcinomas according to germline mismatch repair defects and hMLH1 methylation status. Hum Mol Gen. 2004;13:2303-2311.
  37. Gajate P, Sastre J, Bando I, et al. Influence of KRAS p.G13D mutation in patients with metastatic colorectal cancer treated with cetuximab
    [published online ahead of print April 24, 2012]. Clin Colorectal Cancer.
  38. Modest DP, Reinacher-Schick A, Stintzing S, et al. Cetuximab-based or bevacizumab-based first-line treatment in patients with KRAS p.G13D-mutated metastatic colorectal cancer: a pooled analysis. Anticancer Drugs. 2012;23:666-673.
  39. Modest DP, Jung A, Moosmann N, et al. The influence of KRAS and BRAF mutations on the efficacy of cetuximab-based first-line therapy of metastatic colorectal cancer: an analysis of the AIO KRK-0104-trial. Int J Cancer. 2012;131:980-986.
  40. Peeters M, Douillard J, Van Cutsem E, et al. Mutant (MT) KRAS codon 12 and 13 alleles in patients (pts) with metastatic colorectal cancer (mCRC): assessment as prognostic and predictive biomarkers of response to panitumumab (pmab). J Clin Oncol. 2012;30(suppl). Abstract 3581.
  41. You M, Candrian U, Maronpot RR, et al. Activation of the Ki-ras protooncogene in spontaneously occurring and chemically induced lung tumors of the strain A mouse. Proc Natl Acad Sci U S A. 1989;86:3070-3074.
  42. Paik PK, Johnson ML, D’Angelo SP, et al. Driver mutations determine survival in smokers and never-smokers with stage IIIB/IV lung adenocarcinomas [published online ahead of print May 17, 2012]. Cancer.
  43. Misale S, Yaeger R, Hobor S, et al. Emergence of KRAS mutations and acquired resistance to anti-EGFR therapy in colorectal cancer. Nature. 2012;486:532-536.
  44. Boland CR, Goel A. Microsatellite instability in colorectal cancer. Gastroenterology. 2010;138:2073-2087.
  45. Popovici V, Budinska E, Tejpar S, et al. Identification of a poor-prognosis BRAF-mutant-like population of patients with colon cancer. J Clin Oncol. 2012;30:1288-1295.
  46. Popovici VC, Budinska E, Tejpar S, et al. Molecular and clinicopathologic evidence of heterogeneity in KRAS-mutant colon cancers. J Clin Oncol. 2012;30(suppl). Abstract 3575.
  47. Zhao L, Vogt PK. Hot-spot mutations in p110alpha of phosphatidyl­inositol 3-kinase (pI3K): differential interactions with the regulatory subunit p85 and with RAS. Cell Cycle. 2010;9:596-600.
  48. Whitehall VL, Rickman C, Bond CE, et al. Oncogenic PIK3CA mutations in colorectal cancers and polyps. Int J Cancer. 2012;131:813-820.
  49. Kopetz S, Desai J, Chan E, et al. PLX4032 in metastatic colorectal cancer patients with mutant BRAF tumors. J Clin Oncol. 2010;28(suppl):15s. Abstract 3534.
  50. Kato S, Han SY, Liu W, et al. Understanding the function-structure and function-mutation relationships of p53 tumor suppressor protein by high-resolution missense mutation analysis. Proc Natl Acad Sci U S A. 2003;100:8424-8429.
  51. Warren RS, Atreya CE, Niedzwiecki D, et al. A novel interaction of genotype, gender, and adjuvant treatment in survival after resection of stage III colon cancer: results of CALGB 89803. J Clin Oncol. 2012;30(suppl). Abstract 3517.
  52. Sauna ZE, Kimchi-Sarfaty C. Understanding the contribution of synonymous mutations to human disease. Nat Rev Genet. 2011;12:683-691.
The Last Word - December 21, 2012

Companion Diagnostics and the Paradoxical Return of the Blockbuster Drug

In my article in the inaugural issue of Personalized Medicine in Oncology, I got a bit dogmatic in proclaiming that personalized medicine (PM) drug treatment selectivity spelled the end of population-based medicine, and with it, the blockbuster drug. This is true only insofar as we define blockbuster drug from within [ Read More ]

Uncategorized - December 20, 2012

Institute of Medicine Report: Best Care at Lower Cost: The Path to Continuously Learning Health Care in America. Highlights for the Cancer Community

On September 6, 2012, the Institute of Medicine (IOM) released the report Best Care at Lower Cost: The Path to Continuously Learning Health Care in America.1 The aim of this report, authored by the Committee on the Learning Health Care System in America, was to identify how the effectiveness and [ Read More ]