Relationship of Ki-ras Mutations in Colon Cancers to Tumor Location, Stage, and Survival: A Population-based Study¹

Oncogenic mutation in Ki-ras is one of the most common genetic alterations in colorectal cancer, with alterations in codons 12 and 13 of Ki-ras accounting for ∼85% of these genetic changes in such tumors (1). Some previous studies have demonstrated significant results between Ki-ras mutations in general or specific types of Ki-ras mutations and tumor stage, survival, and/or other clinical variables, whereas others have not (reviewed in Ref. 2). Possible explanations for these discrepant results include non-population-based samples and insufficient power to demonstrate relationships, especially with respect to specific types of mutations. We therefore studied the relationship of codons 12 and 13 Ki-ras mutations to survival, stage, and other clinical variables in a large population-based study of 1413 individuals with colon cancer.

Study participants were black, white, or Hispanic and were from either the Kaiser Permanente Medical Care Program of Northern California or an eight-county area in Utah (Davis, Salt Lake, Utah, Weber, Wasatch,Tooele, Morgan, and Summit counties). Ki-ras mutation results from the Utah sample were reported in a previous study of the relationship between microsatellite instability and Ki-rasand p53 alterations.³ Study participants from the Minnesota center were not included in these analyses because these participants had to give consent again at the time of tissue collection, resulting in extraction of DNA from only 480 cases of the 1290 diagnosed in the area and an unrepresentative sample at the population level. In Utah and the Kaiser Permanente Medical Care Program, we were able to make DNA for 95.6 and 81.1%, respectively, of all cases diagnosed in the area, making this a truly population-based study from these geographic locations. Eligibility criteria for cases included diagnosis with first primary incident colon cancer (ICD-O, Ed. 2: 18.0, 18.2, to 18.9)between October 1, 1991 and September 30, 1994, ages between 30 and 79 years at time of diagnosis, and mentally competent to complete the interview. Cases with cancers of the rectosigmoid junction or rectum(defined as the first 15 cm from the anal opening) or with known familial adenomatous polyposis, ulcerative colitis, or Crohn’s disease were not eligible. All cases were adenocarcinomas or carcinomas.

Information on age at time of diagnosis, sex, tumor site, and tumor stage were available from the Northern California Tumor Registry, the Sacramento Tumor Registry, and the Utah Cancer Registry. These registries are members of the SEER⁴program. Proximal tumors were defined as cecum through transverse colon; distal tumors were those in the splenic flexure, descending, and sigmoid colon. Staging data were summarized as local, regional, or distant, depending on extent of disease and node and other organ involvement using the SEER summary stage codes (3). Vital status, date of death, primary cause of death, and two contributing causes of death were obtained from local tumor registries using death certificate information. Active follow-up of people diagnosed with cancer is done through the cancer registries on a continuous basis. Vital status as of December 30, 1998 was obtained for all study participants. For individuals whose vital status or cause of death could not be determined through local tumor registries, National Death Index tapes were used. Months of survival were calculated by subtracting the date of last contact or death from the date of diagnosis. Deaths from any cause as well as deaths attributed to colon cancer were assessed.

Colon cancer tissue was microdissected and DNA was extracted from formalin-fixed paraffin-embedded tissue blocks as described previously (4). Codons 12 and 13 of the Ki-ras gene were evaluated for mutations. Exon 1 of Ki-ras was amplified as described previously (5), except that primers were tailed with universal primer and reverse primer for sequencing. PCR products were sequenced using prism Big Dye terminators and cycle sequencing with Taq FS DNA polymerase. DNA sequence was collected and analyzed on an ABI prism 377 automated DNA sequencer.

The distribution of specific types of mutations in tumors by patient characteristics was assessed. Differences in the age distribution, sex,tumor site, and tumor stage were determined for tumors with and without Ki-ras mutations were determined usingχ ² tests. Patient and tumor characteristics were also compared for specific types of Ki-ras mutations. Types of mutations assessed were mutations in codon 12, mutations in codon 13, transversions, transitions, and specific bp changes. Transitions and transversions are different classes of bp substitutions that may reflect differences in carcinogen exposure and/or genetic pathways and may therefore define different classes of colon cancers with respect to biological behavior and other clinical variables (6). Proportional hazard models were used to estimate the impact of Ki-ras mutations on hazard of dying during the follow-up period.

Ki-ras mutations were identified in 31.8% of tumors(Table 1). Of these, 77.9% were in codon 12 and 22.1% were in codon 13; 62.5%were transitions, and 37.5% were transversions. Mutations were found in the first and second bases of codons 12 (1G and 2G) and 13 (4G and 5G). The most common types of mutation were 2G→A (Gly→Asp) in codon 12 (31.1% of mutations), 2G→T (Gly→Val) in codon 12 (21.2% of mutations), 5G→A (Gly→Asp) in codon 13 (20.8% of mutations), and 1G→T (Gly→Cys) in codon 12 (9.5% of mutations). Chromatograms demonstrating these mutations are shown in Fig. 1.

Ki-ras mutations were significantly more common in proximal tumors than distal tumors (36.0% versus 26.8%; P < 0.01; Table 2). Tumors diagnosed at advanced disease stage (distant and regional)were significantly more likely to harbor Ki-ras mutations than those diagnosed at a local stage (35.4% and 33.6%versus 27.2%; P < 0.03). No statistically significant relationships between overall Ki-ras mutations in tumors and age at time of diagnosis or gender were detected (Table 2).

Assessment of these same factors with specific types of mutations(Table 3) showed that men were significantly more likely to have transition mutations or 2G→A mutations. All Ki-ras mutations were significantly more likely to occur in tumors in the proximal colon,except for mutations in codon 13, when assessed as a group and 5G→A mutations specifically. Significant differences in tumor stage were detected for codon 12 mutations when taken as a whole, as those with a codon 12 mutation were more likely to have a more advanced tumor at the time of diagnosis.

After adjusting for age at time of diagnosis and tumor stage,individuals with Ki-ras mutations in codon 13 in general and 5G→A mutations specifically had a 40% greater likelihood of dying of colon cancer during the follow-up period than individuals without Ki-ras mutations (Table 4), although this was of borderline significance (P =0.08). No increase in overall mortality was observed.

This is the first population-based study and the largest non-meta analysis of Ki-ras mutations in colon cancer. The overall detection rate for Ki-ras mutations was 31.8%. This is consistent with previous studies that, with rare exceptions (7), have identified Ki-ras gene mutations in approximately 30–40% of colon cancers, and the relative proportions of specific types of Ki-ras gene mutations we identified(Table 1) are very similar to previous studies as well (2, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16). Ki-ras gene mutations were significantly more common in proximal tumors (36%) than distal tumors (26.8%), and significant differences with respect to tumor site were seen for all types of codon 12 mutations but not for codon 13 mutations. A significant relationship between Ki-ras gene mutations and proximal tumor location was seen in one previous study (17). Other studies have not shown such a relationship,but it is possible that smaller studies would have lacked sufficient power to demonstrate the relatively small difference we observed. Even in this study of 1413 cases of colon cancer, we had limited power to evaluate specific associations with specific mutations and more advanced tumors.

As in previous studies (2, 13, 18), there was no significant relationship between overall Ki-ras mutations and gender, although men were significantly more likely to have transition mutations and the 2G→A mutation (Tables 2, 3). It is possible that some lifestyle factor to which men and women are differentially exposed, such as patterns of alcohol or tobacco use,dietary patterns, or hormone replacement therapy, increases or decreases the likelihood of these mutations. Ki-rasmutations, specifically codon 12 mutations, were significantly more common in advanced stage tumors (Table 3). Most previous studies have not shown a relationship between advanced stage and Ki-rasmutations (2, 13, 14, 19, 20). The differences in mutation frequency between early and advanced stage tumors were relatively small in the current study and might not have been significant in studies of lesser power.

In agreement with some previous studies (7, 8, 10, 13, 14, 20), Ki-ras mutations in general were not associated with increased cancer-related mortality (Table 4). Mutations in codon 13, however, specifically the 5G→A mutation, were associated with a 40% greater likelihood of dying, although this difference was of borderline statistical significance (P = 0.08) after adjusting for age and disease stage. A smaller increase in mortality was observed for transition mutations. A previous study also demonstrated significantly poorer survival and disease-free interval associated with the 5G→A mutation (15). That study also described a similar impact on mortality from the 1G→T mutation, but our study does not confirm that association. Other previously reported findings that our study does not confirm include an indolent clinical course associated with 5G→A and 2G→T mutations (9), a poor prognosis associated with the 2G→T mutation and G→T mutations in general (2), and restriction of G→A transition mutations to Dukes’ B tumors and G→T and G→C transversion mutations to Dukes’ C tumors (21). Our study also does not confirm a previous study (15), which showed a poor prognosis associated with Ki-ras mutations in stage 2 tumors, although it should be noted that the SEER staging system used in our study is not directly comparable with that used in the previous study.

It should be noted that analysis of other genetic factors could affect the strength of the relationships we observed. Microsatellite instability, for example, is more common in proximal tumors (22), as are Ki-ras mutations, and recent studies have reported an improved prognosis in colorectal tumors with instability (23). Because we have shown an inverse relationship between microsatellite instability and Ki-rasgene mutations,³ it is possible that adjusting for microsatellite instability would alter the association we identified between a specific type of Ki-ras mutation and a poorer prognosis. This is, perhaps, less likely given the fact that the relationship between Ki-ras and prognosis was only seen with a specific mutation, rather than Ki-ras mutations overall. With respect to other genetic factors, several studies (14) have indicated a poorer prognosis associated with p53 mutations in colon cancer, and a multivariate analysis that includes p53 mutations could also weaken the relationship between Ki-ras and prognosis. It is also possible, however, that some relationships could be strengthened if colon tumors are stratified by other genetic factors. For example,although a recent study found no relationship between Ki-rasmutations and survival overall, tumors with both p53mutations and ras mutations were associated with an adverse prognosis (14). Future studies that evaluate additional genetic factors will be important in refining our understanding of the relationship of Ki-ras gene mutations to other clinical and pathological variables.

In conclusion, our large population-based study demonstrated small but statistically significant relationships between codon 12 Ki-ras mutations and tumor stage, proximal location, and male gender. In addition, the codon 13 Gly→Asp mutation was associated with a 40% increased likelihood of death attributable to colon cancer, although this was of borderline statistical significance. No association between Ki-ras mutations overall and prognosis was seen, and numerous previously reported associations between various Ki-ras gene mutations and tumor stage and/or prognosis were not confirmed by this study.

We acknowledge the contributions and support of Dr. Bette Caan,Judy Morse, Sandra Edwards, and Leslie Palmer to the data collection and tissue processing efforts, Khe Ni Ma for assistance with data analyses, and the sequencing core facility, Melanie Nichols, and Kristen Gruenthal for assistance with genetic analyses.