OUP user menu

KRAS Gene Mutation in Colorectal Cancer Is Correlated With Increased Proliferation and Spontaneous Apoptosis

Xiuli Liu MD, PhD, Maureen Jakubowski MT, Jennifer L. Hunt MD
DOI: http://dx.doi.org/10.1309/AJCP7FO2VAXIVSTP 245-252 First published online: 1 February 2011


KRAS mutation occurs in 30% to 50% of colorectal cancers (CRCs) and has been suggested to be associated with proliferation and decreased apoptosis. In this study, we analyzed KRAS in 198 CRCs and compared the clinicopathologic variables between KRAS-mutated and wild-type CRCs. Also, a subset of 90 and 66 CRCs from this cohort underwent microsatellite instability testing and histomorphologic evaluation, and the frequency of microsatellite instability-high (MSI-H) and histomorphologic variables were compared between KRAS-mutated and wild-type CRCs. Clinicopathologic features (age, sex, and tumor site, depth, size, grade, and metastasis) were not different between KRAS-mutated and wild-type CRCs. Compared with wild-type KRAS CRCs, KRAS-mutated CRCs had a lower frequency of MSI-H (15% vs 42%; P = .015), a higher chance of having brisk mitosis (77% vs 43%, P = .022) and apoptosis (77% vs 28%; P = .00012), and a greater mean of mitotic figures (P = .0002) and apoptotic cells (P = .0008). KRAS mutation was associated with higher tumor cell turnover.

Key Words:
  • Colorectal cancer
  • KRAS
  • Proliferation
  • Apoptosis
  • Microsatellite instability high
  • Polymerase chain reaction
  • Cycling sequencing

Colorectal cancer (CRC) is the second leading cause of cancer-related death in the United States. The development of CRC is a multistep process characterized by accumulation of genetic alterations that have long been considered to occur in a stepwise process. Along the progression from normal colonic epithelial cells, small adenoma, advanced adenoma, and finally to carcinoma, the KRAS oncogene mutation has a role in a significant proportion of CRCs. KRAS has been reported to be mutated in about 30% of colorectal adenomas and 30% to 50% of CRCs.14

The KRAS gene encodes a 21-kDa small protein that is activated transiently as a response to extracellular stimuli or signals such as growth factors, cytokines, and hormones via cell surface receptors.5,6 On its activation, the KRAS protein also is capable of turning off the signaling pathway by catalyzing hydrolysis of guanosine triphosphates (GTP) to guanosine diphosphates. The most common KRAS mutations in codons 12 and 13 are activation mutations, leading to continuous activation of downstream pathways.5,6 The most frequently observed types of mutations in KRAS in all human cancers are G > A transition and G > T transversion. Although the precise molecular and cellular mechanisms that constitute the oncogenic effects of activating KRAS mutations remain incompletely understood, in vitro and animal studies showed that KRASV12 regulated genes involve cytokine signaling, cell adhesion, cell survival, proliferation, apoptosis, and colon development.58

Emerging data strongly suggest that KRAS mutations are predictive of resistance to epidermal growth factor receptor (EGFR) inhibitor treatment in CRCs.5,6 The mechanisms by which KRAS-mutated CRCs are resistant to EGFR inhibitors are not known, although some investigators have suggested that activation mutation of KRAS would replace the dependence of CRCs on increased signaling from upstream EGFR. As an alternative, CRCs KRAS mutations may have decreased potential to undergo apoptosis and imply a resistance to drugs and agents with therapeutic effects through apoptosis pathways.

A few studies have also reported that a KRAS mutation was more commonly seen in well- to moderately differentiated CRCs and in distally located tumors.9,10 Several reports suggested that KRAS mutation was associated with poor prognosis and advanced disease.1,3,10,11 The data regarding the association of KRAS mutation with microsatellite status have been mixed.9,12 Two small studies with relatively detailed morphologic analysis of colorectal adenoma and adenocarcinoma suggested that KRAS mutation is associated with diffuse proliferation and decreased apoptosis.13,14 In CRC, while KRAS mutation was frequently noted in mucinous carcinoma,11 others reported KRAS mutation was more often seen in nonmucinous carcinoma regardless of microsatellite instability (MSI) status.15

In this study, we performed KRAS mutational analysis on 217 CRCs using formalin-fixed, paraffin-embedded tissue by polymerase chain reaction (PCR) and cycle sequencing of the amplified PCR products. A subset of 90 CRCs also had microsatellite stability testing. A detailed analysis of the frequency of each mutation within codons 12 and 13 was performed. Clinicopathologic parameters were obtained, and a detailed histomorphologic analysis was performed. CRCs with mutated KRAS and wild-type KRAS were compared to identify any possible histologic features associated with mutated KRAS in CRCs.

Materials and Methods

Study Group

We included all 217 cases of CRC with KRAS mutation testing in our molecular diagnostic laboratory in the Department of Anatomic Pathology, Cleveland Clinic Foundation, Cleveland, OH, between May 2008 and January 2009. This study was approved by the institutional review board. Since June 2008 in the Cleveland Clinic Foundation, all clinical cases of CRC that are stage 3 or 4 have been tested for KRAS mutations.

PCR Amplification and Sequencing of KRAS

PCR amplification and sequencing of KRAS exon 2, targeting the mutation hotspots of codons 12 and 13, were performed as follows. Briefly, slides from each case were reviewed by a gastrointestinal pathologist (X.L.), and formalin-fixed paraffin-embedded tissue blocks containing the highest density of malignant cells were chosen. For each case, 8 unstained sections were cut and mounted on uncharged glass slides, and the area with the highest density of malignant cells was selected for microdissection. The microdissection targets were confirmed by direct microscopic visualization. Tumor DNA was extracted from the microdissected fragments. The KRAS gene was examined by PCR for a 263-base-pair product that includes the most common mutation sites (codons 12 and 13) using the following primer pairs (forward, 5′-GGTGAGTTTGTATTAAAAGGTACTGG; and reverse, 5′-TCCTGCACCAGTAATATGCA). PCR amplification was performed for 49 cycles, including a 15-second denaturation at 95°C, 15-second annealing at 55°C, and 15-second extension at 72°C following the initial 3-minute denaturation at 95°C. Cycle sequencing was performed for the forward and reverse strands, using the BigDye Terminator kit (Applied Biosystems, Foster City, CA) and the same forward and reverse primers, in separate reactions. The sequence was analyzed using the ABI 3730 (Applied Biosystems). The forward and reverse sequences were visually inspected to identify the presence of point mutations at codon 12 or 13 or other locations within the PCR product.

Microsatellite Stability Analysis

For the study, 90 CRCs were analyzed for high-level MSI (MSI-H) according to the criteria for the determination of MSI-H as previously described.16 Briefly, the microsatellite stability test was performed using a fluorescent PCR-based assay by amplifying 7 loci, including 5 mononucleotide repeats (BAT 25, BAT 26, NR-21, NR-24, and MONO-27) and 2 pentanucleotide repeats (PentaC and PentaD) on genomic DNA extracted from CRC tissue and a paired normal sample (MSI Analysis System, Promega, Madison, WI). The fluorescently labeled, amplified PCR products from CRC and normal samples were analyzed by capillary gel electrophoresis on the ABI 3100 (Applied Biosystems). By comparing the sizes of the PCR products from CRC and normal samples, the presence of MSI was determined by the appearance of new alleles in the CRC sample that were not present in the corresponding normal sample. The 2 loci of pentanucleotide repeats serve to confirm identity. CRCs containing 2 or more loci with MSI were classified as MSI-H.

Histomorphologic Review and Quantitative Assessment of Mitotic and Apoptotic Activity and Tumor-Infiltrating Lymphocytes

Tumor-containing slides from a set of 66 randomly selected, chemoradiation-naive, and surgically resected primary CRCs (including 40 CRCs with wild-type KRAS and 26 cases with mutated KRAS) used for KRAS mutation analysis were further reviewed histologically by one of us (X.L.) to assess the following morphologic features: tumor grade, the presence of dirty necrosis, the presence of any mucinous differentiation, the presence of prominent cribriform architecture (defined as easily noticeable under ×100 magnification), the presence of brisk mitotic activity (defined as ≥3 mitotic figures/high-power field [HPF]), the presence of brisk apoptotic activity (defined as apoptotic cells easily identifiable under ×200 magnification), and the presence of tumor-infiltrating lymphocytes (TILs; defined as ≥2 TILs/HPF).

Further detailed quantitative analysis of mitotic activity, apoptosis, and TILs was performed in 26 cases of KRAS-mutated CRCs and 40 cases of CRCs with wild-type KRAS. More specifically, mitotic figures were counted in every 100 consecutive malignant cells in 10 areas with the most mitotic activity for each tumor. Apoptotic cells were counted in every 100 consecutive malignant cells in 10 areas with the highest apoptotic activity. When apoptotic cells were present on the luminal surface or within the lumen of neoplastic glands, they were excluded, as were areas of extensive necrosis. Only typical apoptotic bodies were counted using the standard morphologic criteria: overall shrinkage and homogeneously dark basophilic nuclei, the presence of nuclear fragments, a sharply delineated cell border surrounded by empty space, and homogeneous eosinophilic cytoplasm. TILs were also counted in every 100 consecutive malignant cells in 10 areas with the most TILs. All assessments were performed on the tumor sections used for KRAS mutation analysis. All sections were examined by 1 observer to ensure internal consistency of interpretation. The mitotic activity, apoptotic cells, and TILs were expressed as mean counts per 100 consecutive malignant cells obtained from the 10 areas.

Statistical Analysis

The χ2 test or Fisher exact test was performed for categorical data. The Student t test was used to analyze continuous data. The results for mitotic activity, apoptotic cells, and TILs were expressed as mean ± SD. All P values were 2-sided, and statistical significance was set at a P value of .05 or less.


Clinicopathologic Features of CRCs in the Study Population

In this study, 217 CRC cases were included, including 109 in-house cases and 108 reference cases from outside facilities. Among them, 198 cases (91.2%) had successful amplification of the KRAS gene and constituted the study population. Of the 19 cases with failed amplification, testing was repeated once in 16, and amplification failed again in 16 of 16; most of these were considered to be due to previous chemoradiation therapy with resultant scarcity of tumor cells and inappropriate fixation of the tissue. The mean age of the patients was 59.8 years, and 49.0% were men Table 1. In the 113 cases with more detailed clinicopathologic data, 23 tumors were in the rectum, 43 were in the right colon, 27 were in the left colon, 12 were liver metastases, and 8 were lung or thoracic metastases. In 95 CRCs for which the depth of invasion was known, 6% (5/88), 6% (5/88), 69% (61/88), and 19% (17/88) of them invaded submucosa (pT1), muscularis propria (pT2), subserosa (pT3), and peritoneal or adjacent organ (pT4), respectively. Approximately 63% (59/93) of CRCs and 25.0% (31/124) had lymph node metastases and distal metastases, respectively, at the time of KRAS mutation testing. The high proportion of lymph node–positive and distant metastatic cases in this study population reflects the practice pattern of oncologists in which they routinely ordered KRAS mutation analysis when EGFR inhibitor therapy was considered (stage 3 or 4 tumor).

Mutational Types of KRAS in the Study Population

In this study, among the 198 cases with analyzable KRAS results, 65 tumors (32.8%) were found to harbor mutated KRAS gene. As shown in Table 2, the KRAS mutations were distributed between codon 12 (45/65 [69%]) and codon 13 (20/65 [31%]). The G > A transitions at nucleotides were the most frequent mutations observed in codons 12 and 13 (61.7%). Among the 65 tumors with mutated KRAS genes, 3 had unique mutation patterns. More specifically, 1 CRC showed only mutated nucleotide T at the first position in codon 12 without an accompanying normal allele, 1 contained only mutated nucleotide A at the second position in codon 13 without an accompanying normal allele, and 1 had mutated nucleotide T at the first and second positions in codon 12. The former 2 cases with mutated nucleotides without an accompanying normal allele may represent a homozygous mutation (C12GGT > TGT with amino acid change from Gly to Cys and C13GGC > GAC with amino acid change from Gly to Asp) or a single mutation in one allele and loss of heterozygosity of the other allele. The third case with 2 mutated nucleotides in codon 12 may represent a double mutation (cis position; GGT > TTT with amino acid change from Gly to Lys) or 2 different mutations in two alleles (trans position; GGT > TGT and GGT > GTT with amino acid changes from Gly to Cys and from Gly to Val, respectively).

View this table:
Table 1
View this table:
Table 2
View this table:
Table 3
View this table:
Table 4

KRAS Mutation Was Not Associated With Clinicopathologic Features of CRCs

Clinicopathologic features of CRCs with mutated KRAS and wild-type KRAS were compared in a subset of CRCs for which clinicopathologic data were available. As shown in Table 1 and Table 3, the KRAS mutation is not associated with age, sex, tumor location, depth of tumor invasion, lymph node metastasis, distant metastasis, tumor grade, tumor size, or angiolymphatic invasion.

KRAS Mutation Was Less Frequently Seen in MSI-H CRCs

Of the 198 CRC cases, 90 tumors (45.5%) were tested for microsatellite stability as well. Among these 90 CRCs tested for microsatellite status, 59 were microsatellite stable (66%) and 31 CRCs were MSI-H (34%). As shown in Table 4, only 13% (4/31) CRCs with mutated KRAS were MSI-H, in contrast with 37% (22/59) CRCs with wild-type KRAS (P = .015).

More KRAS-Mutated CRCs Demonstrated Brisk Mitotic and Apoptotic Activity

Although KRAS activation mutation has been shown to be associated with increased proliferation in cancer cell lines, its exact effect in human cancer is relatively unclear. In a few studies, KRAS mutation has been shown to be associated with a diffuse proliferation pattern, polypoid growth, and high cytologic grade in colorectal adenomas and early cancers.14 In a small study, a KRAS mutation was suggested to be associated with decreased apoptosis.13

Because there is lack of detailed histomorphologic analysis of CRCs with mutated KRAS in the literature, in this study, 40 CRCs with wild-type KRAS and 26 tumors with mutated KRAS were randomly chosen for detailed histomorphologic review by a gastrointestinal pathologist (X.L.) who was blinded to the KRAS mutation results. These 66 tumors were chemoradiation-naive, surgically resected, primary CRCs. Tumors were given a single grade of differentiation (well, moderate, or poor) based on the criteria of Jass et al17 and Greenson et al18 with minor modification. The worst grade of tumor seen was used for the overall grade, unless the worst area was smaller than 10% of the tumor volume and at the advancing margin of the tumor. In addition, morphologic features assessed included the presence of dirty necrosis (>10% of tumor volume, infarcted tumor areas not included), mucinous differentiation (any), cribriform architecture of glands (easily noticeable at ×100 magnification), the presence of brisk mitotic activity (>3 mitotic figures/HPF), and the presence of brisk apoptotic activity (easily identifiable at ×200 magnification). Tumors were also evaluated for TILs as previously described,18 and the tumor was considered to be positive for the presence of TILs if more than 2 lymphocytes per HPF were identified.

As shown in Table 5, the frequency of the presence of dirty necrosis, the presence of poor differentiation, the presence of mucinous differentiation, the presence of TILs, and the presence of cribriform architecture were not different between the 40 wild-type KRAS-containing and the 26 KRAS-mutated CRCs (presence of dirty necrosis, 11 [28%] vs 9 [35%], P = .364; poor tumor differentiation, 11 [28%] vs 10 [38%], P = .252; mucinous differentiation, 20 [50%] vs 9 [35%], P = .281; cribriform architecture, 30 [75%] vs 21 [81%], P = .407; and presence of TILs, 10 [25%] vs 8 [31%], P = .122). However, 77% (20/26) of KRAS-mutated CRCs had brisk mitotic activity compared with 43% (17/40) of CRCs with wild-type KRAS (P = .006). In addition, 77% of the KRAS-mutated CRCs (20/26) demonstrated brisk apoptotic activity compared with only 28% of CRCs (11/40) with wild-type KRAS (P = .00012).

KRAS-Mutated CRCs Were Mitotically and Apoptotically Active by Quantitative Assessment

To confirm our qualitative results, a quantitative assessment of mitotic activity, apoptotic cells, and TILs were performed in that subset of CRCs (40 CRCs with wild-type KRAS and 26 cases with mutated KRAS) using the methods as described in the “Materials and Methods” section. As shown in Figure 1, the quantitative assessment revealed results similar to those obtained by the overall assessment method. The mean ± SD numbers of mitotic figures (in 100 consecutive malignant cells) were 3.15 ± 1.05 (n = 26) in KRAS-mutated CRCs compared with 1.83 ± 1.47 (n = 40) in CRCs with wild-type KRAS (P = .0002). The mean ± SD number of apoptotic cells (in 100 consecutive malignant cells) were 10.9 ± 4.05 (n = 26) in KRAS-mutated CRCs compared with 7.15 ± 4.34 (n = 40) in CRCs with wild-type KRAS (P = .0008). The number of TILs (in 100 consecutive malignant cells) in KRAS-mutated and KRAS wild-type CRCs were not significantly different (2.03 ± 2.95 vs 3.61 ± 6.01, n = 26 and n = 40, respectively, for KRAS-mutated and wild-type CRCs, P = .22).


KRAS is an oncogene that encodes a 21-kDa small protein with GTPase activity.5,6 Oncogenic KRAS alleles that harbor activation mutations produce a protein product with enzymatically impaired GTPase function or that is refractory to GTPase activating proteins. This leads to elevated steady-state levels of RAS-GTP and causes prolonged effector pathway stimulation.5,6 Previous intensive exploration in the field has demonstrated that the RAS-mediated pathway regulates different cellular responses such as proliferation, transformation, differentiation, senescence, and apoptosis.5,6,19,20 Although KRAS mutation has been described in a significant number of human CRCs, detection of the mutation was not known to have clinical relevance and was largely performed in research and clinical trial settings. Emerging data strongly suggest, however, that KRAS mutations in CRCs have a predictive role in determining resistance to EGFR inhibitor (cetuximab and panitumumab) treatment in patients in whom previous conventional chemotherapy regimens failed.21,22 After the initial announcements of these results, KRAS mutational analysis quickly became an important molecular test for patients with CRC who were being considered for EGFR inhibitor treatment. This dramatic shift in practice can be seen in our study in which 217 requests for KRAS mutation analyses have been requested for patients with CRC during the past 9-month period. Currently, only a handful of molecular diagnostic laboratories are offering KRAS mutational analysis, but more laboratories are expected to validate this assay.

View this table:
Table 5

In our study population, the mutational rate for KRAS at codons 12 and 13 was 32.8%, which was similar to the previously reported mutational rates of 30% and 37% in 3 large studies.1,3,23 The mutation at codon 12 accounted for approximately 69% of cases. A G > A transition was the most frequent mutation (61.7% of the total mutations detected). These findings are in line with previously published results.10 In our study, we observed a total of 20 mutations at codon 13, and 1 of them was a silent mutation (1 case). It is interesting that 2 CRCs (3%) in our study showed a single mutant peak on the sequencing electrophoretogram, with no corresponding normal allele. One case showed the mutation at the first position in codon 12, and the other was in the second position in codon 13. The possible explanation is a homozygous mutation with both alleles harboring identical mutations or a single mutant allele with corresponding loss of heterozygosity for the second allele. A third case had mutated nucleotides T at the first and second positions of codon 12, which could represent a double mutation (cis, GGT > TTT in 1 allele with amino acid Gly > Lys) or 2 independent mutations, 1 on each allele (trans, GGT > TGT and GGT > GTT with amino acids Gly > Cys and Gly > Val). Owing to the small number of cases, the biologic significance of these homozygous mutations, double mutation, or single mutation with loss of heterozygosity remains unclear. However, all 3 cases showed poorly differentiated morphologic features (data not shown), and histomorphologic analysis of the tumor with double mutation (cis) or 2 independent mutations (trans) revealed that this tumor was mitotically and apoptotically active.

Figure 1

Correlation of KRAS mutation with mitotic activity, spontaneous apoptosis, and tumor-infiltrating lymphocytes (TILs) in colorectal carcinomas (CRCs). A, Mitotic figures in KRAS-mutated and KRAS wild-type CRCs. B, Apoptotic cells in KRAS mutated and KRAS wild-type CRCs. C, TILs in KRAS-mutated CRCs and KRAS wild-type CRCs. Mitotic figures, apoptotic cells, and TILs were counted in 100 consecutive malignant cells in 10 fields as described in the “Materials and Methods” section in 26 cases of KRAS-mutated CRCs and 40 cases of KRAS wild-type CRCs. The results are expressed as means, as indicated by the numbers on top of the bars.

Our study did not identify any association of KRAS mutation with the following clinicopathologic variables: age, sex, tumor site, depth of tumor invasion, histologic grade, distant metastasis, vascular invasion, and lymphocytic response. Although owing to the relative lack of earlier stage cases the study population may be biased to allow meaningful correlation of KRAS mutation with clinicopathologic variables including nodal status and distant metastases, our current findings were in line with a previous large study of 3,439 cases of CRC collected in multiple centers and many countries.1

We performed a detailed histomorphologic analysis for KRAS-mutated CRCs to identify possible histologic features associated with mutated KRAS in CRCs. For mitotic activity and apoptosis, we did not perform ancillary tests such as immunohistochemical analysis for Ki-67 and the terminal deoxynucleotidyl transferase (TdT)-mediated deoxyuridine triphosphate (dUTP)-biotin end-labeling assay because we wanted to use a simple method that would be applicable in routine surgical pathology practice examining H&E-stained sections. Our results showed that KRAS-mutated CRCs had increased mitotic and apoptotic activity compared with wild-type KRAS tumors. The higher mitotic activity in mutated KRAS CRCs observed in our study is consistent with results previously reported.8,11,14 Brisk apoptotic activity in mutated KRAS CRCs was also observed in our study, consistent with a previous report of an association of KRAS G > A transitions with increased apoptosis,24 but contrasting with the findings reported by Ward et al,13 who found no association. The discrepancy may be due to several factors: The study by Ward et al13 was relatively small, used a different method to measure apoptotic activity (TdT-mediated dUTP-biotin end-labeling), and included more tumors in stages 1 and 2, and the KRAS mutation analysis only detected codon 12 mutations. However, the clinician’s KRAS test-ordering pattern in which KRAS tests were ordered only for patients with stage 3 or 4 CRCs during the study period in our study resulted in a relative lack of earlier stage cases in this retrospective study and could have led to bias.

The exact mechanism by which an activating KRAS mutation leads to relatively brisk spontaneous apoptosis in CRCs remains unclear. It was initially thought that there was a progressive inhibition of apoptosis during colorectal tumorigenesis,25 and KRAS mutation appeared to further inhibit apoptosis in CRCs because CRCs from compound KRASV12G/APC+/1638N mice showed reduced apoptosis relative to tumors arising from APC+/1638N transgenic mice.26 However, emerging data have suggested that KRAS mutation (KRAS13D) promotes the induction of apoptosis in ReovirusT3D-infected or oxaliplatin-treated tumor cells by sensitizing these cells to activation of the intrinsic apoptosis cascade27 and increasing sensitivity of colon cancer cells to 5-fluorouracil-induced apoptosis.28 Additional reports have also showed that oncogenic KRAS mutations mediated apoptosis through NORE1, RASSF1, and other RASSF proteins5,29 and that KRAS mutations induced apoptosis via a p53-independent pathway when NF-κB activation was inhibited.30 Further in vitro experiments in a lung cell line showed that mutant KRASV12 increased COX-2, peroxides, and DNA damage, which led to activation of an intrinsic apoptotic cascade.31 However, the intermediate steps between oncogenic KRAS mutation, COX-2 activation/up-regulation, DNA damage, NORE1/RASSF1 and other RASSF proteins, and final apoptosis of cancer cells are largely unexplored in CRCs. Further studies in these fields will yield more mechanistic insights into KRAS mutation-mediated spontaneous apoptosis in CRC.

We also assessed KRAS in relation to MSI and found that KRAS mutations were less frequently seen in CRCs with MSI-H. This finding is consistent with previously reported results.12 Most of the MSI-H CRCs in our study had loss of hMLH1 protein (but no other mismatch repair [MMR] proteins) shown by immunohistochemical analysis and, therefore, were most likely sporadic tumors (data not shown). Our results, along with results reported by Oliveira et al12 strongly suggest that sporadic MSI-H CRCs have a lower frequency of KRAS mutation compared with sporadic microsatellite-stable CRCs and MSI-H CRCs arising in the setting of hereditary nonpolyposis CRC. Furthermore, in a small study, Zhao et al32 reported an association of sequence alterations in hMLH1 with sporadic KRAS mutations in sporadic MSI-H CRCs. Therefore, the emerging data appear to support that mutations in MMR genes are associated with KRAS mutation in hereditary nonpolyposis CRC and sporadic MSI-H CRCs. It is interesting that of 31 MSI-H CRCs in the current study, 4 had KRAS mutation and 3 of them (75%) were C13GGC > GAC with amino acid change from Gly to Asp, therefore suggesting a possible association of C13GGC > GAC with MSI-H in CRCs because the frequency of C13GGC > GAC in the current entire cohort was only 28% (18/65). However, large studies are needed to confirm this possible association. The interplay of promoter hypermethylation of hMLH1 and other genes, MMR gene mutation, and KRAS mutation (including mutational location and type) is complex, and detailed analysis of larger studies will be essential to enhance our understanding of early predominant genetic alterations during tumorigenesis of sporadic CRCs.

Our study showed an excellent success rate of KRAS mutation analysis in routinely processed clinical surgical pathology material using the traditional cycle sequencing reaction. Approximately 30% of CRCs had KRAS mutation at codons 12 and 13. KRAS mutation was less frequently seen in MSI-H CRCs. Furthermore, our study showed that KRAS mutation in CRCs was associated with higher mitotic activity and brisker spontaneous apoptosis. Additional work in larger series is needed to determine whether these histologic parameters may be a screening tool for triaging cases for up-front KRAS mutational testing. It will also be important to determine whether these parameters have biologic significance in the background of KRAS mutations and MSI in larger series including early- and late-stage CRCs.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
View Abstract