- Email: [email protected]
Genetic and molecular diversity of colon cancer hepatic metastases
Genetic and molecular diversity of colon cancer hepatic metastases Craig A. Messick, MD,a Julian Sanchez, MD,a Kathryn L. DeJulius, MS,a,b James M. Church, MB, ChB,a,b and Matthew F. Kalady, MD,a,b Cleveland, OH
Background. Colon cancer arises through distinct molecular pathways resulting in diverse tumor populations demonstrated by differences in microsatellite instability (MSI), CpG island methylator phenotype (CIMP), and mutations in oncogenes KRAS and BRAF. Although these molecular differences are well-described for primary neoplasms, the molecular nature of hepatic metastases is not wellcharacterized. This study seeks to describe molecular characteristics of colon cancer hepatic metastases in terms of oncogenic pathway. Methods. Tumor DNA was isolated from fresh frozen hepatic metastases from colon cancer and analyzed for MSI by polymerase chain reaction (PCR)-based microsatellite analysis and for CIMP using MethyLight quantitative PCR. KRAS and BRAF oncogenes were analyzed for DNA mutations. Metastases were classified by their molecular and genetic features. Unfortunately, tissue from the primary neoplasms from these patients were not available Results. Thirty patients with liver metastases from colon cancer were studied. Molecular analysis revealed 10% (3/30) were MSI-H, 10% (3/30) were CIMP positive, 33% (10/30) had KRAS mutations, and none had BRAF mutations. Literature describing primary colon cancers reports an incidence of approximately 20% MSI-H, 20% CIMP-positive, 35% KRAS mutants, and 17% BRAF mutants. Conclusion. Hepatic metastases from colon cancer, like primary colon adenocarcinomas, show genetic and molecular diversity. Furthermore, hepatic metastases may have a different incidence of MSI and methylation compared with primary neoplasms. These differences could impact treatment decisions. (Surgery 2009;146:227-31.) From the Department of Colorectal Surgery,a and the Department of Cancer Biology,b Digestive Disease Institute, Cleveland Clinic, Cleveland, OH
COLORECTAL ONCOGENESIS is a complex sequence of events resulting in the transformation of normal colonic epithelium into malignant tumor cells whose unchecked growth leads to continued progression and metastasis. The current understanding of colorectal oncogenesis is based on the concept of 3 or more molecular mechanisms underlying the change from benign to malignant disease: chromosomal instability (CIN), DNA microsatellite instability (MSI), and hypermethylation of specific tumor Presented at the Academic Surgical Congress. The authors make the following disclosures: James M. Church, MB, ChB: Myriad Genetics, speaker, honorarium; Salix, consultant, honorarium; Cleveland Clinic/Cologene, employee, salary. Matthew F. Kalady, MD: Genzyme, speaker’s bureau, honorarium. Accepted for publication June 4, 2009. Reprint requests: Craig A. Messick, MD, Department of Colorectal Surgery, The Cleveland Clinic, 9500 Euclid Avenue, A30, Cleveland, OH 44106. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2009 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2009.06.003
suppressor genes resulting in the CpG island methylator phenotype (CIMP), as well as others.1-6 Importantly, these distinct pathways defined by molecular and genetic features correspond with distinct clinical phenotypes and are also associated with differing responses to particular chemotherapy and clinical outcomes.5,6 Treatment decisions may be made based on the characteristics of the neoplasms. For example, MSIhigh (MSI-H) colon cancers treated by resection alone enjoy a survival advantage compared with microsatellite stable (MSS) neoplasms.7 MSI-H neoplasms, however, do not respond as well to 5-fluourouracil--based chemotherapy and stage III MSI-H colon cancer patients seem to do worse when treated with 5-fluourouracil compared with MSS patients.7 Similarly, CIMP status has been shown to affect response to chemotherapy and overall survival.8-10 Furthermore, metastatic neoplasms that display mutations in the oncogenes KRAS and BRAF are resistant to adjuvant treatment with cetuximab.11-13 Despite directed therapy based on molecular and genetic characteristics of the primary neoplasm, SURGERY 227
228 Messick et al
best responses are still only approximately 15--40% for stage III disease8,12 and median survival for stage IV disease is variable at 10--22 months.8,14 Although patient-specific factors may contribute to chemotherapy resistance, the rates of low treatment success remain incompletely explained. One hypothesis is that colon cancer metastases are genetically heterogeneous and possibly more than, or in different ways from, the primary neoplasms. Thus, a standard approach to treating metastatic disease using the same chemotherapy regimen for all patients is unlikely to be uniformly effective. If chemotherapy could be tailored to the genetics of the metastases, better response rates may be achieved. Although described for primary colon cancers, molecular characteristics of metastases are not well described. This study describes the molecular and genetic diversity of colon adenocarcinoma hepatic metastases in the context of how it may potentially influence clinical decision making. METHODS Tissue collection and patient information. This study was approved by the Cleveland Clinic Institutional Review Board. Tissue from resected colon cancer hepatic metastases was obtained from a prospectively collected and maintained frozen tissue biobank. All available colon cancer hepatic metastases were included. A gastrointestinal pathologist confirmed the diagnosis of metastatic adenocarcinoma by review of hematoxylin and eosin--stained slides. Only tissues sections containing at least 60% adenocarcinoma were included. Tumor information and patient demographics were recorded. DNA isolation. Frozen tissue blocks were sectioned using a cryostat into seven to ten 10-mm sections then subjected to an overnight proteinase K digestion. Using the Qiagen BioRobot EZ1 workstation, genomic DNA was isolated with the Qiagen EZ1 DNA Tissue Kit as per the manufacturer’s instructions (Qiagen, Valencia, CA). Genetic analyses. Microsatellite analysis: Isolated genomic DNA from hepatic metastases was tested for MSI using a panel of 10 mononucleotide, dinucleotide, and penta-mono-tetra compound markers as previously described.5,6 Additional, normal, nonmalignant, adjacent hepatic tissue was used as a source of genomic DNA to which tumor microsatellite status was compared for each individual patient. Briefly, standard polymerase chain reaction (PCR) techniques were employed for DNA amplification and then both tumor and normal adjacent tumor DNA were sequenced. All samples were read and scored for their microsatellite
Surgery August 2009
status by 2 independent evaluators. Neoplasms with 3 (30%) or more of the 10 markers exhibiting instability were considered to be MSI-H; 1 or 2 markers exhibiting instability were considered to be MSI-low (MSI-L); those with no markers exhibiting instability were considered to be MSS. For this study, neoplasms that were MSI-L and MSS were grouped together and designated MSS.15 CIMP analysis: CIMP status was determined as described previously by Eads et al,16 and as reported by our group.5,6 Briefly, genomic DNA underwent a sodium bisulfite conversion and universally methylated human DNA was used as positive control. All sodium bisulfite conversion reactions were repeated twice to ensure validity of the results. MethyLight quantitative PCR was then performed on a panel of 5 markers (CACNA1G, IGF2, NEUROG1, RUNX3, and SOCS1). The percent of methylated reference was then calculated as described previously.17 A given marker was considered positive if the percent of methylated reference was greater than 10 and samples with 3 or more positive markers were considered CIMP positive (CIMP+) and those with 2 or fewer CIMP negative (CIMP-N). Mutation analysis: KRAS and BRAF mutations were evaluated by PCR sequencing analysis. Genomic DNA was amplified at 2 KRAS exons (exon 3 codon 61, exon 2 codon 12, and exon 2 codon 13) and 2 BRAF exons (exon 11 codon 468, exon 15 codon 596, and exon 15 codon 599) by standard PCR. All PCR products were purified using the Qiagen QiaQuick Kit. Sequences were analyzed for mutations using FinchTV version 1.4.0 (Geospiza, Seattle, WA). RESULTS Patient demographics and tumor characteristics. Thirty patients with available metastatic colon adenocarcinoma tissue to the liver were included for analysis. The group consisted of 18 males and 12 females with a median age at hepatic resection of 62 years (range, 38--79). After their primary colon cancer resection, 67% (20/30) of patients were alive at a median follow-up of 44 months (range, 11--218) with the median time to distant recurrence at 24 months (range, 5--96). Thirty-three percent (10/30) of these patients developed a second recurrence after liver metastasectomy with a median time to recurrence of 15.7 months (range, 2--90). Of the 30 patients with hepatic metastases, 14 of the primary colon cancers were located in the right colon, 13 in the left colon, and the documented location of 3 primary colon neoplasms was not available. A total of 36 hepatic metastases were
Messick et al 229
Surgery Volume 146, Number 2
Table I. Molecular correlates of colon cancer hepatic metastases Subject no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Total (%)
MSI-H S S S S H S S S S S S S S S S S S S S S S S H S S S H S S S 3/30 (10%)
CIMP
KRAS
BRAF
+ + + 3/30 (10%)
Mut WT WT WT Mut WT Mut WT WT WT WT WT WT WT Mut WT WT WT WT WT WT WT WT WT Mut WT WT WT WT WT WT WT WT WT WT WT WT WT WT WT Mut WT WT WT Mut WT Mut WT WT WT Mut WT WT WT Mut WT WT WT WT WT 10/30 0/30 (33%) (0%)
CIMP, CpG island methylator phenotype; H, high; MSI, microsatellite instability; Mut, mutant; S, stable; WT, wild type.
identified from the 30 individuals. Sixty percent (18/30) of patients had liver metastases located in the right liver lobe accounting for 67% (24/36) of lesions, and 40% (12/30) of patients had lesions located in the left lobe accounting for 33% (12/36) of lesions. There were no patients with synchronous lesions located in both lobes. The median tumor size of the liver metastases was 4.5 cm (range, 1.5--12.0). Molecular and genetic analyses. There were 30 metastatic liver tissues available for molecular and genetic analysis. Tumors were heterogeneous according to these traits. Three of 30 (10%) were MSI-H, 3 (10%) were CIMP+, 10 (33%) were KRAS mutant, and there were no BRAF mutations. The complete molecular profile for each metastatic lesion is shown in Table I.
DISCUSSION This study reports the molecular and genetic diversity of colon cancer hepatic metastases that parallels known colorectal neoplastic pathways. Our data reveal that molecular heterogeneity exists within colon cancer hepatic metastases and suggests that differences may also exist between metastatic and primary lesions. Existing differences between primary and metastatic lesions could have important clinical implications regarding adjunctive or palliative chemotherapy decisions. Differences among colon cancers according to oncogenic pathway are well described. The CIN pathway, characterized by a combination of mutations in tumor suppressor genes and proto-oncogenes, and loss of heterozygosity in various chromosomes,1 is the most common colon cancer pathway and accounts for the majority of colon cancers.3-6 The methylator pathway (CIMP) and mutator pathway (Lynch syndrome) are less common and represent approximately 15--20% and 3--5% of colon cancers, respectively.3-6 No previous work has examined the molecular characteristics of metastases based on the described oncogenic pathways of colorectal cancer. Despite the small number of studies describing molecular, genetic, and epigenetic characteristics of hepatic metastases,12,18-23 the overall classification of hepatic metastases based on known molecular correlates has not been explored. When analyzing colon cancer hepatic metastases in this way, differences in the proportion of neoplasms developing through the 3 molecular pathways are observed. Specifically, lesser incidences of MSI-H, CIMP-positivity, and BRAF mutations were recognized within liver metastases when compared with unmatched primary colon cancers.3-5,24 The reported percentage and contribution of these pathways to different cohorts of patients with colorectal cancer are shown in Table II. As expected, the majority of metastatic lesions were consistent with the CIN pathway as characterized by microsatellite stability and CIMP negativity. This observation is supported by the fact that MSS and CIMP-N neoplasms have a greater metastatic potential compared with MSI-H neoplasms18; however, the percentage of liver metastases that were MSI-H (10%) is considerably less than would be expected from the incidence of MSI-H in primary neoplasms.3,4 The differences between primary tumors and metastases are even more dramatic when comparing the incidence of MSI-H hepatic metastases from other studies which report a range of 0--3%.12,19 Two possible explanations for our
230 Messick et al
Surgery August 2009
Table II. Primary and review articles reporting molecular correlates in primary colorectal cancer MSI-H Makinen, 200724 Jass, 20074 Ogino and Goel, 20083 Kalady et al, 20095*
15% 15% 15% 28%
CIMP+ 30–50%y 20% 20% 28%
KRAS mutant
BRAF mutant
40% N/A N/A 33%
15% 20% N/A 17%
*Represents percentage of methylated tumors. yRepresents colon cancers only.
greater incidence of MSI-H metastasis are the study population and the MSI detection method. First, our study excluded rectal cancers, which have a very low percentage of MSI and a greater percentage of MSS cancers.5 Thus, excluding rectal cancers lowers the total number of MSS primary cancers; therefore, the incidence of MSI-H in metastases is greater. Secondly, we utilized a 10-marker microsatellite panel, which has greater sensitivity and is designed to maximize the identification of neoplasms that display instability in additional markers than other panels that use 5 markers.3 Only 1 previous study has evaluated CIMP in resected liver metastases with a reported incidence of 23%.20 The authors concluded that, although gene methylation is an early event in oncogenesis and is maintained through disease progression including liver metastases, there is frequent variability as to which genes get methylated during oncogenesis.20 Therefore, CIMP will be variable depending on the method used to define CIMP. Despite our study using different CIMP markers than the abovementioned study,20 the incidence was similar. In both instances, the rate of CIMP in hepatic metastases was also lower than would be expected according to primary tumor characteristics. There is a high correlation between CIMP positivity and mutations in the oncogene BRAF because they are linked in the serrated methylator oncogenic pathway for colorectal cancer development.2 In primary colorectal cancers, Weisenberger et al2 demonstrated an odds ratio of more than 200 for being CIMP+ if the same neoplasm also harbored a BRAF mutation.2 No BRAF mutations, however, were detected in our study, despite 3 neoplasms having CIMP. The BRAF mutation rate in liver metastases has been reported at approximately 7%,21 and our lack of BRAF mutations may be due to statistical bias given our low number of neoplasms. Either way, the expected BRAF mutation rate in hepatic metastases was less than would be expected based on previous reported incidences in primary neoplasms. Mutations in the oncogene KRAS, which is associated with CIN, did not
show significant differences from other reports or from expected rates based on primary lesions. Thirty-three percent of our hepatic metastases had KRAS mutations, which is consistent with other published reports ranging from 30% to 36%.12,22 The finding of tumor heterogeneity within colon cancer liver metastases has important clinical implications. Molecular characteristics of primary neoplasms have been shown not only to predict response rates to chemotherapy but also to be independent predictors of survival in advanced colorectal cancer.7,8,12 Thus, molecular heterogeneity within metastatic deposits could likely also impact patient survival. Additionally, observed differences in molecular characteristics between primary neoplasms and metastatic disease could impact patient care. Currently, targeted chemotherapy, which is designed to prevent or treat metastatic disease, is guided by molecular and genetic features of primary neoplasms. If metastatic deposits have different characteristics than the primary neoplasm, targeting treatment based on the primary neoplasm may lead to ineffective therapy. This effect could be substantial, even if it only occurred in a small percentage of patients. Possible explanations for the different pathway proportions observed in the liver metastases include the following: Liver metastases are actually similar to the primary lesions, but may seem to be different owing to the small numbers observed; certain molecular types of cancer are more prone to metastasize than others18; or neoplasms ‘‘change’’ their molecular nature in the process of metastasis. Studies are underway currently by our group on matched pairs of primary and metastatic colon and rectal cancers that will facilitate resolving this question. This shift in proportion of molecular characteristics could be one possible reason why overall successful chemotherapy response rates still only approach 50%.7,11,12 Furthermore, these changes may explain a discrepancy in the literature that describes opposing effects of CIMP on response to therapy.8-10 Therefore, treatment decisions might
Surgery Volume 146, Number 2
be better directed by molecular characteristics of metastatic lesions rather than the primary neoplasm, leading to a more individualized treatment algorithm. In conclusion, this study evaluates collectively the molecular correlates of colon cancer hepatic metastases according to defined colorectal neoplastic pathways. Colon cancer hepatic metastases, like primary colon adenocarcinomas, display genetic and molecular diversity. Our data suggest that a lesser proportion of liver metastases may exhibit MSI and have a lower degree of methylation compared with primary neoplasms. Although further matched tumor studies need to be completed to confirm these data and to elucidate mechanisms, the heterogeneity of metastatic disease should be acknowledged when discussing adjuvant therapies in treating colon cancer. REFERENCES 1. Vogelstein B, Fearon ER, Hamilton SR, Kern SE, Preisinger AC, Leppert M, et al. Genetic alterations during colorectaltumor development. N Engl J Med 1988;319:525-32. 2. Weisenberger DJ, Siegmund KD, Campan M, Young J, Long TI, Faasse MA, et al. CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer [see comment]. Nat Genet 2006;38:787-93. 3. Ogino S, Goel A. Molecular classification and correlates in colorectal cancer. J Mol Diagn 2008;10:13-27. 4. Jass JR. Classification of colorectal cancer based on correlation of clinical, morphological and molecular features. Histopathology 2007;50:113-30. 5. Kalady MF, Sanchez J, Manilich E, Hammel J, Casey G, Church J. Divergent oncogenic changes influence survival differences between colon and rectal adenocarcinomas. Dis Colon Rectum 2009;52:1039-45. 6. Sanchez J, Aung PP, Merkulova A, Plummer S, Skacel M, Krumroy L, et al. Genetic and epigenetic classifications define clinical phenotypes and determine patient outcomes in colorectal cancer. Br J Cancer 2009. In press. 7. Ribic CM, Sargent DJ, Moore MJ, Thibodeau SN, French AJ, Goldberg RM, et al. Tumor microsatellite-instability status as a predictor of benefit from fluorouracil-based adjuvant chemotherapy for colon cancer. N Engl J Med 2003;349:247-57. 8. Shen L, Catalano PJ, Benson AB, III, O’Dwyer P, Hamilton SR, Issa JP, et al. Association between DNA methylation and shortened survival in patients with advanced colorectal cancer treated with 5-fluorouracil based chemotherapy. Clin Cancer Res 2007;13:6093-8. 9. Van Rijnsoever M, Elsaleh H, Joseph D, McCaul K, Iacopetta B. CpG island methylator phenotype is an independent predictor of survival benefit from 5-fluorouracil in stage III colorectal cancer. Clin Cancer Res 2003;9:2898-903.
Messick et al 231
10. Iacopetta B, Kawakami K, Watanabe T. Predicting clinical outcome of 5-fluorouracil-based chemotherapy for colon cancer patients: is the CpG island methylator phenotype the 5-fluorouracilresponsive subgroup? Int J Clin Oncol 2008;13:498-503. 11. Lievre A, Bachet JB, Boige V, Cayre A, Le Corre D, Buc E, et al. KRAS mutations as an independent prognostic factor in patients with advanced colorectal cancer treated with cetuximab. J Clin Oncol 2008;26:374-9. 12. Rosty C, Chazal M, Etienne MC, Letoublon C, Bourgeon A, Delpero JR, et al. Determination of microsatellite instability, p53 and K-RAS mutations in hepatic metastases from patients with colorectal cancer: relationship with response to 5-fluorouracil and survival. Int J Cancer 2001;95:162-7. 13. Di Nicolantonio F, Martini M, Molinari F, Sartore-Bianchi A, Arena S, Saletti P, et al. Wild-type BRAF is required for response to panitumumab or cetuximab in metastatic colorectal cancer. J Clin Oncol 20082008;26:5705--12. 14. Masi G, Cupini S, Marcucci L, Cerri E, Loupakis F, Allegrini G, et al. Treatment with 5-fluorouracil/folinic acid, oxaliplatin, and irinotecan enables surgical resection of metastases in patients with initially unresectable metastatic colorectal cancer. Ann Surg Oncol 2006;13:58-65. 15. Tomlinson I, Halford S, Aaltonen L, Hawkins N, Ward R, Tomlinson I, et al. Does MSI-low exist? J Pathol 2002;197: 6-13. 16. Eads CA, Danenberg KD, Kawakami K, Saltz LB, Blake C, Shibata D, et al. MethyLight: a high-throughput assay to measure DNA methylation. Nucleic Acids Res 2000;28:E32. 17. Trinh BN, Long TI, Laird PW. DNA methylation analysis by MethyLight technology. Methods 2001;25:456-62. 18. Buckowitz A, Knaebel HP, Benner A, Blaker H, Gebert J, Kienle P, et al. Microsatellite instability in colorectal cancer is associated with local lymphocyte infiltration and low frequency of distant metastases. Br J Cancer 2005;92:1746-53. 19. Haddad R, Ogilvie RT, Croitoru M, Muniz V, Gryfe R, Pollet A, et al. Microsatellite instability as a prognostic factor in resected colorectal cancer liver metastases. Ann Surg Oncol 2004;11:977-82. 20. Kim YH, Petko Z, Dzieciatkowski S, Lin L, Ghiassi M, Stain S, et al. CpG island methylation of genes accumulates during the adenoma progression step of the multistep pathogenesis of colorectal cancer. Genes Chromosomes Cancer 2006;45:781-9. 21. Miranda E, Destro A, Malesci A, Balladore E, Bianchi P, Baryshnikova E, et al. Genetic and epigenetic changes in primary metastatic and nonmetastatic colorectal cancer. Br J Cancer 2006;95:1101-7. 22. Santini D, Loupakis F, Vincenzi B, Floriani I, Stasi I, Canestrari E, et al. High concordance of KRAS status between primary colorectal tumors and related metastatic sites: implications for clinical practice. Oncologist 2008;13:1270-5. 23. Artale S, Sartore-Bianchi A, Veronese SM, Gambi V, Sarnataro CS, Gambacorta M, et al. Mutations of KRAS and BRAF in primary and matched metastatic sites of colorectal cancer. J Clin Oncol 2008;26:4217-9. 24. Makinen MJ. Colorectal serrated adenocarcinoma. Histopathology 2007;50:131-50.
Background. Colon cancer arises through distinct molecular pathways resulting in diverse tumor populations demonstrated by differences in microsatellite instability (MSI), CpG island methylator phenotype (CIMP), and mutations in oncogenes KRAS and BRAF. Although these molecular differences are well-described for primary neoplasms, the molecular nature of hepatic metastases is not wellcharacterized. This study seeks to describe molecular characteristics of colon cancer hepatic metastases in terms of oncogenic pathway. Methods. Tumor DNA was isolated from fresh frozen hepatic metastases from colon cancer and analyzed for MSI by polymerase chain reaction (PCR)-based microsatellite analysis and for CIMP using MethyLight quantitative PCR. KRAS and BRAF oncogenes were analyzed for DNA mutations. Metastases were classified by their molecular and genetic features. Unfortunately, tissue from the primary neoplasms from these patients were not available Results. Thirty patients with liver metastases from colon cancer were studied. Molecular analysis revealed 10% (3/30) were MSI-H, 10% (3/30) were CIMP positive, 33% (10/30) had KRAS mutations, and none had BRAF mutations. Literature describing primary colon cancers reports an incidence of approximately 20% MSI-H, 20% CIMP-positive, 35% KRAS mutants, and 17% BRAF mutants. Conclusion. Hepatic metastases from colon cancer, like primary colon adenocarcinomas, show genetic and molecular diversity. Furthermore, hepatic metastases may have a different incidence of MSI and methylation compared with primary neoplasms. These differences could impact treatment decisions. (Surgery 2009;146:227-31.) From the Department of Colorectal Surgery,a and the Department of Cancer Biology,b Digestive Disease Institute, Cleveland Clinic, Cleveland, OH
COLORECTAL ONCOGENESIS is a complex sequence of events resulting in the transformation of normal colonic epithelium into malignant tumor cells whose unchecked growth leads to continued progression and metastasis. The current understanding of colorectal oncogenesis is based on the concept of 3 or more molecular mechanisms underlying the change from benign to malignant disease: chromosomal instability (CIN), DNA microsatellite instability (MSI), and hypermethylation of specific tumor Presented at the Academic Surgical Congress. The authors make the following disclosures: James M. Church, MB, ChB: Myriad Genetics, speaker, honorarium; Salix, consultant, honorarium; Cleveland Clinic/Cologene, employee, salary. Matthew F. Kalady, MD: Genzyme, speaker’s bureau, honorarium. Accepted for publication June 4, 2009. Reprint requests: Craig A. Messick, MD, Department of Colorectal Surgery, The Cleveland Clinic, 9500 Euclid Avenue, A30, Cleveland, OH 44106. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2009 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2009.06.003
suppressor genes resulting in the CpG island methylator phenotype (CIMP), as well as others.1-6 Importantly, these distinct pathways defined by molecular and genetic features correspond with distinct clinical phenotypes and are also associated with differing responses to particular chemotherapy and clinical outcomes.5,6 Treatment decisions may be made based on the characteristics of the neoplasms. For example, MSIhigh (MSI-H) colon cancers treated by resection alone enjoy a survival advantage compared with microsatellite stable (MSS) neoplasms.7 MSI-H neoplasms, however, do not respond as well to 5-fluourouracil--based chemotherapy and stage III MSI-H colon cancer patients seem to do worse when treated with 5-fluourouracil compared with MSS patients.7 Similarly, CIMP status has been shown to affect response to chemotherapy and overall survival.8-10 Furthermore, metastatic neoplasms that display mutations in the oncogenes KRAS and BRAF are resistant to adjuvant treatment with cetuximab.11-13 Despite directed therapy based on molecular and genetic characteristics of the primary neoplasm, SURGERY 227
228 Messick et al
best responses are still only approximately 15--40% for stage III disease8,12 and median survival for stage IV disease is variable at 10--22 months.8,14 Although patient-specific factors may contribute to chemotherapy resistance, the rates of low treatment success remain incompletely explained. One hypothesis is that colon cancer metastases are genetically heterogeneous and possibly more than, or in different ways from, the primary neoplasms. Thus, a standard approach to treating metastatic disease using the same chemotherapy regimen for all patients is unlikely to be uniformly effective. If chemotherapy could be tailored to the genetics of the metastases, better response rates may be achieved. Although described for primary colon cancers, molecular characteristics of metastases are not well described. This study describes the molecular and genetic diversity of colon adenocarcinoma hepatic metastases in the context of how it may potentially influence clinical decision making. METHODS Tissue collection and patient information. This study was approved by the Cleveland Clinic Institutional Review Board. Tissue from resected colon cancer hepatic metastases was obtained from a prospectively collected and maintained frozen tissue biobank. All available colon cancer hepatic metastases were included. A gastrointestinal pathologist confirmed the diagnosis of metastatic adenocarcinoma by review of hematoxylin and eosin--stained slides. Only tissues sections containing at least 60% adenocarcinoma were included. Tumor information and patient demographics were recorded. DNA isolation. Frozen tissue blocks were sectioned using a cryostat into seven to ten 10-mm sections then subjected to an overnight proteinase K digestion. Using the Qiagen BioRobot EZ1 workstation, genomic DNA was isolated with the Qiagen EZ1 DNA Tissue Kit as per the manufacturer’s instructions (Qiagen, Valencia, CA). Genetic analyses. Microsatellite analysis: Isolated genomic DNA from hepatic metastases was tested for MSI using a panel of 10 mononucleotide, dinucleotide, and penta-mono-tetra compound markers as previously described.5,6 Additional, normal, nonmalignant, adjacent hepatic tissue was used as a source of genomic DNA to which tumor microsatellite status was compared for each individual patient. Briefly, standard polymerase chain reaction (PCR) techniques were employed for DNA amplification and then both tumor and normal adjacent tumor DNA were sequenced. All samples were read and scored for their microsatellite
Surgery August 2009
status by 2 independent evaluators. Neoplasms with 3 (30%) or more of the 10 markers exhibiting instability were considered to be MSI-H; 1 or 2 markers exhibiting instability were considered to be MSI-low (MSI-L); those with no markers exhibiting instability were considered to be MSS. For this study, neoplasms that were MSI-L and MSS were grouped together and designated MSS.15 CIMP analysis: CIMP status was determined as described previously by Eads et al,16 and as reported by our group.5,6 Briefly, genomic DNA underwent a sodium bisulfite conversion and universally methylated human DNA was used as positive control. All sodium bisulfite conversion reactions were repeated twice to ensure validity of the results. MethyLight quantitative PCR was then performed on a panel of 5 markers (CACNA1G, IGF2, NEUROG1, RUNX3, and SOCS1). The percent of methylated reference was then calculated as described previously.17 A given marker was considered positive if the percent of methylated reference was greater than 10 and samples with 3 or more positive markers were considered CIMP positive (CIMP+) and those with 2 or fewer CIMP negative (CIMP-N). Mutation analysis: KRAS and BRAF mutations were evaluated by PCR sequencing analysis. Genomic DNA was amplified at 2 KRAS exons (exon 3 codon 61, exon 2 codon 12, and exon 2 codon 13) and 2 BRAF exons (exon 11 codon 468, exon 15 codon 596, and exon 15 codon 599) by standard PCR. All PCR products were purified using the Qiagen QiaQuick Kit. Sequences were analyzed for mutations using FinchTV version 1.4.0 (Geospiza, Seattle, WA). RESULTS Patient demographics and tumor characteristics. Thirty patients with available metastatic colon adenocarcinoma tissue to the liver were included for analysis. The group consisted of 18 males and 12 females with a median age at hepatic resection of 62 years (range, 38--79). After their primary colon cancer resection, 67% (20/30) of patients were alive at a median follow-up of 44 months (range, 11--218) with the median time to distant recurrence at 24 months (range, 5--96). Thirty-three percent (10/30) of these patients developed a second recurrence after liver metastasectomy with a median time to recurrence of 15.7 months (range, 2--90). Of the 30 patients with hepatic metastases, 14 of the primary colon cancers were located in the right colon, 13 in the left colon, and the documented location of 3 primary colon neoplasms was not available. A total of 36 hepatic metastases were
Messick et al 229
Surgery Volume 146, Number 2
Table I. Molecular correlates of colon cancer hepatic metastases Subject no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Total (%)
MSI-H S S S S H S S S S S S S S S S S S S S S S S H S S S H S S S 3/30 (10%)
CIMP
KRAS
BRAF
+ + + 3/30 (10%)
Mut WT WT WT Mut WT Mut WT WT WT WT WT WT WT Mut WT WT WT WT WT WT WT WT WT Mut WT WT WT WT WT WT WT WT WT WT WT WT WT WT WT Mut WT WT WT Mut WT Mut WT WT WT Mut WT WT WT Mut WT WT WT WT WT 10/30 0/30 (33%) (0%)
CIMP, CpG island methylator phenotype; H, high; MSI, microsatellite instability; Mut, mutant; S, stable; WT, wild type.
identified from the 30 individuals. Sixty percent (18/30) of patients had liver metastases located in the right liver lobe accounting for 67% (24/36) of lesions, and 40% (12/30) of patients had lesions located in the left lobe accounting for 33% (12/36) of lesions. There were no patients with synchronous lesions located in both lobes. The median tumor size of the liver metastases was 4.5 cm (range, 1.5--12.0). Molecular and genetic analyses. There were 30 metastatic liver tissues available for molecular and genetic analysis. Tumors were heterogeneous according to these traits. Three of 30 (10%) were MSI-H, 3 (10%) were CIMP+, 10 (33%) were KRAS mutant, and there were no BRAF mutations. The complete molecular profile for each metastatic lesion is shown in Table I.
DISCUSSION This study reports the molecular and genetic diversity of colon cancer hepatic metastases that parallels known colorectal neoplastic pathways. Our data reveal that molecular heterogeneity exists within colon cancer hepatic metastases and suggests that differences may also exist between metastatic and primary lesions. Existing differences between primary and metastatic lesions could have important clinical implications regarding adjunctive or palliative chemotherapy decisions. Differences among colon cancers according to oncogenic pathway are well described. The CIN pathway, characterized by a combination of mutations in tumor suppressor genes and proto-oncogenes, and loss of heterozygosity in various chromosomes,1 is the most common colon cancer pathway and accounts for the majority of colon cancers.3-6 The methylator pathway (CIMP) and mutator pathway (Lynch syndrome) are less common and represent approximately 15--20% and 3--5% of colon cancers, respectively.3-6 No previous work has examined the molecular characteristics of metastases based on the described oncogenic pathways of colorectal cancer. Despite the small number of studies describing molecular, genetic, and epigenetic characteristics of hepatic metastases,12,18-23 the overall classification of hepatic metastases based on known molecular correlates has not been explored. When analyzing colon cancer hepatic metastases in this way, differences in the proportion of neoplasms developing through the 3 molecular pathways are observed. Specifically, lesser incidences of MSI-H, CIMP-positivity, and BRAF mutations were recognized within liver metastases when compared with unmatched primary colon cancers.3-5,24 The reported percentage and contribution of these pathways to different cohorts of patients with colorectal cancer are shown in Table II. As expected, the majority of metastatic lesions were consistent with the CIN pathway as characterized by microsatellite stability and CIMP negativity. This observation is supported by the fact that MSS and CIMP-N neoplasms have a greater metastatic potential compared with MSI-H neoplasms18; however, the percentage of liver metastases that were MSI-H (10%) is considerably less than would be expected from the incidence of MSI-H in primary neoplasms.3,4 The differences between primary tumors and metastases are even more dramatic when comparing the incidence of MSI-H hepatic metastases from other studies which report a range of 0--3%.12,19 Two possible explanations for our
230 Messick et al
Surgery August 2009
Table II. Primary and review articles reporting molecular correlates in primary colorectal cancer MSI-H Makinen, 200724 Jass, 20074 Ogino and Goel, 20083 Kalady et al, 20095*
15% 15% 15% 28%
CIMP+ 30–50%y 20% 20% 28%
KRAS mutant
BRAF mutant
40% N/A N/A 33%
15% 20% N/A 17%
*Represents percentage of methylated tumors. yRepresents colon cancers only.
greater incidence of MSI-H metastasis are the study population and the MSI detection method. First, our study excluded rectal cancers, which have a very low percentage of MSI and a greater percentage of MSS cancers.5 Thus, excluding rectal cancers lowers the total number of MSS primary cancers; therefore, the incidence of MSI-H in metastases is greater. Secondly, we utilized a 10-marker microsatellite panel, which has greater sensitivity and is designed to maximize the identification of neoplasms that display instability in additional markers than other panels that use 5 markers.3 Only 1 previous study has evaluated CIMP in resected liver metastases with a reported incidence of 23%.20 The authors concluded that, although gene methylation is an early event in oncogenesis and is maintained through disease progression including liver metastases, there is frequent variability as to which genes get methylated during oncogenesis.20 Therefore, CIMP will be variable depending on the method used to define CIMP. Despite our study using different CIMP markers than the abovementioned study,20 the incidence was similar. In both instances, the rate of CIMP in hepatic metastases was also lower than would be expected according to primary tumor characteristics. There is a high correlation between CIMP positivity and mutations in the oncogene BRAF because they are linked in the serrated methylator oncogenic pathway for colorectal cancer development.2 In primary colorectal cancers, Weisenberger et al2 demonstrated an odds ratio of more than 200 for being CIMP+ if the same neoplasm also harbored a BRAF mutation.2 No BRAF mutations, however, were detected in our study, despite 3 neoplasms having CIMP. The BRAF mutation rate in liver metastases has been reported at approximately 7%,21 and our lack of BRAF mutations may be due to statistical bias given our low number of neoplasms. Either way, the expected BRAF mutation rate in hepatic metastases was less than would be expected based on previous reported incidences in primary neoplasms. Mutations in the oncogene KRAS, which is associated with CIN, did not
show significant differences from other reports or from expected rates based on primary lesions. Thirty-three percent of our hepatic metastases had KRAS mutations, which is consistent with other published reports ranging from 30% to 36%.12,22 The finding of tumor heterogeneity within colon cancer liver metastases has important clinical implications. Molecular characteristics of primary neoplasms have been shown not only to predict response rates to chemotherapy but also to be independent predictors of survival in advanced colorectal cancer.7,8,12 Thus, molecular heterogeneity within metastatic deposits could likely also impact patient survival. Additionally, observed differences in molecular characteristics between primary neoplasms and metastatic disease could impact patient care. Currently, targeted chemotherapy, which is designed to prevent or treat metastatic disease, is guided by molecular and genetic features of primary neoplasms. If metastatic deposits have different characteristics than the primary neoplasm, targeting treatment based on the primary neoplasm may lead to ineffective therapy. This effect could be substantial, even if it only occurred in a small percentage of patients. Possible explanations for the different pathway proportions observed in the liver metastases include the following: Liver metastases are actually similar to the primary lesions, but may seem to be different owing to the small numbers observed; certain molecular types of cancer are more prone to metastasize than others18; or neoplasms ‘‘change’’ their molecular nature in the process of metastasis. Studies are underway currently by our group on matched pairs of primary and metastatic colon and rectal cancers that will facilitate resolving this question. This shift in proportion of molecular characteristics could be one possible reason why overall successful chemotherapy response rates still only approach 50%.7,11,12 Furthermore, these changes may explain a discrepancy in the literature that describes opposing effects of CIMP on response to therapy.8-10 Therefore, treatment decisions might
Surgery Volume 146, Number 2
be better directed by molecular characteristics of metastatic lesions rather than the primary neoplasm, leading to a more individualized treatment algorithm. In conclusion, this study evaluates collectively the molecular correlates of colon cancer hepatic metastases according to defined colorectal neoplastic pathways. Colon cancer hepatic metastases, like primary colon adenocarcinomas, display genetic and molecular diversity. Our data suggest that a lesser proportion of liver metastases may exhibit MSI and have a lower degree of methylation compared with primary neoplasms. Although further matched tumor studies need to be completed to confirm these data and to elucidate mechanisms, the heterogeneity of metastatic disease should be acknowledged when discussing adjuvant therapies in treating colon cancer. REFERENCES 1. Vogelstein B, Fearon ER, Hamilton SR, Kern SE, Preisinger AC, Leppert M, et al. Genetic alterations during colorectaltumor development. N Engl J Med 1988;319:525-32. 2. Weisenberger DJ, Siegmund KD, Campan M, Young J, Long TI, Faasse MA, et al. CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer [see comment]. Nat Genet 2006;38:787-93. 3. Ogino S, Goel A. Molecular classification and correlates in colorectal cancer. J Mol Diagn 2008;10:13-27. 4. Jass JR. Classification of colorectal cancer based on correlation of clinical, morphological and molecular features. Histopathology 2007;50:113-30. 5. Kalady MF, Sanchez J, Manilich E, Hammel J, Casey G, Church J. Divergent oncogenic changes influence survival differences between colon and rectal adenocarcinomas. Dis Colon Rectum 2009;52:1039-45. 6. Sanchez J, Aung PP, Merkulova A, Plummer S, Skacel M, Krumroy L, et al. Genetic and epigenetic classifications define clinical phenotypes and determine patient outcomes in colorectal cancer. Br J Cancer 2009. In press. 7. Ribic CM, Sargent DJ, Moore MJ, Thibodeau SN, French AJ, Goldberg RM, et al. Tumor microsatellite-instability status as a predictor of benefit from fluorouracil-based adjuvant chemotherapy for colon cancer. N Engl J Med 2003;349:247-57. 8. Shen L, Catalano PJ, Benson AB, III, O’Dwyer P, Hamilton SR, Issa JP, et al. Association between DNA methylation and shortened survival in patients with advanced colorectal cancer treated with 5-fluorouracil based chemotherapy. Clin Cancer Res 2007;13:6093-8. 9. Van Rijnsoever M, Elsaleh H, Joseph D, McCaul K, Iacopetta B. CpG island methylator phenotype is an independent predictor of survival benefit from 5-fluorouracil in stage III colorectal cancer. Clin Cancer Res 2003;9:2898-903.
Messick et al 231
10. Iacopetta B, Kawakami K, Watanabe T. Predicting clinical outcome of 5-fluorouracil-based chemotherapy for colon cancer patients: is the CpG island methylator phenotype the 5-fluorouracilresponsive subgroup? Int J Clin Oncol 2008;13:498-503. 11. Lievre A, Bachet JB, Boige V, Cayre A, Le Corre D, Buc E, et al. KRAS mutations as an independent prognostic factor in patients with advanced colorectal cancer treated with cetuximab. J Clin Oncol 2008;26:374-9. 12. Rosty C, Chazal M, Etienne MC, Letoublon C, Bourgeon A, Delpero JR, et al. Determination of microsatellite instability, p53 and K-RAS mutations in hepatic metastases from patients with colorectal cancer: relationship with response to 5-fluorouracil and survival. Int J Cancer 2001;95:162-7. 13. Di Nicolantonio F, Martini M, Molinari F, Sartore-Bianchi A, Arena S, Saletti P, et al. Wild-type BRAF is required for response to panitumumab or cetuximab in metastatic colorectal cancer. J Clin Oncol 20082008;26:5705--12. 14. Masi G, Cupini S, Marcucci L, Cerri E, Loupakis F, Allegrini G, et al. Treatment with 5-fluorouracil/folinic acid, oxaliplatin, and irinotecan enables surgical resection of metastases in patients with initially unresectable metastatic colorectal cancer. Ann Surg Oncol 2006;13:58-65. 15. Tomlinson I, Halford S, Aaltonen L, Hawkins N, Ward R, Tomlinson I, et al. Does MSI-low exist? J Pathol 2002;197: 6-13. 16. Eads CA, Danenberg KD, Kawakami K, Saltz LB, Blake C, Shibata D, et al. MethyLight: a high-throughput assay to measure DNA methylation. Nucleic Acids Res 2000;28:E32. 17. Trinh BN, Long TI, Laird PW. DNA methylation analysis by MethyLight technology. Methods 2001;25:456-62. 18. Buckowitz A, Knaebel HP, Benner A, Blaker H, Gebert J, Kienle P, et al. Microsatellite instability in colorectal cancer is associated with local lymphocyte infiltration and low frequency of distant metastases. Br J Cancer 2005;92:1746-53. 19. Haddad R, Ogilvie RT, Croitoru M, Muniz V, Gryfe R, Pollet A, et al. Microsatellite instability as a prognostic factor in resected colorectal cancer liver metastases. Ann Surg Oncol 2004;11:977-82. 20. Kim YH, Petko Z, Dzieciatkowski S, Lin L, Ghiassi M, Stain S, et al. CpG island methylation of genes accumulates during the adenoma progression step of the multistep pathogenesis of colorectal cancer. Genes Chromosomes Cancer 2006;45:781-9. 21. Miranda E, Destro A, Malesci A, Balladore E, Bianchi P, Baryshnikova E, et al. Genetic and epigenetic changes in primary metastatic and nonmetastatic colorectal cancer. Br J Cancer 2006;95:1101-7. 22. Santini D, Loupakis F, Vincenzi B, Floriani I, Stasi I, Canestrari E, et al. High concordance of KRAS status between primary colorectal tumors and related metastatic sites: implications for clinical practice. Oncologist 2008;13:1270-5. 23. Artale S, Sartore-Bianchi A, Veronese SM, Gambi V, Sarnataro CS, Gambacorta M, et al. Mutations of KRAS and BRAF in primary and matched metastatic sites of colorectal cancer. J Clin Oncol 2008;26:4217-9. 24. Makinen MJ. Colorectal serrated adenocarcinoma. Histopathology 2007;50:131-50.