|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology ras Mutations Are Associated With Aggressive Tumor Phenotypes and Poor Prognosis in Thyroid Cancer
From the Departments of Pathology, Epidemiology and Public Health, and Psychiatry, Yale University School of Medicine, New Haven, CT; Cancer Epidemiology Service, National Center for Epidemiology, National Institute of Health Carlos III, Madrid; Department of Pathology, Oviedo University School of Medicine, Oviedo; and Department of Pathology, Navarra University School of Medicine, Pamplona, Spain. Address reprint requests to Giovanni Tallini, MD, Anatomia Patologica, Università di Bologna-Ospedale Bellaria, Via Altura 3, 40139 Bologna, Italy; e-mail: Giovanni.Tallini{at}ausl.bologna.it. or Ginesa Garcia-Rostan, MD, Instituto de Patologia e Immunologia Molecular de Universidade do Porto, Rua Roberto Frias, 4200 Porto, Portugal; e-mail: grostan{at}ipatimup.pt.
Purpose: ras oncogenic activation has long been demonstrated in thyroid carcinomas of follicular cell derivation, but no consistent relationship has been shown between mutations and clinicopathologic features. Materials and Methods: We analyzed H-, K-, and N-ras mutations by polymerase chain reaction–single-strand conformational polymorphism followed by DNA sequencing in 125 thyroid carcinoma specimens from 107 patients, to include tumors covering the entire spectrum of thyroid tumor differentiation. Results: Mutations were identified in four (8.2%) of 49 well-differentiated carcinomas (WDCs; two [6.7%] of 30 of the tumors were papillary carcinomas, two [10.5%] of 19 of them were follicular carcinomas), in 16 (55.2%) of 29 poorly differentiated carcinomas (PDCs), and in 15 (51.7%) of 29 undifferentiated carcinomas, with a significant association between ras mutation and poorly or undifferentiated tumors (P < .001). Twenty-six (74.3%) of 35 patients with ras-mutated tumors died as a result of disease as opposed to 23 (31.9%) of 72 patients with tumors lacking the mutations. Among patients with differentiated thyroid carcinomas (WDC and PDC), 11 (55.0%) of 20 patients with mutated tumors died as a result of disease as opposed to nine (15.5%) of 58 patients with wild-type ras tumors, and the correlation was independent of tumor differentiation and stage (P = .016). K-ras codon 13 mutations (all with G-A nucleotide transitions resulting in Gly>Asp substitution) and single activating mutations in any of the ras genes were also independent predictors of poor survival in differentiated thyroid carcinomas (P = .027 and P = .007, respectively). Conclusion: These findings demonstrate that ras mutations are a marker for aggressive cancer behavior and indicate a possible role of ras genotyping to identify thyroid carcinoma subsets associated with poor prognosis.
THE THREE members of the ras gene family (H-, K-, and N-ras) encode membrane-associated guanine nucleotide-binding proteins (p21ras). Point mutations affecting the guanosine triphosphate (GTP)-binding domain (codons12/13) or the GTPase domain (codon 61) determine the replacement of specific amino acid residues that lock p21ras in the active GTP-bound form, resulting in constitutive activation of the protein and tumor development.1 Oncogenic mutations of H-, K-, and N-ras were among the first genetic changes to be identified in tumors originating from the thyroid follicular epithelium, and numerous reports have documented their occurrence in many different types of thyroid tumors.1 The prevalence of ras mutations shows considerable variability among the different series, and environmental factors such as radiation exposure may influence both occurrence and pattern of ras mutation.2–4 Despite the numerous reports, relatively few studies have analyzed the significance of the ras mutation status for tumor prognosis and its impact on survival. The few studies that have fully addressed these issues have been limited to specific types of thyroid cancer or of ras mutation.5 To analyze the influence of oncogenic ras on the patient’s clinical course, we have therefore genotyped for H-, K-, and N-ras more than 100 thyroid carcinomas to include the entire spectrum of differentiation from well-differentiated carcinomas (WDCs) to undifferentiated (anaplastic) carcinomas (UDCs). Polymerase chain reaction–single-strand conformational polymorphism (PCR-SSCP), followed by DNA sequencing of each individual shifted band was used for mutation detection.
Patients With Tumors We analyzed 107 patients who underwent surgery for thyroid carcinoma. Diagnostic material on these patients was retrieved from the files of the Pathology Departments at Yale New Haven Hospital, Yale University (New Haven, CT; 55 patients) and Covadonga Hospital, University of Oviedo (Oviedo, Spain; 52 patients). Patients were chosen randomly among those with detailed clinical and follow-up data to cover the entire spectrum of differentiation for tumors of follicular cell origin. All histologic diagnoses were reviewed according to established histologic criteria.6 Patients with UDC received palliative treatment, whereas those with differentiated thyroid carcinoma (WDC and PDC) underwent total thyroidectomy followed by postoperative iodine treatment, according to standard clinical protocols. Forty-nine patients died as a result of disease during follow-up; all cancer survivors were observed for a median period of 84 months (range, 11 to 262 months) or until death. Processing of samples and of patient information proceeded in agreement with review board approved protocols.
DNA Extraction
PCR-SSCP
Sequence Analysis All bands from samples exhibiting reproducible mobility shifts after independent repeat of the PCR-SSCP assay were excised, eluted, and amplified by PCR using the same primer set and reaction conditions described above. The reamplified products were separated, purified, and analyzed by the W.M. Keck Biotechnology Resource Laboratory at Yale University using an automated Applied Biosystems 373A Stretch DNA sequencer (Perkin-Elmer, Norwalk, CT). Nucleotide sequencing from both the sense and antisense orientation was performed for confirmation. All mutated patient samples were further verified by repeating the PCR-SSCP assay.
Statistical Analysis
Pathologic and Clinical Data The pathologic and clinical features of the tumors are summarized in Table 2  . Similar to previous descriptions,7 the tumors were divided into three types on the basis of their degree of differentiation. The first set consisted of well-differentiated thyroid carcinomas (16 patients from Spain; 33 patients from the United States) of either papillary or follicular type. The second set consisted of PDCs (16 patients from Spain; 13 patients from the United States) exhibiting the features of insular carcinoma (12 patients)9 or consisting of neoplasms with a trabecular or solid (comedo type) growth pattern with nuclear hyperchromasia, high mitotic activity, and necrosis (poorly differentiated thyroid carcinoma not otherwise specified, 17 patients).6 On the basis of their morphologic features, the PDCs were also classified as papillary or follicular (Table 2). The third set included anaplastic UDCs (20 patients from Spain; nine patients from the United States). Histologic classification of thyroid carcinomas in the three groups according to their degree of differentiation successfully stratified the mortality risk for thyroid carcinoma (Fig 1A), thus justifying the validity of this approach. Differentiated thyroid carcinomas (WDC and PDC) are traditionally subclassified into papillary and follicular histotypes. There was no significant difference in survival after conventional morphologic subclassification of the differentiated carcinoma group (PDC and WDC, 78 patients) into differentiated follicular and differentiated papillary cancer (Fig 1B). Patients with oncocytic tumors (Hürthle cell carcinomas) had a higher mortality compared with patients with other differentiated carcinomas but the difference was not statistically significant (data not shown). Because there was a significant association between tumor origin (Spain v United States) and poorly or undifferentiated tumors (P = .006), tumor origin was included as a covariate in assessing the association between ras mutation and clinicopathologic parameters or survival (see subsection, ras Mutation Pattern and Correlation With Clinicopathologic Parameters).
ras Mutation Pattern and Correlation With Clinicopathologic Parameters The SSCP pattern and sequence analysis of common types of ras mutations are illustrated in Figure 2 . The results of H-, K- and N-ras mutation analysis and the correlation of ras mutation status with clinicopathologic features are summarized in Tables 3 and 4, respectively. ras mutations were identified in 35 of 107 tumors (32.7%). K-ras had the largest number of mutations (26 of 107 tumors, 24.3%), all of them occurring either at codon 12 or at codon 13 of exon 1. N-ras was mutated in nine of 107 tumors (8.4%) and H-ras was mutated in five of 107 tumors (4.7%). The majority of nucleotide changes were transitions (26 of 107 tumors, 24.3%). Multiple activating mutations were detected in 17 of the 35 tumors that harbored oncogenic ras and were restricted to PDC and UDC.
Mutational analysis in the eight PDCs and six UDCs with additional DNA samples from tumor recurrence, metastases, and areas with different histologic appearance demonstrated that ras mutations segregated with the less-differentiated portions of the tumor and that the type and pattern of ras mutations was consistent with that identified in the primary lesion.
Amino acid changes at the ras GTP-binding domain (corresponding to codons 12 and 13 of exon 1) involved replacement of glycine with aspartic acid (30 mutations), cysteine (13 mutations), serine (eight mutations), alanine (four mutations), and valine (three mutations; Table 3
ras mutations were present in four of 49 (8.2%) WDCs (two of 30 [6.7%] of well-differentiated papillary carcinomas and two of 19 [10.5%] of well-differentiated follicular carcinomas), in 16 of 29 (55.2%) PDCs, and in 15 of 29 (51.7%) UDCs, with a significant association between ras mutation and poorly or undifferentiated tumor phenotypes (P < .001,
In general, ras mutations were associated not only with histologic features (ie, loss of tumor differentiation) but also with clinicopathologic parameters indicative of aggressive behavior, such as large tumor size and vascular invasion (Table 4
ras Mutations and Disease-Specific Survival
Given the uniformly fatal outcome of patients with UDC (all of them were dead as a result of disease within 1 year of diagnosis; Table 2  ) survival analysis was also performed, after exclusion of UDC patients, on the remaining differentiated thyroid carcinomas (WDC and PDC). The results, which are illustrated in Figure 4, demonstrate that the ras mutation status is associated with tumor-related death. Table 5 summarizes the results of multivariate analyses. Specific types of ras mutations were entered as covariates in separate models, together with tumor differentiation (WDCs v PDCs), tumor stage, and origin (Spain v United States). As expected, tumor stage was always an independent predictor of survival, with a relative risk ranging from 12 to 18 for stage 3 tumors and from 12 to 44 for stage 4 tumors. ras mutations in general, as well as the specific ras mutation patterns shown in Table 5, were independently associated with survival and were more powerful predictors of outcome than the morphologic diagnosis of poor tumor differentiation. Patients with K-ras codon 13 mutated tumors, with tumors harboring Gly>Asp substitutions (Fig 4), and with tumor mutations caused by nucleotide transitions (not shown) had similar survival curves because all of the patients with K-ras codon 13 mutations featured second nucleotide G-A transitions in codon 13, resulting in Gly>Asp substitutions. Conversely, K-ras codon 13 was mutated in the majority of the differentiated carcinomas with Gly>Asp substitutions (11 of 14 patients) and with nucleotide transition mutations (11 of 16 patients). Tumor origin did not influence survival. Similarly, histologic classification of differentiated thyroid tumors as papillary or follicular did not have a statistical influence on the association between ras mutation status and survival.
Activating H-, K-, and N-ras mutations represent the most common type of abnormality of a dominant oncogene in human cancer and have been identified in many different types of tumors, with specificity and type of mutation varying in relation to the tumor type.1,10 Numerous studies have addressed the relationship between ras mutations and the clinicopathologic features of the tumors harboring the mutation. Several studies, including a prospective study11 and large meta-analyses,12,13 have shown that ras mutations are associated with poor prognosis in colorectal adenocarcinoma and that different gene mutations have different prognostic impact.11–13 Both prospective14 and retrospective15,16 analyses have shown that K-ras mutations are associated with poor prognosis in non–small-cell lung carcinoma. The type of K-ras mutation may also influence survival in pancreatic adenocarcinoma,17 whereas N-ras mutations are associated with failure to achieve complete remission in acute myeloid leukemia.18 Constitutive activation of all three ras oncogenes (H-, K-, and N-ras) is known to occur among tumors that originated from the follicular epithelium of the thyroid gland.19 However, there are significant discrepancies related to the overall frequency of ras mutations (ranging from 7% to 62%)20,21 and their prevalence in specific thyroid tumors. No consistent relationship between tumor histotype or biologic behavior and one particular pattern of ras activation can be inferred from a review of the literature. Although it is difficult to explain this lack of consistency, the mutation screening methods, the selection of patients, and the design of individual studies are critical to identify specific associations between ras mutational status and clinical or pathologic parameters. We have used PCR-SSCP, a technique that, despite its high sensitivity22 and its widespread use for mutation detection in human cancer, has only infrequently been applied to the study of thyroid neoplasms.21,23 We have also selected tumors that include the full spectrum of differentiation observed in thyroid cancer of follicular cell derivation with a large number of patients—the largest series so far analyzed for ras mutations—to allow for meaningful statistical analysis. This study demonstrates that ras mutations define a subset of thyroid carcinoma characterized by aggressive behavior. This is indicated by the close relationship between oncogenic ras and the loss of those histologic features that characterize well-differentiated thyroid tumor phenotypes.6 Remarkably, oncogenic K-ras not only correlates with the loss of tumor differentiation but also with the presence of distant metastases, independent of tumor differentiation. This is consistent with the ras role in modulating cell motility and invasiveness, and with the recent observation that K-ras mutations are detectable in the large majority of early metastatic deposits in the bone marrow of patients with colonic adenocarcinoma.24 Potentially relevant for patient management is the finding that ras mutations are associated with poor prognosis among differentiated carcinomas (the WDC and PDC tumor group) independent of tumor stage and of whether the tumor is subclassified morphologically as well or poorly differentiated, papillary or follicular. As expected, stage exhibited the strongest correlation with survival but the ras mutation status was a more powerful indicator of outcome than the histologic diagnosis of PDC. Patients with UDC, because of to the rapidly fatal outcome of this tumor type, represented a bias for the survival analysis that justified their separation from the rest of the tumors. Indeed, UDC has long been recognized as a distinctive type of thyroid malignancy characterized by complete loss of tumor differentiation and an aggressive behavior resulting in the patient death as a result of uncontainable tumor growth in the neck.6 The vast majority of the remaining types of thyroid carcinoma are indolent tumors, but a small minority of them can be difficult to control and may ultimately result in patient death.25 These neoplasms include a rather heterogeneous morphologic spectrum and often exhibit the histologic features of PDC.6 Unlike the morphologic criteria that define UDC, those used by different pathologists for the diagnosis of poorly differentiated carcinomas are not always comparable. The results of this study indicate that ras mutation analysis may provide an objective tool to identify those thyroid tumors which, apart from the fatal but rare UDC, are also associated with patient death. Consistent with previous studies,19,26 we have identified activating mutations of all three ras genes in thyroid cancer. We have demonstrated for the first time that H-, K- or N-ras may occur in each of the three differentiation types of thyroid carcinoma. Similar to what has been shown for other tumor types with a high prevalence of oncogenic ras, such as colonic11–13 and pancreatic17 adenocarcinoma, specific mutation patterns may have a different influence on metastatic potential and survival. K-ras codon 13 mutations (all of which involved second nucleotide G-A transitions in codon 13 resulting in Gly>Asp substitution) were a marker for distant metastases and, among patients with differentiated thyroid cancer, poor survival. This molecular change was present in the majority of patients with nucleotide transitions and Gly>Asp mutations, which explains the influence that both alterations have on survival. K-ras codon 13 Gly>Asp mutations have been associated with an increased risk of disease recurrence in a prospective study of colonic adenocarcinoma.11 The K-ras codon 12 Gly>Val mutation, which has been associated with poor survival in colon cancer,13 was present in three of our patients, all of whom had UDC, indicating that it may represent a marker for tumor aggressiveness in thyroid cancer as well. Single activating ras mutations were a marker for poor survival in the differentiated carcinoma group and were present in the majority of the patients with PDC who died as a result of disease (data not shown), whereas multiple mutations did not have an independent influence on survival. The results of this study demonstrate a clear link between ras mutations and poor prognosis. They therefore provide a rational basis for treating thyroid cancer with chemotherapeutic agents that target ras, such as farnesyl transferase inhibitors. The greatest potential for detecting a benefit in clinical trials with these drugs should be in patients with tumors that are highly likely to harbor mutated ras,27 such as poorly and undifferentiated thyroid cancers. The recent demonstration that the farnesyl transferase inhibitor manumycin A inhibits UDC growth both in vitro and in nude mouse xenografts28 is consistent with this hypothesis. In summary, we show that ras mutations are a marker for aggressive thyroid cancer behavior and poor outcome. Although additional investigations and particularly prospective studies are necessary to elucidate further the relationship between ras oncogene activation and thyroid neoplasia, our results indicate that ras genotyping may be of significant value as a prognostic indicator and may provide the rationale for novel treatment modalities.
Supported in part by FIS grants (files 97/5063 and 98/5022) from the Spanish Government to G.G.R. and by a grant from the Thyroid Research Advisory Council of Knoll Pharmaceutical Company (grant SYN 0400 08) to G.T.
1. Bos JL: Ras oncogenes in human cancer: A review. Cancer Res 49:4682–4689, 1989 2. Wright PA, Williams ED, Lemoine NR, et al: Radiation-associated and ‘spontaneous’ human thyroid carcinomas show a different pattern of ras oncogene mutation. Oncogene 6:471–473, 1991[Medline] 3. Challeton C, Bounacer A, Du Villard JA, et al: Pattern of ras and gsp oncogene mutations in radiation-associated human thyroid tumors. Oncogene 11:601–603, 1995[Medline] 4. Suchy B, Waldmann V, Klugbauer S, et al: Absence of RAS and p53 mutations in thyroid carcinomas of children after Chernobyl in contrast to adult thyroid tumours. Br J Cancer 77:952–955, 1998[Medline] 5. Hara H, Fulton N, Yashiro T, et al: N-ras mutation: An independent prognostic factor for aggressiveness of papillary thyroid carcinoma. Surgery 116:1010–1016, 1994[Medline] 6. Rosai J, Carcangiu ML, DeLellis RA: Tumors of the thyroid gland, in Atlas of Tumor Pathology (series 3, fascicle 5). Washington, DC, Armed Forces Institute of Pathology, 1992
7. Garcia-Rostan G, Camp RL, Herrero A, et al: b-catenin dysregulation in thyroid neoplasms: Down-regulation, aberrant nuclear expression and ctnnb1 exon 3 mutations are markers for aggressive tumor phenotypes and poor prognosis. Am J Pathol 158:987–996, 2001 8. Efron B, Tibshirani RJ, An Introduction to the Bootstrap. London, United Kingdom, Chapman & Hall, 1993 9. Carcangiu ML, Zampi G, Rosai J: Poorly differentiated ("insular") thyroid carcinoma: A reinterpretation of Langhans’ "wuchernde struma." Am J Surg Pathol 8:655–668, 1984[Medline] 10. Capella G, Cronauer-Mitra S, Peinado MA, et al: Frequency and spectrum of mutations at codons 12 and 13 of the C-K-ras gene in human tumors. Environ Health Perspect 93:125–131, 1991[Medline] 11. Cerottini JP, Caplin S, Saraga E, et al: The type of K-ras mutation determines prognosis in colorectal cancer. Am J Surg 175:198–202, 1998[CrossRef][Medline]
12. Andreyev HJ, Norman AR, Cunningham D, et al: Kirsten ras mutations in patients with colorectal cancer: The multicenter "RASCAL" study. J Natl Cancer Inst 90:675–684, 1998 13. Andreyev HJ, Norman AR, Cunningham D, et al: Kirsten ras mutations in patients with colorectal cancer: The ‘RASCAL II’ study. Br J Cancer 85:692–696, 2001[CrossRef][Medline]
14. Nelson HH, Christiani DC, Mark EJ, et al: Implications and prognostic value of K-ras mutation for early-stage lung cancer in women. J Natl Cancer Inst 91:2032–2038, 1999 15. Kwiatkowski DJ, Harpole DH Jr, Godleski J, et al: Molecular pathologic substaging in 244 stage I non-small-cell lung cancer patients: Clinical implications. J Clin Oncol 16:2468–2477, 1998[Abstract] 16. Broermann P, Junker K, Brandt BH, et al: Trimodality treatment in stage III nonsmall cell lung carcinoma: Prognostic impact of K-ras mutations after neoadjuvant therapy. Cancer 94:2055–2062, 2002[CrossRef][Medline] 17. Kawesha A, Ghaneh P, Andren-Sandberg A, et al: K-ras oncogene subtype mutations are associated with survival but not expression of p53, p16(INK4A), p21(WAF1), cyclin D1, erbB-2 and erbB-3 in resected pancreatic ductal adenocarcinoma. Int J Cancer 89:469–474, 2000[CrossRef][Medline]
18. Kiyoi H, Naoe T, Nakano Y, et al: Prognostic implication of FLT3 and N-RAS gene mutations in acute myeloid leukemia. Blood 93:3074–3080, 1999
19. Lemoine NR, Mayall ES, Wyllie FS, et al: Activated ras oncogenes in human thyroid cancers. Cancer Res 48:4459–4463, 1988 20. Horie H, Yokogoshi Y, Tsuyuguchi M, et al: Point mutations of ras and Gs alpha subunit genes in thyroid tumors. Jpn J Cancer Res 86:737–742, 1995[CrossRef][Medline] 21. Pilotti S, Collini P, Mariani L, et al: Insular carcinoma: A distinct de novo entity among follicular carcinomas of the thyroid gland. Am J Surg Pathol 21:1466–1473, 1997[CrossRef][Medline] 22. Emanuel JR, Damico C, Ahn S, et al: Highly sensitive nonradioactive single strand conformational polymorphism: Detection of Ki-ras mutations. Diagn Mol Pathol 5:260–264, 1996[CrossRef][Medline] 23. Ezzat S, Zheng L, Kolenda J, et al: Prevalence of activating ras mutations in morphologically characterized thyroid nodules. Thyroid 6:409–416, 1996[Medline]
24. Solakoglu O, Maierhofer C, Lahr G, et al: Heterogeneous proliferative potential of occult metastatic cells in bone marrow of patients with solid epithelial tumors. Proc Natl Acad Sci U S A 99:2246–2251, 2002 25. Harness JK, McLeod MK, Thompson NW, et al: Deaths due to differentiated thyroid cancer: A 46-year perspective. World J Surg 12:623–629, 1988[CrossRef][Medline] 26. Manenti G, Pilotti S, Re FC, et al: Selective activation of ras oncogenes in follicular and undifferentiated carcinomas. Eur J Cancer 30:987–993, 1994
27. Rowinsky EK, Windle JJ, Von Hoff DD: Ras protein farnesyltransferase: A strategic target for anticancer therapeutic development. J Clin Oncol 17:3631–3652, 1999
28. Yeung S-C J, Xu G, Pan J, et al: Manumycin enhances the effect of paclitaxel on anaplastic thyroid carcinoma cells. Cancer Res 60:650–655, 2000 Submitted October 24, 2002; accepted June 9, 2003. This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
2 test for trend). Among the 78 patients of the differentiated carcinoma group (PDC and WDC), ras mutations were identified in 11 of 44 (25%) thyroid carcinomas of papillary histotype and in nine of 34 (26%) of follicular histotype. There was no significant correlation between the type of ras mutations and whether a differentiated thyroid tumor was classified morphologically as papillary or follicular carcinoma, or whether it exhibited oncocytic features. Samples originating from Spain had an overall higher prevalence of activating ras mutation because of the higher proportion of PDC and UDC among the Spanish patients. ras mutations were not associated with patient sex, age, or the presence of lymph node metastases. 
