|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Thymidylate Synthase Protein Expression in Primary Colorectal Cancer: Lack of Correlation With Outcome and Response to Fluorouracil in Metastatic Disease SitesFrom the Cancer Research Centre, Department of Oncology, Queen’s University Belfast, Belfast, Northern Ireland; Clinical Investigations Program, Robert H. Lurie Comprehensive Cancer Center of Northwestern University, Chicago, IL; Eastern Cooperative Oncology Group Statistical Center, Department of Biostatistical Science, Dana-Farber Cancer Institute, Boston, MA; Northwestern Memorial Hospital, Chicago, IL; University of Pennsylvania, Philadelphia, PA; and Medicine Branch, National Cancer Institute, Bethesda, MD. Address reprint requests to Patrick G. Johnston, MD, PhD, Cancer Research Centre, Department of Oncology, Queen’s University Belfast, University Floor, Belfast City Hospital, 97 Lisburn Road, Belfast BT9 7AB, Northern Ireland; email: oncology{at}qub.ac.uk.
Purpose: The aim of this study was to investigate the utility of quantitating thymidylate synthase (TS) in the primary tumor as a surrogate for metastatic disease sites to predict the likelihood of response and outcome to fluorouracil (FU) treatment in patients with metastatic colorectal cancer. Methods: TS protein expression was evaluated using the TS 106 antibody and the avidin biotin labeling immunohistochemical technique in primary tumor samples from 219 patients with metastatic colorectal cancer. The patients were a representative sample of those patients enrolled into the Eastern Cooperative Oncology Group E2290 protocol that evaluated five separate FU-containing regimens in patients with metastatic residual or recurrent colorectal carcinoma. Results: Our retrospective analysis found that the level and extent of TS protein expression in the primary tumor did not correlate with overall survival in patients with metastatic or recurrent colorectal cancer. A trend toward a direct correlation between the level of TS protein expression and response was noted in tumors that expressed high TS levels. This response advantage for patients expressing high TS levels in the primary tumor was apparent regardless of what FU-based treatment the patient received but was most apparent in the subgroup treated with leucovorin, in which the level of TS expression and response to FU and leucovorin reached statistical significance (P = .034). No significant interaction could be detected between the addition of leucovorin to FU and the level of TS expression in the primary tumor. Conclusion: This study demonstrated that measurement of TS protein levels in the primary tumor tissue does not aid in predicting outcome or response to FU in a metastatic disease site. These assays must be performed on biopsy tissue from the metastatic disease site that is used to radiologically assess response and outcome to treatment.
COLORECTAL CANCER is a highly treatable and often curable disease when localized to the bowel. It is the second most frequently diagnosed malignancy and the second most common cause of cancer death in the western world.1 Surgery is the primary treatment and results in cure in approximately 50% of patients. Recurrence following surgery is a major problem and often is the ultimate cause of death. The use of chemotherapy in patients with advanced or recurrent metastatic colorectal cancer has been suboptimal and despite several decades of active investigation, fluorouracil (FU) still remains one of the most active agents used in the treatment of this disease. Thymidylate synthase (TS), which catalyzes the conversion of 2-deoxyuridine 5-monophosphate and 5–10-methylene tetrahydrofolate to deoxythymidine monophosphate, is an important target for fluoropyrimidine drugs.2,3 Inhibition of TS along and incorporation of fluorodeoxyuracil triphosphate incorporation directly into DNA result in disruption of DNA synthesis and repair. Several lines of evidence now support the concept that insufficient TS inhibition may represent a major mechanism of resistance to fluoropyrimidines as well as the newer folate-based TS inhibitors.4–7 Moreover, studies examining the importance of TS expression indicate that TS expression predicts overall outcome and response to cytotoxic therapy in several major tumor types. In the adjuvant setting TS is one of the most promising predictors of disease-free survival and overall survival (OS) in colorectal cancer.8,9 In the setting of advanced metastatic disease, both high levels of TS mRNA and TS protein in malignant tissue from the metastatic disease site predict poor response to fluoropyrimidine-based therapy.10,11 This is also true for other solid tumors such as breast, gastric, and head and neck cancer, in which FU is part of the chemotherapy treatment. These studies suggest that the ability to predict response and outcome on the basis of TS expression may provide the opportunity to select those patients most likely to benefit from TS-directed therapy.12 In this study we wished to determine whether immunohistochemical quantitation of TS protein expression in the tissue from a patient’s primary colon tumor could be used as a surrogate for the metastatic disease sites to predict the likelihood of response or determine the outcome of FU-based chemotherapy treatment for metastatic disease.
A total of 1,120 eligible patients with residual or recurrent colorectal cancer were entered onto an Eastern Cooperative Oncology Group/Cancer and Leukemia Group B (ECOG/CALGB) protocol E2290/CALGB9092.13 This investigation was designed to study mechanism-directed regimens in maximizing the efficacy of FU in advanced colorectal cancer. This study was conducted between November 1990 and 1995 and included an evaluation of FU versus FU plus N-(phosphonacetyl)-L-aspartate, FU plus oral leucovorin, FU plus intravenous leucovorin, or FU plus interferon alpha (Fig 1  ). Patients were required to have histologically confirmed metastatic residual or recurrent adenocarcinoma of the colon or rectum with unresectable tumor.
The TS study samples consisted of tissue sections from paraffin-embedded tissues available from 219 ECOG patients entered onto the E2290 study. These represented all available tissue sections from the E2290 protocol. These samples were stored at the Department of Pathology, Northwestern University (Chicago, IL). The laboratory data were merged with the clinical and pathological data kept separately in the ECOG Coordinating Center (Brookline, MA). The study was approved by the institutional review board.
TS Immunohistochemistry Tissue specimens were deparaffinized in 100% xylene and rehydrated through graded alcohol solutions, rinsed in phosphate-buffered saline (PBS) for 5 minutes, then in distilled H2O for 5 minutes. Endogenous peroxidase activity was inhibited by incubating the slides in 3% H202 for 10 minutes, followed by a 5-minute rinse in distilled H2O and a 5-minute rinse in PBS. The tissues were blocked with 2% horse serum for 30 minutes to reduce nonspecific staining and then incubated for 60 minutes with TS primary antibody at a 1:500 dilution. The slides were washed four times with PBS for 10 minutes and then incubated in biotinylated goat antimouse secondary antibody for 30 minutes. The slides were again washed four times with PBS buffer and then incubated with avidin-biotin complex for 30 minutes. After slides were washed, the chromogen, diaminobenzidine (0.7 mg/mL), was applied for 5 minutes. After another wash cycle, the tissues were counterstained with Mayer’s hematoxylin. The slides were rinsed, dehydrated in graded alcohol solutions, fixed in xylene, and mounted on glass cover slips with permount (ProSciTech, Kelso, Queensland, Australia). To ensure consistent staining, a control adenocarcinoma sample with well-characterized staining was included in each staining run. To control against the possibility of nonspecific staining, each sample was stained with a nonspecific mouse immunoglobulin G as a primary antibody (Vector Laboratories, Burlingame, CA) at a 1:10,000 dilution. Each slide was analyzed to verify that staining was absent.
Tissue Evaluation
Statistical Analysis
Comparison of TS Study Population to Overall E2290 Study Two hundred and nineteen (n = 219) patient tumor samples were evaluable for TS staining from the E2290 cohort. This represented 28% of the overall E2290 study population. The patient characteristics for those cases included in the TS analysis were compared to the remaining patients from the E2290 population. The comparison demonstrated no difference between the two groups including analysis by treatment arm, sex, tumor differentiation, age, obstruction, progression-free survival, and survival. A comparison of the TS study patient characteristics and the remaining E2290 population is outlined in Table 1 .
Pathologic Correlations The TS staining in the primary colorectal tumor specimens was predominantly a granular cytoplasmic staining pattern noted in the tubular and cribriform glands of the colorectal cancers. TS staining was also present in lymphocytes adjacent to the tumors and some minor cross-reactivity was observed with smooth muscle cells. Heterogeneity of TS staining was observed both within individual tumors and between different tumors. Of the 219 tumor samples studied, 44 (20%) demonstrated low TS staining, whereas 175 (80%) tumors had high levels of TS staining (Table 2 ). The TS staining intensity was not significantly associated with histologic grade or differentiation (P > .1).
Clinical Correlations TS expression and the distribution of TS staining (F v D) in the primary tumor did not correlate with overall survival in these patients with metastatic or recurrent disease (P > .1). Of the 44 patients with tumors that expressed low TS, 27% remained alive at 2 years, whereas among the 175 patients with tumors that expressed high TS, only 22% were alive (P > .1; Fig 2 ). Thus, in the setting of recurrent or metastatic disease, OS was independent of the level and extent of TS expression in the primary tumor.
The pattern of response at the metastatic disease site was also correlated to the level of TS expression in the primary tumor sample. A trend toward a direct correlation between the level of TS expression and response was noted in the tumors that expressed high TS. Of the 44 patients with low TS levels in the primary tumor, only one patient (2%) demonstrated a response, compared with 20 of 175 patients (11%) with high TS expression (P = .084; Table 3 ). This response advantage for patients expressing high TS levels in the primary tumor was apparent regardless of what FU-based treatment the patient received but was most apparent in the subgroup treated with leucovorin, in which the level of TS expression and response to FU and leucovorin reached statistical significance (P = .034; Table 4).
We also examined whether any interaction could be detected between TS expression levels in the primary tumor and the addition of leucovorin to FU treatment. This comparison was done by comparing the FU-alone and FU and leucovorin arms. No significant interaction could be detected between the addition of leucovorin to FU and the level of TS expression in the primary tumor. The leucovorin treatment by TS level were also examined using exact logistic regression models with appropriate interaction terms in the models and these interactions also were not significant (P > .1).
TS plays a key role in DNA nucleotide precursor synthesis and represents an important therapeutic target for FU-based therapy. Several recent preclinical and clinical studies have indicated that the level of TS expression in gastrointestinal tumors represents one of the most promising predictors of response to FU-based chemotherapy in advanced colorectal cancer. These studies have demonstrated an inverse correlation between the level of TS expression and response to FU.10–12,14 In these studies the level of TS expression was measured in tumor tissue from metastatic disease sites using reverse transcriptase polymerase chain reaction or immunohistochemical methodologies. More recent studies have examined the clinical importance of TS gene polymorphisms (two TSER*2 or three TSER*3 28 base pairs tandem repeats) in the TS promotor region.15 These studies have demonstrated that three copies of the tandem repeat generate approximately 2.6-fold higher TS expression than two copies. Moreover, preliminary clinical studies have suggested that stage III colorectal cancer patients with the TSER*3/TSER*3 genotype do not derive a survival benefit from adjuvant FU treatment that is observed in patients with the TSER*2/TSER*2 or TSER*2/TSER*3 genotypes.15 This study examined whether the expression of TS in the primary tumor could be used as a surrogate to measuring TS expression from the metastatic disease site. This study included primary tumor samples from 219 patients treated on the ECOG protocol E2290. The original treatment study evaluated five separate FU treatment regimens in patients with histologically confirmed metastatic residual or recurrent adenocarcinoma of the colon or rectum. The key findings from our analysis were that no association could be found between TS expression in the primary tumor and OS. Although there was a trend toward a better response in the metastatic disease site in patients whose primary tumors had high TS levels, this did not reach significance. These results were consistent across all treatment arms and reached significance only in those patients treated with FU and leucovorin. The lack of association between TS levels in the primary tumor and patient survival after recurrence of the disease in metastatic sites is not surprising and most likely relates to an altered molecular phenotype and clinical behavior in the disease at the metastatic site. However, the trend toward an improved response rate at the metastatic site on the basis of high TS levels in the primary tumor is more difficult to explain and is in complete contradiction to those studies that have demonstrated a clear inverse relationship between high TS expression and FU responsiveness.10–12,14 This may be a chance finding; however, it is possible that those patients whose primary disease expresses high TS levels may be more likely to relapse with FU-responsive disease, possibly because of lower TS expression in the metastatic clone. This could only be verified by a large study designed to examine TS in both the primary tumor and the metastatic disease site at the time of recurrence. To date relatively few reports have examined the correlation of TS expression between primary tumor versus metastatic disease sites in colorectal cancer. Aschele et al16 measured TS immunohistochemically in 18 patients with both primary tumors and corresponding hepatic metastases. They found that TS expression did not correlate in 11 of the 18 patients, and that the level of TS immunostaining was lower in the hepatic metastases than the primary tumor in 10 of the 18 patients.16 Other investigators have examined TS expression in various metastatic disease sites and noted significant variability in TS levels on the basis of the metastatic disease site. Gorlick et al17 reported that TS expression was significantly higher in pulmonary compared with hepatic metastases from colorectal cancer patients, and Cascinu et al18 observed higher TS levels in abdominal masses compared with liver metastases from colorectal tumors. This may explain why those patients who present with lung metastases or significant abdominal disease tend to be refractory to FU treatment. In addition, a study by Findlay et al19 found that TS levels in primary colorectal tissue do not predict response to FU chemotherapy in metastatic disease sites. Taken together, these studies indicate that metastases from a colorectal tumor do not necessarily contain similar levels of TS enzyme compared with TS levels in the primary tumor tissue. These findings also may be relevant to those studies performed in the adjuvant disease setting in which TS is measured in the primary tumor. In these patients with locally advanced colorectal cancer, the micrometastases-like metastatic disease sites may have different levels of TS compared with levels in the primary tumor. As a consequence, the use of primary tissue to predict an effect on the micrometastases in the adjuvant setting may be flawed because it may be similar to the situation in patients with advanced disease as demonstrated by this study. This may help explain the disparate and confusing results that have been obtained in the adjuvant disease studies relating to a treatment by outcome interaction with TS.20 The fact that such variability and heterogeneity of TS expression exists between differing tissues underlies this lack of association and our inability to use the primary tumor tissue to predict responsiveness in the metastatic disease setting. This variability in expression has also been demonstrated for other biologic markers such as BCL-2–associated X protein, a member of the BCL-2 protein family.21 The reasons for such variability are complex but may relate to the role of TS in cell cycle regulation and cellular turnover as well as its role in drug resistance. It may also relate to a changing molecular phenotype in the tumor as the cancer develops the ability to metastasize to distant sites. Moreover, the semiquantitative nature of the immunohistochemical technique, the age of the tissues stained, the staining technique, the TS antibody used for TS analysis, and the interobserver variation that occurs in the immunohistochemical evaluation of marker expression studies also may account for some of the variation and heterogeneity seen. This study indicates that primary colorectal tissue cannot be used as a reliable surrogate for tissue from metastatic disease sites to predict response to FU-based regimens. Therefore, to predict the likelihood of response and outcome to FU, TS analysis must be performed on biopsy material obtained from the metastatic site or sites. The feasibility of performing these types of analyses in the clinical setting on patients with metastatic colorectal cancer is not straightforward. It requires an invasive procedure, multiple biopsies, and exposure of patients to increased risk of morbidity. Moreover, laboratory support with the expertise and proper quality assurance to run these tests must be available. Nonetheless, these tests could have a positive impact in that the use of FU-based chemotherapy could be targeted to those individuals most likely to benefit, and those patients unlikely to benefit could be spared the harmful side effects of FU-based chemotherapy. In conclusion, this study has demonstrated that TS levels in the primary tumor tissue do not aid in predicting outcome or response to FU in a metastatic disease site. TS is an important predictor of response to FU-based chemotherapy regimens, however, these assays must be performed on biopsy tissue from the metastatic disease site that is used to radiologically assess response and outcome of FU treatment.
1. Johnston PG, Allegra CJ: Colorectal cancer biology: Clinical implications. Semin Oncol 22:418–432, 1995[Medline] 2. Danenberg PV: Thymidylate synthase: A target enzyme in cancer chemotherapy. Biochim Biophys Acta 473:73–92, 1997 3. Grem, JL: Fluorinated pyrimidines, in Chabner BA, Collins, JM (eds): Cancer Chemotherapy, Principles and Practice. Philadelphia, PA, Lippincott, 1990, pp 180–224
4. Johnston PG, Drake JC, Trepel J, et al: The immunological quantitation of thymidylate synthase using the monoclonal antibody TS106 in 5-FU sensitive and resistant human cancer cell lines. Cancer Res 52:4306–4312, 1992 5. Washtein WL: Increased levels of thymidylate synthase in cells exposed to fluorouracil. Mol Pharmacol 25:171–177, 1984[Abstract] 6. Jackman AL, Kelland LR, Kimbell R, et al: Mechanisms of acquired resistance to the quinazoline thymidylate synthase inhibitor ZD1694 (Tomudex) in one mouse and three human cell lines. Br J Cancer 71:914–924, 1995[Medline] 7. Drake JC, Allegra CJ, Moran RG, et al: Resistance to Tomudex (ZD1694) multifactorial in human breast and colon carcinoma cell lines. Biochem Pharmacol 51:1349–1355, 1996[CrossRef][Medline]
8. Johnston PG, Fisher ER, Rockette HE, et al: The role of thymidylate synthase expression in prognosis and outcome of adjuvant chemotherapy in patients with rectal cancer. J Clin Oncol 12:2640–2647, 1994 9. Lenz H-J, Danenberg KD, Leichman CG, et al: p53 and thymidylate synthase expression in untreated stage II colon cancer: Associations with recurrence, survival and site. Clin Cancer Res 4:1227–1234, 1998[Abstract] 10. Leichman CG, Lenz H, Leichman L, et al: Quantitation of intratumoral thymidylate synthase expression predicts for disseminated colorectal cancer response and resistance to protracted-infused fluorouracil and weekly leucovorin. J Clin Oncol 15:3223–3229, 1997[Abstract]
11. Johnston PG, Lenz HJ, Danenberg KD, et al: Thymidylate synthase protein and gene expression predicts for response to 5-fluorouracil leucovorin chemotherapy in patients with colorectal and gastric cancer. Cancer Res 55:1407–1412, 1995 12. Chu E, Grem JL, Johnston PG, et al: New concepts for the development and use of antifolates. Stem Cells 14:41–46, 1996[Abstract]
13. O’Dwyer PJ, Manola J, Valone FH, et al: Fluorouracil modulation in colorectal cancer: Lack of improvement with N-phosphonoacetyl-L-aspartic acid or oral leucovorin or interferon, but enhanced therapeutic index with weekly 24-hour infusion schedule. An Eastern Cooperative Oncology Group/Cancer and Leukemia Group B Study. J Clin Oncol 19:2413–2421, 2001 14. Paradiso A, DeLena M, Stea B, et al: Thymidylate synthase and p53 primary tumor expression as predictive factors for advanced colorectal cancer patients. Br J Cancer 82:560–567, 2000[CrossRef][Medline] 15. Marsh S, McLeod HL: Thymidylate synthase pharmacogenomics in colorectal cancer. Clin Colorectal Cancer 1:175–178, 2001[Medline]
16. Aschele C, Debernardis D, Tunesi G, et al: Thymidylate synthase protein expression in primary colorectal cancer compared with corresponding distant metastases and relationship to clinical response to 5-fluorouracil. Clin Cancer Res 6:4797–4802, 2000
17. Gorlick R, Metzger R, Danenberg KD, et al: Higher levels of thymidylate synthase gene expression are observed in pulmonary as compared to hepatic metastases of colorectal adenocarcinoma. J Clin Oncol 16:1465–1469, 1998
18. Cascinu S, Aschele C, Barni S, et al: Thymidylate synthase protein expression in advanced colon cancer: Correlation with site of metastases and clinical response to leucovorin modulated bolus 5-fluorouracil. Clin Cancer Res 5:1996–1999, 1999 19. Findlay MP, Cunningham D, Morgan G, et al: Lack of correlation between thymidylate synthase levels in primary colorectal cancer and subsequent response to chemotherapy. Br J Cancer 75:903–909, 1997[Medline]
20. Allegra CJ, Parr Al, Wold LE, et al: Investigation of the prognostic and predictive value of thymidylate synthase, p53, Ki-67 in patients with locally advanced colon cancer. J Clin Oncol 20:1735–1743, 2002
21. Jansson A, Sun X-F: Bax expression decreases significantly from primary tumor to metastasis in colorectal cancer. J Clin Oncol 20:811–816, 2002 Submitted July 8, 2002; accepted October 1, 2002.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
-2a, recombinant interferon alpha-2a.
2 test or Fisher’s exact test when appropriate. All P values presented are two-sided. 
