ANTICANCER RESEARCH 27: (2007)

Transcription

1 Gene Expression Profile of 5-Fluorouracil Metabolic Enzymes in Primary Colorectal Cancer: Potential as Predictive Parameters for Response to Fluorouracil-based Chemotherapy MIZUNOBU KINOSHITA, YASUHIRO KODERA, KENJI HIBI, GORO NAKAYAMA, TAMOTSU INOUE, NORIFUMI OHASHI, YUICHI ITO, MASAHIKO KOIKE, MICHITAKA FUJIWARA and AKIMASA NAKAO Department of Surgery II, Nagoya University Graduate School of Medicine, 65, Tsurumai-cho, Showa-ku, Nagoya, Aichi , Japan Abstract. Background: Thymidylate synthase (TS), dihydropyrimidine dehydrogenase (DPD), thymidine phosphorylase (TP) and orotate phosphoribosyltransferase (OPRT) have been reported to be predictive parameters for the efficacy of fluoropyrimidine-based chemotherapy. Therapy guided by chemotherapy sensitivity and resistance assays may lead to rational treatment decisions. Materials and Methods: mrna expression of TS, DPD, TP, and OPRT were quantified by reverse-transcriptase polymerase chain reaction (RT-PCR) after harvesting from paraffin embedded specimens through microdissection. In vitro chemosensitivity testing by histoculture drug response assay (HDRA) was performed with fresh specimens of the primary tumor from 49 patients with colorectal cancer. Correlations between the results of a chemosensitivity test (the T/C ratio; the percentage of optical density of a tumor treated with anticancer drugs in relation to the optical density of the tumor cultured in RPMI 1640 medium only) and the gene expression were assessed. Results: the gene expression of TS, TP, and OPRT had no correlation with clinicopathological factors, survival, and T/C ratio. The DPD mrna levels (0.295 vs , p=0.2460) and OPRT/DPD ratio (5.535 vs p=0.226) had a weak correlation with the T/C ratio. Of the eleven patients who were actually treated with chemotherapy, the responders had higher OPRT/DPD ratios (8.065 vs , p=0.090). Conclusion: The DPD mrna level and OPRT/DPD ratio evaluated from paraffin embedded specimens are candidates for further evaluation as predictors of response against 5- fluorouracil-based chemotherapy in colorectal cancer. Correspondence to: Yasuhiro Kodera, Department of Surgery II, Nagoya University Graduate School of Medicine, 65, Tsurumaicho, Showa-ku, Nagoya, Aichi , Japan. Tel: , Fax: , ykodera@med.nagoya-u.ac.jp Key Words: Chemotherapy, histoculture drug response assay (HDRA), dihydropyrimidine dehydrogenase (DPD), orotate phosphoribosyltransferase (OPRT). Although surgery remains important in the treatment of colorectal cancer, evidence obtained through phase III trials has pointed to a significant survival benefit conferred by 5- fluorouracil (5FU)-based postoperative adjuvant chemotherapy in the case of stage III cancer (1). 5FU is also used extensively for the treatment of advanced/metastatic disease (2). In addition to the establishment of standards of care, researchers now seek methods to adequately identify patients who may benefit from a given treatment. This could be achieved by various techniques evaluating the malignant potential of individual carcinomas (3) and detecting the minimal residual disease that may lead to relapse (4). In addition, measurement of the enzymes that are related to the metabolism or mechanism of action of drugs may enable the prediction of response to the treatment (5-9). Thymidylate synthase (TS) is an essential DNA synthesizing enzyme that is suppressed by 5-fluoro-deoxyuridine-monophosphate, an active metabolite of fluorouracil (10). Dihydropyrimidine dehydrogenase (DPD) metabolizes and inactivates fluorouracil, and its expression has been reported to correlate with drug resistance as well as toxicity (11). Fluorouracil is converted to active metabolites by phosphorylation through three different pathways, and thymidine phosphorylase (TP) and orotate phosphoribosyltransferase (OPRT) are the key enzymes in two of these pathways (12). The expression levels of these enzymes, alone and in combination, have shown potential as predictive parameters for fluorouracil-based chemotherapy in metastatic colorectal cancer (5-9). In addition to its use for tailoring treatments for patients with advanced colorectal cancer, TS is known also as a prognostic determinant (13). Recently, Danenberg et al. have developed a method to quantify mrna levels of these enzymes in cancer cells selectively harvested from paraffin embedded specimens by means of microdissection (14). In the current study, the mrna levels were quantified by this method in 52 consecutive specimens collected during surgery for colorectal cancer. Although evaluation of the correlation between the mrna expression and the clinical response to the /2007 $

2 chemotherapy was of primary interest, only a fraction of patients have been treated with chemotherapy for metastatic or recurrent disease to date. Additionally, further analysis of the data in relation to the outcome of patients receiving postoperative adjuvant chemotherapy is required. For these reasons, the correlation between the mrna levels and a surrogate marker for response to chemotherapy had to be introduced and evaluated. A strategy that predicts the response to chemotherapy without exploiting specific mechanisms underlying drug resistance is the application of an in vitro chemosensitivity test, in which the viability of cancer cells in relation to the controls is quantified after exposure to anticancer drugs for a predetermined period of time (15). In the current study, sensitivity to the fluorouracil evaluated by the histoculture drug response assay (HDRA), an in vitro chemosensitivity test (16, 17), was considered as a surrogate marker for response to this drug and correlations of drug sensitivity with mrna expression of the four enzymes were evaluated. Materials and Methods Patients. Tissue samples were collected at surgery from 52 patients with colorectal cancer who were recruited during the period from July 2002 to January 2005 and underwent surgery at the Department of Gastroenterological Surgery II, Nagoya University, Japan. The experimental application of the chemosensitivity test had been approved by the Institutional Review Board and written informed consent was obtained from all patients. Fresh specimens were sampled from the primary lesion immediately after surgical resection, washed twice in saline, immersed in Hank's solution, and subjected to in vitro chemosensitivity testing by HDRA. Ten Ìm slices from the paraffin embedded specimens were later used for quantification of TS, DPD, TP, and OPRT mrna by quantitative reverse-transcriptase polymerase chain reaction (RT-PCR). The clinical course of all patients was followed and responses to the chemotherapeutic regimens were recorded for patients who received chemotherapy. Histoculture drug response assay. HDRA was performed as described previously (16). Fluorouracil was purchased from Kyowa Hakko (Tokyo, Japan). Tumor tissue (10 mg) was freshly harvested from each surgically resected specimen, excluding necrotic or infected portions and avoiding contamination with blood or bowel contents, washed in Hank's solution, trimmed of all adipose tissue and placed in a 24-well microplate (Becton Dickinson Labware, Franklin Lakes, NJ, USA) in which 1-cm square gelatin sponges (Upjohn, Kalamazoo, MI, USA) immersed in RPMI 1640 medium (Nihon Seiyaku, Tokyo, Japan) supplemented with 20% fetal calf serum and containing the anticancer agents at various concentrations had been placed. The tissue was then cultured for seven days at 37ÆC with 5% CO 2. A mixed solution of 100 ÌL of 0.6 mg/ml collagenase (230 U/mg; Worthington Biochemical, Freehold, NJ, USA) and 100 ÌL of 0.4% MTT (3-[4,5-dimethyl-2- thiazolyl]-2-5-dimethyl- 2H-tetrazolium bromide) in 0.1 M Na succinate was added, and the optical density at 540 nm was measured using an Easy Reader (SLT-Lab Instruments, Salzburg, Table I. PCR primers and probes. Primer/Probe TS2-764F (18bp) TS2-830R (17bp) Probe TS2-785T (21bp) DPD3a-51F (19bp) DPD3a-134R (20bp) Probe DPD3a-71T (29bp) TP3-700F (17bp) TP3-770R (20bp) Probe TP3-722T (25bp) Sequence GCCTCGGTGTGCCTTTCA CCCGTGATGTGCGCAAT 6FAM-TCGCCAGCTACGC CCTGCTCA AGGACGCAAGGAGGGTTTG GTCCGCCGAGTCCTTACTGA 6FAM-CAGTGCCTACAGTCTCGA GTCTGCCAGTG CCTGCGGACGGAATCCT GCTGTGATGAGTGGCAGGCT 6FAM-CAGCCAGAGATGTGAC AGCCACCGT OPRT5-496F (25bp) TAGTGTTTTGGAAACTGTT GAGGTT OPRT5-586R (20bp) CTTGCCTCCCTGCTCTCTGT Probe OPRT5-528T (27bp) 6FAM-TGGCATCAGTGACCTT CAAGCCCTCCT -actin-592f (18bp) -actin-651r (22bp) Probe b-actin-611t (18bp) TGAGCGCGGCTACAGCTT TCCTTAATGTCACGCACGATTT 6FAM-ACCACCACGGCCGAGCGG Austria) after three hours of incubation. Chemosensitivity was calculated as a ratio of the optical density of a tumor treated with anticancer drugs in proportion to the optical density of the tumor cultured in RPMI 1640 medium only (T/C ratio). The specimens with a T/C ratio of less than 30% were considered as sensitive to fluorouracil. Initially, the chemosensitivity assay for fluorouracil was attempted with three concentrations, 50 Ìg/mL, 150 Ìg/mL and 300 Ìg/mL, based on the results of in vitro testing using several colorectal cancer xenografts (data not shown). The optimal concentration was then determined as 150 Ìg/mL. Microdissection. A representative formalin-fixed, paraffinembedded tumor specimen was selected by a pathologist after examination of hematoxylin and eosin stained slides. Ten Ìm thick sections were stained with neutral fast red to enable visualization of histology for laser capture microdissection (P.A.L.M. Microlaser Technologies AG, Munich, Germany), which was performed to ensure that only tumor cells were studied. RNA extraction, cdna synthesis and RT-PCR. The RNA was isolated from the formalin-fixed, paraffin-embedded (FFPE) specimens using a novel, proprietary procedure (Response Genetics, Los Angeles, CA: United States Patent Number 6,248,535). After RNA isolation, the cdna was derived from each sample and target cdna sequences were amplified by quantitative PCR using a fluorescence-based real-time detection method (ABI PRISM 7900 Sequence Detection System, Applied Biosystems, Foster City, CA) as previously described (14). The 25 ÌL PCR reaction mixture contained 600 nmol/l of each primer (Table π), 200 nmol/l each of datp, dctp, and dgtp, 400 Ìmol/L dutp, 852

3 Kinoshita et al: 5FU Metabolic Enzyme Gene Expression in CRC 5.5 mmol/l MgCl 2 and 1x Taqman buffer A containing a reference dye (all reagents were supplied by Applied Biosystems, Foster City, CA). The PCR conditions were 50ÆC for 10 seconds and 95ÆC for 10 minutes, followed by 42 cycles at 95ÆC for 15 seconds and 60ÆC for one minute. Each reported gene expression value represents the average of a minimum of three separate PCR reactions. Statistical analyses. Differences in the T/C ratio between different age groups (<60 versus 60 years of age), gender, clinical stages and location of the tumor were evaluated by Student's t-test. Pearson's regression coefficient was calculated to evaluate the correlation between the mrna expressions of the various enzymes. The difference in survival between sensitive and resistant groups was evaluated by log-rank test. Results Patient characteristics. The HDRA at the relevant fluorouracil concentration of 150 Ìg/mL was successfully performed in 49 out of 52 patients (35 males and 14 females; with the median age of 66 years, range, years). The HDRA of three patients failed due to infection of the samples. The location of the tumor was the cecum and ascending colon in 12 patients, the transverse colon in five patients, the descending, sigmoid and rectosigmoid colon in 15 patients and the rectum in 17 patients. The clinical stage of the patients determined by the TNM classification was Stage I in three, Stage II in 17, Stage IIIA in 12, Stage IIIB in six, and Stage IV in 11 patients. Surgery with curative intent was performed in 34 patients, whereas 15 patients underwent palliative resection. The gene expressions of TS, DPD, TP, and OPRT were successfully measured in 48, 40, 47, and 49 out of the 49 patients, respectively. The gene expressions of TS, DPD, TP and OPRT did not correlate with the age or gender of the patients, location of the tumor or clinical stage, with the exception that the DPD mrna level was significantly lower in the rectosigmoid and the rectal cancer compared with the transverse colon cancer (0.270 and vs , p=0.003 and 0.043), lower in Stage II cancer compared with Stage IV cancer (0.248 vs , p=0.002) and that the gene expression of OPRT was significantly lower in women compared with men (1.024 vs , p=0.002). There was significant correlation between the gene expression of DPD and that of TP (p<0.001, R=0.602), whereas no correlation was observed between OPRT and TS or any other pairs of the four 5FU-related enzymes gene expressions (Figure 1). The HDRA identified 13 potential responders and 36 non-responders. The differences in mrna expression of TS, TP and OPRT were not significant between the potential responders and the non-responders. The mrna expression of DPD had a weak tendency to be lower among the responders although the difference did not reach statistical significance (0.295 vs , p=0.246) (Figure 2). Figure 1. Correlation between mrna expression levels of DPD and TP. DPD, dehydropyrimidine dehydrogenase; TP, thymidine phosphorylase. There was a significant correlation between the gene expression of DPD and that of TP (p<0.001, R=0.602), whereas no correlation was observed between OPRT and TS or any other pairs of the four 5FU-related enzymes gene expressions. The difference in the OPRT/DPD ratio also had a weak tendency to be higher for the responders (5.535 vs , p=0.226). Association of chemosensitivity and mrna with the outcome of the patients. A partial response was observed in two out of 11 patients who were treated with chemotherapy for metastatic disease. The partial response was observed in one out of three potential responders as predicted by the chemosensitivity test (T/C ratio of less than 30%) and one out of eight non-responders (T/C ratio exceeding 30%). DPD mrna levels (0.190 vs , p=0.108) tended to be lower and the OPRT/DPD ratio (8.065 vs , p=0.090) higher for the actual responders (Figures 3 and 4). Discussion Intratumoral gene expression and the activities of several enzymes related to the metabolism of 5FU have been shown to correlate with the sensitivity to this key drug in the treatment of colorectal cancer. A high OPRT mrna level or OPRT/DPD mrna ratio has been shown to be associated with prolonged survival in patients receiving oral tegafururacil and leucovorin for metastatic disease (8) and those receiving intraarterial chemotherapy for liver metastases (6), while high OPRT activity and low DPD activity has been shown to correlate with chemosensitivity as defined by in vitro chemosensitivity tests (19). In the current study, the mrna levels were evaluated from the paraffin-embedded specimens by a method involving microdissection, extraction of RNA and RT-PCR. This technique was found to be successful in terms 853

4 Figure 2. mrna expression levels in HDRA identified potential responders and non responders. HDRA, histoculture drug response assay; TS, thymidylate synthase; DPD, dehydropyrimidine dehydrogenase; TP, thymidine phosphorylase; OPRT, orotate phosphoribosyltransferase. HDRA identified 13 potential responders and 36 non-responders. Differences in mrna expression of TS, TP and OPRT were not significant between the potential responders and non-responders. The mrna expression of DPD had a weak tendency to be lower among the potential responders although the difference did not reach a statistical significance (0.295 vs , p=0.246). of high detection rates and may in future enable retrospective evaluation of the expression of various other relevant mrnas using archival specimens. While the microdissection technique allows the analysis to be confined strictly to neoplastic cells, this could be misleading for evaluating signal transduction pathways such as cytokine/receptor systems that involve interactions with the microenvironment. The possible role of TP in angiogenesis (20), for instance, had to be disregarded in the current study. When discussing the mrna expression of an enzyme, questions are always raised as to whether the mrna level correlates with the actual enzymatic activity. Evaluation of the enzymatic activities usually calls for a large amount of fresh samples, however, and the results are highly dependent on how promptly the samples were collected and stored. It is for this reason that mrna was the target in the current study. The DPD mrna was found to be strongly expressed in Stage IV disease and transverse colon cancer and the OPRT mrna expression was significantly higher in men. Coregulation of DPD and TP as reported by Inokuchi et al. (18) was suggested also in the current study, whereas the similar correlation between TS and OPRT reported by the same authors was lacking. Attempts were made to evaluate the correlation between mrna levels of the 5FU-related enzymes and the response to chemotherapy. The clinical stages of the patients involved were varied in the current series, however, and overall survival of the patients may not reflect benefits conferred by chemotherapy. Thus, a surrogate marker for the efficacy of the fluoropyrimidines was needed and the T/C ratio evaluated by in vitro chemosensitivity test was employed. The T/C ratio evaluated by HDRA has been observed to predict response to 854

5 Kinoshita et al: 5FU Metabolic Enzyme Gene Expression in CRC Figure 3. mrna expression in chemotherapy treated patients. TS, thymidylate synthase; DPD, dehydropyrimidine dehydrogenase; TP, thymidine phosphorylase; OPRT, orotate phosphoribosyltransferase. A partial response was observed in two out of 11 patients who were treated with chemotherapy for metastatic disease. DPD mrna levels tended to be lower for the actual responders (0.190 vs , p=0.108). Figure 4. OPRT/DPD of chemotherapy treated patients. Responders and non-responders identified by HDRA (histoculture drug response assay). PD, progressive disease; PR, partial response; DPD, dehydropyrimidine dehydrogenase; OPRT, orotate phosphoribosyltransferase. The difference in OPRT/DPD ratio had a weak tendency to be higher for the potential responders as determined by histoculture drug response assay (5.535 vs p=0.226) and for the actual responders (8.065 vs , p=0.090). 855

6 adjuvant chemotherapy in stage III gastric cancer (17) and seemed to be predictive also for esophageal cancer (16). However, formal validation that the in vitro chemosensitivity test accurately predicts outcome needs a well-designed prospective phase III trial that has not been performed to date, and this is a fundamental weakness of this study. In addition, mrna levels and T/C ratios have been evaluated with primary carcinomas, whereas the targets of chemotherapy for recurrent or metastatic disease are usually metastatic lesions. The profile of 5FU-related enzymes in the primary could be different from that of the metastatic lesion (21) and the predictive value of the chemosensitivity test could be limited except in the adjuvant setting. Nevertheless, weak but promising correlations of the DPD mrna levels and the OPRT/DPD ratios were observed with the T/C ratios against 5FU and also with the clinical response following chemotherapy as evaluated in a small proportion of the patients. To summarize, evaluation of the OPRT/DPD ratio by a technique involving microdissection and RT-PCR may predict the outcome of patients with colorectal cancer who undergo 5FU-based chemotherapy. However, a prospective study testing the current assay for the mrnas needs to be done in a larger group of patients, perhaps in those who undergo a pivotal clinical trial for adjuvant chemotherapy, to confirm its usefulness. Acknowledgements Authors are indebted to Ms. Sawako Kato for her excellent technical assistance. References 1 Leichman CG: Adjuvant therapy for colon cancer 2005: new options in the twenty-first century. Surg Oncol Clin N Am 15: , Moertel CG: Chemotherapy for colorectal cancer. New Engl J Med 330: , Elsaleh H, Powell B, McCaul K et al: p53 alteration and microsatellite instability have predictive value for survival benefit from chemotherapy in stage III colorectal carcinoma. Clin Cancer Res 7: , Liefers GJ, Cleton-Jansen AM, Van de Velde CJH et al: Micrometastases and survival in stage II colorectal cancer. N Engl J Med 339: , Adlard JW, Richman SD, Seymour MT and Quirke P: Prediction of the response of colorectal cancer to systemic therapy. Lancet Oncol 3: 75-82, Matsuyama R, Togo S, Shimizu D et al: Predicting 5-fluorouracil chemosensitivity of liver metastases from colorectal cancer using primary tumor specimens: Three-gene expression model predicts clinical response. Int J Cancer 119: , Ichikawa W, Uetake H, Shirota Y et al: Combination of dihydropyrimidine dehydrogenase and thymidylate synthase gene expressions in primary tumors as predictive parameters for the efficacy of fluoropyrimidine-based chemotherapy for metastatic colorectal cancer. Clin Cancer Res 9: , Ichikawa W, Uetake H, Shirota Y et al: Both gene expression for orotate phosphoribosyl transferase and its ratio to dihydropyrimidine dehydrogenase influence outcome following fluoropyrimidine-based chemotherapy for metastatic colorectal cancer. Br J Cancer 89: , Ochiai T, Mishimura K, Noguchi H et al: Prognostic impact of orotate phosphoribosyl transferase among 5-fluorouracil metabolic enzymes in resectable colorectal cancers treated by oral 5-fluorouracil-based adjuvant chemotherapy. Int J Cancer 118: , Danenberg PV: Thymidylate synthase-a target enzyme in cancer chemotherapy. Bichim Biophys Acta 473: 73-92, Heggie GD, Sommadossi JP, Cross DS et al: Clinical pharmacokinetics of 5-fluorouracil and its metabolites in plasma, urine, and bile. Cancer Res 47: , Peters GJ, Laurensse E, Leyva A et al: Sensitivity of human, murine, and rat cells to 5-fluorouracil and 5'-deoxy-5- fluorouridine in relation to drug-metabolizing enzymes. Cancer Res 46: 20-28, Edler D, Kressner U, Ragnhammar P et al: Immunohistochemically detected thymidylate synthase in colorectal cancer: An independent prognostic factor of survival. Clin Cancer Res 6: , Salonga D, Danenberg KD, Johnson M et al: Colorectal tumors responding to 5-fluorouracil have low gene expression levels of dihydropyrimidine dehydrogenase, thymidylate synthase and thymidine phosphorylase. Clin Cancer Res 6: , Samson DJ, Seidenfeld J, Ziegler K et al: Chemotherapy sensitivity and resistance assays: a systematic review. J Clin Oncol 22: , Suda S, Akiyama S, Sekiguchi H et al: Evaluation of the histoculture drug response assay as a sensitivity test for anticancer agents. Surgery Today 32: , Abe S, Kubota T, Matsuzaki SW et al: Chemosensitivity test is useful in evaluating the appropriate adjuvant cancer chemotherapy for stage III non-scirrhous and scirrhous gastric cancers. Anticancer Res 19: , Inokuchi M, Uetake H, Shirota Y et al: Gene expression of 5- fluorouracil metabolic enzymes in primary colorectal cancer and corresponding liver metastasis. Cancer Chemother Pharmacol 53: , Isshi K, Sakuyama T, Gen T et al: Predicting 5-FU sensitivity using human colorectal cancer specimens: comparison of tumor dihydropyrimidine dehydrogenase and orotate phosphoribosyl transferase activities with in vitro chemosensitivity to 5-FU. Int J Clin Oncol 7: , Toi M, Rahman MA, Bando H and Chow LWC: Thymidine phosphorylase (platelet-derived endothelial-cell growth factor) in cancer biology and treatment. Lancet Oncol 6: , Kuramochi H, Hayashi K, Uchida K et al: 5-Fluorouracilrelated gene expression levels in primary colorectal cancer and corresponding liver metastasis. Int J Cancer 119: , Received November 27, 2006 Revised February 2, 2007 Accepted February 6,