Polymorphisms in the Thymidylate Synthase Gene Predict Response to 5-Fluorouracil Therapy in Colorectal Cancer

Abstract

Introduction Advances in the field of pharmacogenomics have enhanced our knowledge of the molecular basis of cancer and provided a link between detection of genetic polymorphisms and response to treatment. Different variations of genetic polymorphisms, such as tandem repeats, single nucleotide polymorphisms, or nucleotide deletions, can lead to alterations in the transcription, translation, or stability of certain gene products that could be critical for the clinical activity and/or toxicity of chemotherapeutic drugs. One such target is thymidylate synthase (TS), an enzyme that catalyzes the reductive methylation of deoxyuridine-5'monophosphate (dUMP) to deoxythymidine-5'-monophosphate (dTMP) in the cell and generates the sole intracellular de novo source of thymidylate, a key precursor needed for DNA biosynthesis. This folate-dependent enzyme is an important target for chemotherapy drugs such as 5fluorouracil (5-FU), fluorodeoxyuridine (FUDR), oral 5-FU prodrugs (eg, uracil/tegafur, S-1, and capecitabine), and other novel folate-based drugs (eg, raltitrexed and pemetrexed). These agents directly inhibit TS enzymatic function, thereby resulting in thymidylate depletion with subsequent DNA damage and cell death. In the treatment of colorectal cancer (CRC), 5-FU–based chemotherapeutic regimens have yielded overall response rates ranging from 10% to 30% for advanced disease1 and provide approximately 15% improvement in 5year survival as adjuvant therapy for stage III disease.2 One key question that remains, however, is how to best predict which patients will benefit from such a therapeutic treatment strategy. Several studies have observed that high levels of TS expression at the protein or messenger RNA (mRNA) level, in the primary tumor or at the metastatic sites, have correlated with a lack of response to 5-FU–based chemotherapy. In contrast, low expression levels of TS have been associated with a positive clinical response. In order for these drugs to be effective in killing a tumor cell, active enzyme levels have to be reduced for a sufficient period for cell cycle arrest to be initiated. If TS expression is increased, then standard drug doses would be ineffective in attenuating enzyme activity. This reasoning has been supported in a retrospective analysis of TS levels, as measured by immunohistochemistry, in tumor specimens of patients with advanced CRC. Among 48 patients who received 5-FU–based therapy, the overall response rate was lower (24% vs. 67%; P = 0.003), and median time to progression (6.2 months vs. 9.6 months; P = 0.005) and median overall survival (OS) times (15.4 months vs. 18.4 months; P = 0.02) were shorter in patients with high tumor TS expression compared with patients with low TS expression.3 Discrepancies have been reported in TS expression levels detected in the primary colorectal tumor versus hepatic metastases. In 14 patients with CRC and hepatic metastases matched samples, TS mRNA levels, as measured by semiquantitative reverse-transcriptase polymerase chain reaction, were higher in the primary tumor (mean, 1.14 ± 0.34) compared with the liver metastases (mean, 0.76 ± 0.37; Wilcoxon test; P < 0.05).4