Search

Nucleotide excision repair (NER) and base excision repair (BER) pathways are DNA repair pathways that are important in carcinogenesis and in response to DNA-damaging chemotherapy. ERCC1 and ERCC2 are important molecular markers for NER; XRCC1 and PARP1 are important molecular markers for BER. Functional polymorphisms have been described that are associated with altered expression levels of these genes and with altered DNA repair capability. We assayed genomic DNA from 156 Americans of European descent and 164 Americans of African descent for the allelic frequencies of specific polymorphisms of ERCC1 N118N (500C>T), ERCC1 C8092A, ERCC2 K751Q (2282A>C), XRCC1 R399Q (1301G>A), XRCC1 R194W (685C>T), and PARP1 V762A (2446T>C). Differences were observed between Americans of European descent and Americans of African descent in the allelic frequencies of the ERCC1 N118N polymorphism (P < 0.000001). Differences were also observed between these two ethnic groups for ERCC2 K751Q (P = < 0.006675), XRCC1 R399Q (P < 0.000001), and PARP1 V762A (P = 0.000001). The ERCC1 N118N polymorphic variant that is seen most commonly in Americans of European descent is associated with a measurable reduction in NER function. ERCC1-mediated reduction in NER functionality affects the repair of cisplatin-DNA lesions.

Colorectal cancer is the second leading cause of cancer-related mortality among U.S. adults. In 2004, treatment costs for colorectal cancer were $8.4 billion. There is substantial evidence that colorectal cancer incidence and mortality are reduced with regular screening. The natural history of this disease is also well described: most colorectal cancers develop slowly from preexisting polyps. This slow development provides an opportunity to intervene with screening tests, which can either prevent colorectal cancer through the removal of polyps or detect it at an early stage. However, much less is known about how best to implement an effective colorectal cancer screening program. Screening rates are low, and uninsured persons, low-income persons, and persons who have not visited a physician within a year are least likely to be screened. Although the Centers for Disease Control and Prevention (CDC) has 15 years of experience supporting the National Breast and Cervical Cancer Early Detection Program for the underserved population, a similar national program for colorectal cancer is not in place. To explore the feasibility of implementing a national program for the underserved U.S. population and to learn which settings and which program models are most viable and cost-effective, CDC began a 3-year colorectal cancer screening demonstration program in 2005. This article describes briefly this demonstration program and the process CDC used to design it and to select program sites. The multiple-methods evaluation now under way to assess the program's feasibility and describe key outcomes is also detailed. Evaluation results will be used to inform future activities related to organized screening for colorectal cancer.

Beta-elemene has recently raised interest in P.R. China as a novel antitumor plant drug isolated from the Chinese medicinal herb Zedoary. To explore potentially useful combinations of beta-elemene with taxanes in the clinic, we characterized the effects of beta-elemene combined with taxanes in human lung cancer cells using a median effect analysis, micronucleus assay, apoptotic detection, and determination of gene expression in the signaling pathways of apoptosis. The synergistic analysis indicated that the interactions of beta-elemene with paclitaxel or docetaxel ranged from slight synergism to synergism. Combinations of beta-elemene with docetaxel induced much stronger synergistic interactions in p53 mutant H23 cells and p53 null H358 cells than in p53 wild-type H460 and A549 cells. Similar synergistic interactions were observed by micronucleus assay, apoptotic detection, and determination of apoptotic gene expression. Our findings indicate that the synergistic effects achieved with combinations of beta-elemene and taxanes are related to the augmented cytotoxic efficacy of taxanes owing to the action of beta-elemene. In H460 and A549 cells, dose-dependent upregulation of p53 protein expression was observed in cultures treated with docetaxel alone and with docetaxel plus beta-elemene, whereas no significant change in p53 expression was observed in any of the treatment groups in H23 cells. Fas revealed no alteration of expression with any of the treatments in this study. However, the combination treatments induced increased cytochrome c release from mitochondria, significant caspase-8 and -3 cleavage, and downregulation of Bcl-2 and Bcl-XL expression. These results suggest that, although p53 plays an important role in taxane-induced cell death, apoptosis induced by beta-elemene or in combination with docetaxel thereof seems to be initiated through a p53- and Fas-independent pathway via mitochondria in our lung cancer cells. The suppression of specific 'survival' gene expression appears to be the key action leading to the synergistic effect of combination treatments with beta-elemene and taxanes. Finally, the beta-elemene-induced alteration of cell membrane permeability, which has potential to result in enhanced cellular uptake of taxanes, may also contribute to the synergistic interactions of the combination treatments.

ERCC1 is a critical gene within the nucleotide excision repair pathway. Overexpression of ERCC1 through promoter-mediating transcriptional regulation is associated with repair of cisplatin-induced DNA damage and clinical resistance to platinum-chemotherapy. Several transcriptional repressors and activators within the 5'-flanking region of the ERCC1 gene may be involved in the up-regulation of this gene. Minimal sequence within the promoter region required for ERCC1 transcription was analyzed by CAT assay and demonstrated that the region of -220 to -110 is essential to constitutive expression of ERCC1 gene in ovarian cancer cell line A2780/CP70. A more forward upstream region seems to be responsible for cisplatin-induced expression. Study of the functional cis-element in this region by electrophoretic mobility shift assay indicates that a MZF1-like site as well as an AP1-like site responded in a time-dependent manner to cisplatin stimulation with altered binding activities. EMSA with MZF1 ZN1-4 consensus oligonucleotides suggests that the MZF1 N-terminal domain of zinc finger cluster may bind to the MZF1-like site of the ERCC1 promoter region. MZF1 mRNA in A2780/CP70 cells decreased upon cisplatin exposure as analyzed by quantitative PCR, suggesting that MZF1 may mediate cisplatin-invoked gene expression in these cells. Overexpression of MZF1 repressed the ERCC1 promoter activity as determined in co-transfection assay, suggesting that MZF1 might be a repressor of ERCC1 transcription upon cisplatin exposure. In summary, our studies revealed a core promoter region and adjacent drug-responsible region within the ERCC1 promoter. The drug-responsible region contains cis-elements of activator, AP1 and repressor, MZF1. In response to cisplatin treatment, decreased MZF1 and increased AP1 binding activities appear to be the leading mechanism of up-regulation of ERCC1 expression. Our findings imply potential therapeutic strategies to antagonize drug resistant mechanisms in treatment of human ovarian cancer.

The emergence of multidrug-resistant bacteria has become a global crisis. Accumulating evidence shows that bacteriophages (phages) can rescue animals from a variety of lethal infections and be effective in treating drug-resistant infections in humans. Enterobacteriaceae, producing extended spectrum beta-lactamase enzymes (ESBLs), are resistant to a broad range of beta-lactamase antibiotics. One of the most common ESBL-producing gram-negative bacilli in Enterobacteriaceae is Escherichia coli. Since ESBL-producing E. coli poses a formidable challenge in the management of critically ill patients with bacterial infections, we undertook this study to explore the possible therapeutic utility of phages to control ESBL-producing E. coli infections. The phage Ø9882 used in this study was isolated from our hospital sewage and has lytic activity against a broad range of clinical isolates of ESBL-producing E. coli. ESBL-producing E. coli strains (n=30) were isolated in the clinic, and one of them was used to induce bacteremia in a murine model. Bacteremia was established by intraperitoneal (i.p.) injection of 3 x 10(7) CFU/ml, the minimum lethal dose (MLD) of bacterium in this animal model. Mice infected with the MLD of this strain alone died within 14 h, whereas a single i.p. inoculation of Ø9882 (MOI > or =10(-4)) given 40 min after the bacterial challenge led to 100% survival at 24-168 h, compared to 0% survival of saline-treated controls. Protection was obtained even when administration of the phage was delayed up to 60 min after the bacterial infection and the survival rate of infected animals was 60% at 168 h. Furthermore, it was shown that the therapeutic efficacy of Ø9882 in lethal systemic infection in our model is due to the functional capability of the phage and not the nonspecific immune effects. Our data both in vitro and in vivo revealed that: i) the protection of mice from death occurred only in animals infected with selected bacterial strains and the virulent phage specific to them; ii) when the phages were heat-inactivated, survival of the infected mice was strikingly decreased to 0; and iii) the level of antibody against the phage was not substantially elevated when the bacteremic animals were protected by the phage. The present findings indicate that phages can effectively rescue our mouse model from bacteremia and death, and thus provide the rationale and framework to evaluate the therapeutical efficacy of lytic phages against fatal ESBL-producing E. coli infections in humans.

The use of 'double-thymidine block' was the first widely accepted method for inducing cell synchrony and remains one of the most effective and frequently used techniques for analyzing the cell cycle today. While thymidine is in itself an inhibitor of DNA replication, thymidine blocks are typically used to generate cell synchrony at the G1/S boundary. We have previously presented the first evidence that shows the growth imbalance and altered expression levels of cyclins A, B1, D3 and E in MOLT-4 cells synchronized in the cell cycle by thymidine. The major objective of the present study was to compare the levels of cyclins A, B1, D3 and E in G1 phase of the cell cycle between synchronized and unperturbed asynchronously growing human lymphocyte leukemia MOLT-4 cells. Here, we demonstrate that the sorted, asynchronously growing MOLT-4 cells had considerably lower levels of cyclins A, B1, D3 and E than their counterparts of the cells arrested in G1/S phase, as assessed by flow cytometry. In addition, we confirmed these results by using post-sorting Western blotting, a new method we recently developed for examining protein expression in specific phases of sorted, synchronized or asynchronously growing cells. Our findings revealed that the levels of cyclins D3 and E in the asynchronously growing MOLT-4 cells were significantly lower than those in synchronized cultures. Interestingly, protein expression levels of cyclins A and B1 in the asynchronously growing MOLT-4 cells were barely measurable, suggesting that these proteins were either not expressed or under detectable levels. These studies indicate that our synchronization protocol may have disturbed cell proliferation and metabolism as evidenced by significant differences in the expression of cyclins between asynchronously growing and synchronized cells, and further suggest that the levels of cyclins A, B1, D3 and E in synchronized cultures cannot represent those in unperturbed, asynchronously growing cells. Thus, it appears that thymidine-treated, synchronized cells may not be suitable experimental models for the study of normal cell cycle.

Caspase-3 is a critical effector caspase for apoptosis, which cleaves proteins, including cytoskeletal and associated proteins, kinases, and members of the Bcl-2 family of apoptosis-related proteins. This leads to changes in apoptotic morphology, such as membrane externalization and cytoplasm and nuclear condensation. It has been reported that pro-caspase-3 is activated in the cytosol. However, it remains obscure how caspase-3 activation correlates to serial changes in cell morphology during apoptosis. The current study was therefore undertaken to assess the relationship between caspase-3 activation and its subcellular localization and alterations in apoptotic morphology in MOLT-4 human leukemia cells exposed to X-ray irradiation. Fluorescence labeled inhibitor of caspases (FLICA) was used to detect caspase-3 activity in apoptotic cells in this project; cell morphology and caspase-3 sub-localization were determined by confocal microscopy. Our data showed that MOLT-4 cells presented typical morphological changes in apoptosis, such as membrane reversion, DNA fragmentation, and formation of apoptotic cell bodies following 10 Gray (Gy) of X-ray irradiation. Caspase-3 was activated 2 h after X-ray irradiation, and its activity increased markedly after 4-6-h exposure. Membrane reversion in MOLT-4 leukemia cells was detected by Annexin V assay at 4 h following X-ray irradiation, 2 h after the elevated caspase-3 activity was measured. Cytologically, activation of caspase-3 was first observed close to the inside surface of the cellular membrane, then transferred to the cytoplasm, and finally translocated to the nuclear region. We conclude that caspase-3 is activated in MOLT-4 cells following exposure to X-rays, and that the enhanced caspase-3 activity and its sub-localization shifting is correlated to changes in apoptotic morphology. The spatial shift of activated caspase-3 in X-ray-induced apoptotic MOLT-4 leukemia cells is a process of crucial importance for apoptosis.

Our previous studies revealed a splicing variant (lacking a 42 base pair segment) within the 5'-UTR of the ERCC1 gene, a critical component of the nucleotide excision repair (NER) pathway that plays an important role in the development of chemoresistance in platinum-based anticancer therapy. This 42-bp segment seems to possess a regulatory function in ERCC1 expression and representing the level of clinical response to platinum-treatment in ovarian cancer patients. To confirm the existence of the 42-bp deletion and to investigate the 42-bp function, we performed several experiments and assays. Northern blot analysis and RNase protection assay provide evidence that the 42-bp deletion occurs at RNA level of ERCC1 5'-UTR in both ovarian cancer cell lines and ovarian cancer tissues. Luciferase assay suggests that this gene fragment possesses a regulatory function as an enhancer of ERCC1 gene expression in ovarian cancer cells. In Electrophoretic Mobility Shift Assay (EMSA), a shift band present in the ovarian cancer cell line extracts is consistent with the presence of an intracellular protein that recognizes this specific 42-bp sequence. Further, specific EMSA results with 42-bp probe mutated at the site of RFX-1 indicate different putative-DNA binding proteins, rather than RFX-1. We conclude that the 42-bp sequence within the 5'-UTR influences the expression of ERCC1 and hence can influence response to cisplatin in ovarian cancer therapy.

Expression of cyclin E is believed to be a critical factor promoting cell entry into the S-phase and cell proliferation. Indeed, normal proliferating cells and most tumor cell lines are characterized by the existence of a minimal cyclin E threshold level in the G1-phase, and only those cells expressing cyclin E over this threshold enter into the S-phase of the cell cycle. However, through studying clinical tumor tissue specimens, we recently observed that some cancer cells can enter into the S-phase with minimal levels of cyclin E expression. In an effort to establish an in vitro cell model system for studying the mechanisms underlying this phenomenon, we treated MOLT-4 lymphocyte leukemia cells with 50 mM caffeine and found that the levels of cyclin E expression were decreased markedly in these cells following 2 to 4-h exposure to caffeine. Quite unexpectedly, we observed that the percentage of the cells progressing through the S-phase increased despite the reduced levels of cyclin E, as analyzed for the cellular DNA contents, expression of nuclear-bound PCNA, immunolabelling with Ki-67 antibody and incorporation of BrdU. In fact, these cells entered into the S-phase with a level of cyclin E well below the threshold level for untreated cells, thus suggesting that lower levels of cyclin E expression are associated with cell proliferation under certain circumstances. We speculate that caffeine may enhance MOLT-4 cell entrance into the S-phase through activation of Cdc25, which in turn activates cyclin-dependent protein kinases (CDKs) including CDK2 and drives the cell cycle progression; while degradation of cyclin E by the ubiquitin/proteasome pathway may account for the decreased levels of cyclin E in these cells. Our findings from both the MOLT-4 cell line and patients' cancer tissues may help decipher the mystery of the deregulation of cell cycle progression and carcinogenesis in some malignant tumors.

Ovarian cancer is the leading cause of death from gynecological malignancy and has the worst prognosis of all gynecological cancers. Vascular endothelial growth factor (VEGF) plays an important role in ovarian cancer development. 9-beta-D-Arabinofuranosyl-2-fluoroadenine (Fara-A), a nucleotide analog, is frequently used in treating certain types of cancer. However, the effectiveness of Fara-A on ovarian cancer cells is unknown. In this study, we found that Fara-A inhibited VEGF expression in human ovarian cancer cells. Fara-A inhibited VEGF transcriptional activation through hypoxia-inducible factor 1 (HIF-1). HIF-1 is composed of HIF-1alpha and -1beta subunits. Fara-A inhibited expression of HIF-1alpha but not HIF-1beta. Overexpression of HIF-1alpha reversed Fara-A-inhibited VEGF transcriptional activation. Our results demonstrated that Fara-A inhibited VEGF transcriptional activation through HIF-1alpha expression. Fara-A partly inhibited HIF-1alpha mRNA levels. Fara-A blocked the activation of AKT but not of ERK1/2. Overexpression of AKT reversed the Fara-A-inhibited VEGF transcriptional activation, suggesting that Fara-A inhibits VEGF expression via phosphatidylinositol 3-kinase/AKT signaling. These results demonstrate a new function of Fara-A in inhibiting VEGF and HIF-1alpha expression and identify a potential molecular mechanism of the regulation.

Background: CHK2 kinase is a tumor suppressor that plays important role in DNA damage signaling, cell cycle regulation and DNA damage induced apoptosis. CHK2 kinase expression was known to be ubiquitous in mammalian cells. CHK2-/- cells were remarkably resistant to DNA damage induced apoptosis, mimicking the clinical behavior of non-small cell lung cancer to conventional chemo and radiation therapy., Result: We reported that the CHK2 expression is diminished or absent in both non-small cell lung cancer (NSCLC) cell lines and clinical lung cancer tumor specimens. The absent CHK2 expression in NSCLC was due to hypermethylation of the CHK2 gene promoter, preventing from binding of a transcriptional factor, leading to silence of the CHK2 gene transcription., Conclusion: Since the CHK2 null mice showed a remarkable radioresistance, which bear significant similarity to clinical behavior of NSCLC, down-regulation of CHK2 kinase expression by CHK2 gene silencing and methylation in non-small cell lung cancer suggest a critical role of CHK2 kinase in DNA damage induced apoptosis and a novel mechanism of the resistance of NSCLC to DNA damage based therapy.

Vascular endothelial growth factor (VEGF) has been shown to play an important role in tumor growth and progression. However, the clinical implications of VEGF expression in ovarian tumors are not fully understood. We therefore investigated the serum level of VEGF in patients with ovarian tumors and explored the potential use of VEGF as a tumor marker for diagnosis, treatment and prognosis of human ovarian cancer. The serum VEGF (sVEGF) levels in 120 patients with ovarian carcinoma, 25 patients with benign ovarian tumor and 90 healthy female blood donors were measured by an enzyme-linked immunosorbent assay in this study. We also determined the levels of sVEGF in patients with epithelial ovarian cancer before and after surgery. Our results showed that: (i) ovarian cancer patients had significantly higher levels of sVEGF compared to those of patients with benign ovarian tumor or those of healthy individuals. As a cut-off at 100 pg/ml, the sensitivity and specificity of sVEGF levels for diagnosing ovarian carcinoma were 77.1% and 87%, respectively. (ii) sVEGF levels were markedly elevated in patients with advanced stage or poorly-differentiated ovarian cancer, as well as in those with more ascites (>500 ml), as compared to patients with early stage and well-differentiated ovarian cancer, or those with less ascites (<500 ml). However, there was no significant difference in sVEGF levels among different pathological subtypes of ovarian carcinoma. (iii) The post-operative sVEGF levels were significantly lower than the pre-operative sVEGF levels. (iv) We measured significantly higher levels of sVEGF in patients with metastasis as compared to patients lacking metastasis. Lastly (v) the average survival-time in patients with higher levels of sVEGF (>100 pg/ml) was 28 months, while the average survival-time in patients with lower levels of sVEGF (<100 pg/ml) was 35 months, indicating that the elevations in sVEGF level are correlated with patient survival and tumor metastasis in ovarian carcinoma. These data suggest that VEGF may be a useful serological biomarker for clinical diagnosis and prognosis of ovarian cancer, for follow-up of ovarian tumor metastasis and for monitoring the efficacy of therapy in patients with ovarian carcinomas.

Purpose: Carboxyamidotriazole (CAI) is a cytostatic inhibitor of nonvoltage-operated calcium channels and calcium channel-mediated signaling pathways. It inhibits angiogenesis, tumor growth, invasion, and metastasis. We hypothesized that CAI would promote disease stabilization lasting >/= 6 months in patients with relapsed ovarian cancer., Patients and Methods: Patients with epithelial ovarian cancer, good end-organ function, measurable disease, and three or fewer prior regimens were eligible. Oral CAI was given daily using a pharmacokinetic-dosing approach to maintain plasma concentrations between 2 and 4 microg/mL. Radiographic imaging to assess response was performed every 8 weeks. Positive outcome included stabilization or improvement of disease lasting >/= 6 months. Plasma vascular endothelial growth factor (VEGF), interleukin (IL)-8, and matrix metalloproteinase (MMP)-2 were measured., Results: Thirty-six patients were assessable for primary end point analysis, and 38 were assessable for toxicity. Forty-four percent of patients had three prior regimens, more than 50% had four or more disease sites, and 48% had liver metastases. Thirty-three patients reached the targeted concentration range during the first cycle. Eleven patients (31%) attained the >/= 6-month outcome end point, with one partial response (8 months) and three minor responses (8, 12+, and 13 months). Median time to progression was 3.6 months (range, 1.6 to 13.3 months). CAI was well tolerated, with mostly grade 1 to 2 toxicity. Grade 3 events included fatigue (5%), vomiting (2%), neutropenic fever (2%), and neutropenia (2%). There were no grade 4 adverse events. No associations between VEGF, IL-8, and MMP-2 with CAI concentration or clinical outcome were observed., Conclusion: CAI is a potential agent for additional study in the stabilization of relapsed ovarian cancer. Given a limited toxicity profile, it may have utility as a maintenance therapeutic agent for this disease.

Purpose: The purpose of the present work was to investigate the relationship between mRNA expression of ERCC1 and XPB, two key genes in the nucleotide excision repair pathway, and clinical resistance of platinum-chemotherapy in histological subtypes of epithelial ovarian cancer., Experimental Design: mRNA levels of ERCC1 and XPB in epithelial ovarian cancer specimens from 126 different individuals were assessed using reverse transcription-PCR and followed by Southern hybridization methodology. Data were analyzed by linear regression analyses and by exhaustive regression analyses., Results: Five different histological types of tumors were examined; serous (n = 76), mucinous (n = 11), clear cell (n = 9), poorly differentiated (n = 9), and endometroid (n = 21). Numerical values for mRNA expression levels were based on internal controls for a stable comparative cell line and for beta-actin. Median values for ERCC1 and XPB mRNAs within clear cell tumors were, on average, >2-fold higher than the other histological tumor types. Linear regression analyses suggest a continuum of nucleotide excision repair gene expression among these cell types, and exhaustive regression analyses demonstrate that the higher mRNA levels seen in clear cell tumors are highly statistically significant., Conclusions: We conclude that mRNA levels of ERCC1 and XPB tend to be higher in clear cell tumors as opposed to other types of epithelial ovarian cancer. This is consistent with the long-standing observation that clear cell tumors are more likely to show de novo drug resistance against DNA damaging agents in the clinic.

Inherited breast and ovarian cancers account for 10% of all breast and ovarian cancers. Relative to sporadic breast and ovarian cancers, these cancers tend to occur at an earlier age and grow more aggressively. Women with BRCA1 and BRCA2 mutations (BRCA1/2 mutation) have a 65% to 85% cumulative lifetime risk of developing invasive breast cancer and a 15% to 65% cumulative lifetime risk of developing invasive ovarian cancer. Identification of patients with the mutation is therefore crucial, because preventive measures such as prophylactic bilateral mastectomy, prophylactic bilateral salpingpo-oophorectomy and chemoprevention with Tamoxifen can prevent breast and ovarian cancer. Likewise, genetic counseling prior to testing is important, considering the major impact of the test results on an individual's life.

Anti-angiogenic therapy represents one of the most promising treatment modalities for human cancer. Thalidomide (alpha-N-phthalimidoglutarimide) is a potent inhibitor of angiogenesis, and it is reported to overcome classical drug resistance in human multiple myeloma cells. However, the effect of this agent on the expression of angiogenic growth factors in cisplatin-resistant tumors is largely unknown. In the current study, we showed that thalidomide suppressed the expression of vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) in cisplatin-resistant human A549DDP lung carcinoma cells. The mRNA levels of VEGF and bFGF were markedly decreased in the A549DDP cells treated with the therapeutic concentrations of thalidomide (0.6-6 micrograms/ml), as determined by RT-PCR analysis. Consistent with these results, thalidomide also significantly reduced the protein levels of VEGF and bFGF in these cells in a dose- and time-dependent manner. This study provided evidence to support the potential therapeutic applications of thalidomide in cisplatin-resistant human lung cancer and other tumors.

Thalidomide requires cytochrome P450 (CYP)-catalyzed biotransformation for its antiangiogenic property, and CYP2C19 is responsible for 5-hydroxylation and 5'-hydroxylation of thalidomide in human. This study explored a hypothesis that patients with poor metabolizing phenotype of CYP2C19 receive little benefit from thalidomide treatment and that the poor metabolizer genotype is associated with lower ability to form the metabolites. A case-control study was conducted with 63 patients with prostate cancer who had been enrolled in a randomized phase II trial of thalidomide monotherapy (200 to 1,200 mg/day). CYP2C19 polymorphism (CYP2C19(*)2, CYP2C19(*)3, CYP2C19(*)4) was compared with clinical events (prostate-specific antigen (PSA) decline) and formations of the hydroxylated metabolites. Two patients were homozygous for the variant CYP2C19(*)2 allele (poor metabolizing phenotype). Both of these were included in the 25 patients whose PSA failed to demonstrate a decline. While 32% and 48% of the patients had quantifiable levels of 5-hydroxythalidomide and cis-5'-hydroxythalidomide, respectively, these metabolite were below quantification in both poor metabolizing patients. None had CYP2C19(*)3 or CYP2C19(*)4 alleles. Although this study had no power to detect the statistical significance of the CYP2C19 genotype, the findings were consistent with our hypothesis. The role of CYP2C19 polymorphism in thalidomide treatments remains to be elucidated.