Pancreatic ductal adenocarcinoma is the eighth most common cancer with the lowest overall 5-year relative survival rate of any tumor type today. Expression profiling using microarrays has been widely used to identify genes associated with pancreatic cancer development. To extract maximum value from the available gene expression data, we applied a meta-analysis to search for commonly differentially expressed genes in pancreatic ductal adenocarcinoma. We obtained data sets from four different gene expression studies on pancreatic cancer. We selected a consensus set of 2984 genes measured in all four studies and applied a meta-analysis approach to evaluate the combined data. Of the genes identified as differentially expressed, several were validated using RT-PCR and immunohistochemistry. Additionally, we used a class discovery algorithm to identify a gene expression signature. Our meta-analysis revealed that the pancreatic cancer gene expression data sets shared a significant number of up- and downregulated genes, independent of the technology used. This interstudy crossvalidation approach generated a set of 568 genes that were consistently and significantly dysregulated in pancreatic cancer. Of these, 364 (64.1%) were upregulated and 204 (35.9%) were downregulated in pancreatic cancer. Only 127 (22%) were described in the published individual analyses. Functional annotation of the genes revealed that genes presumably associated with the cell adhesion-mediated drug resistance pathway are frequently overexpressed in pancreatic cancer. Meta-analysis is an important tool for the identification and validation of differentially expressed genes. These could represent good candidates for novel diagnostic and therapeutic approaches to pancreatic cancer.

Beckwith-Wiedemann syndrome (BWS) is an overgrowth syndrome associated with macroglossia, abdominal wall defects, ear anomalies, and an increased risk for embryonic tumors. Reported tumor risk estimates vary between 4% and 21%. It has been hypothesized that tumor predisposition in BWS is related to the imprinting status of the H19 and LIT1 genes on chromosome 11p15. A loss of imprinting (LOI) of H19 implies a higher tumor risk. However, a systematic analysis of available data is lacking. Therefore, we performed a review and meta-analysis of reported associations between the imprinting status of the LIT1 and H19 genes and the risk for tumor development in BWS. Five publications suitable for meta-analysis were identified by electronic database searches. Sufficient data were available for 402 out of 520 patients. Patients were divided into four groups based on the imprinting status of H19 and LIT1: group I with LOI of LIT1 (45%); group II with LOI of H19 (9%); group III with LOI of LIT1 and LOI of H19 (21%); and group IV with normal imprinting patterns (26%). Differences in tumor risk between groups were studied with random effects meta-analysis. Tumors occurred in 55 patients. The odds of tumor development was significantly lower in group I when compared to group II (OR=0.06; 95% CI: 0.02-0.21) and group III (OR=0.12; 95% CI: 0.04-0.37). Tumor risk did not differ significantly between groups II and III (OR=1.40; 95% CI: 0.56-3.50). Compared to group IV, tumor risk was significantly lower in group I (OR=0.33; 95% CI: 0.12-0.87) and higher in groups II (OR=4.0; 95% CI: 1.5-10.4) and III (OR=2.6; 95% CI: 1.2-5.7). Tumor incidence rate for group IV was 10.6% (95% CI: 3.6-17.7). Calculated absolute risks were 3% for group I, 43% for group II, and 28% for group III, respectively. No Wilms tumor was seen in group I. In total, other tumors were seen with comparable frequencies in groups I-III. The results show a strong association between a LOI of H19 and especially Wilms tumor development in BWS.

Glutathione s-transferase (GST) polymorphisms (GSTM1, GSTP1 and GSTT1) have been considered as risk factors for developing acute leukaemia in a number of studies; however the overall results of such studies are inconsistent. To investigate a putative association of GST polymorphisms with the risk of acute leukaemia, we performed a systematic review and meta-analysis of 30 published case-control studies. To take into account the possibility of heterogeneity across the studies, a statistical test was performed. The pooled odds ratios (ORs) were assessed using both a fixed-effects and a random-effects model. The pooled OR of acute leukaemia risks associated with GSTM1 null genotype, GSTP1 Val105 allele and GSTT1 null genotype were 1.22 (95% confidence interval (CI) 1.07-1.38), 1.07 (95% CI 1.00-1.13) and 1.19 (95% CI 1.00-1.41), respectively. Significantly increased risk of acute lymphoblastic leukaemia associated with GSTM1 and GSTT1 null genotypes was observed. Their pooled ORs were 1.24 (95% CI 1.17-1.31) and 1.30 (95% CI 1.06-1.60), respectively. We also found substantial evidence of heterogeneity between the studies. Our results suggest that GSTM1 and GSTT1, but not GSTP1 polymorphisms, appear to be associated with a modest increase in the risk of acute lymphoblastic leukaemia. It is conceivable that GSTM1 and GSTT1 null genotypes may thus play a role in leukemogenesis. A review of the 30 case-control studies indicates that greater attention should be paid to the design of future studies.

Phase I/II enzymes metabolize environmental carcin-ogens and several functional polymorphisms have been reported in their encoding genes. Although their significance with regard to esophageal carcinogenicity has been examined epidemiologically, it remains controversial. The present systematic review of the literature was performed to clarify associations.

Microsatellite instability (MSI) testing is a common screening procedure used to identify families that may harbor mutations of a mismatch repair (MMR) gene and therefore may be at high risk for hereditary colorectal cancer. A reliable estimate of sensitivity and specificity of MSI for detecting germline mutations of MMR genes is critical in genetic counseling and colorectal cancer prevention. Several studies published results of both MSI and mutation analysis on the same subjects. In this article we perform a meta-analysis of these studies and obtain estimates that can be directly used in counseling and screening. In particular, we estimate the sensitivity of MSI for detecting mutations of MSH2 and MLH1 to be 0.81 (0.73-0.89). Statistically, challenges arise from the following: (a) traditional mutation analysis methods used in these studies cannot be considered a gold standard for the identification of mutations; (b) studies are heterogeneous in both the design and the populations considered; and (c) studies may include different patterns of missing data resulting from partial testing of the populations sampled. We address these challenges in the context of a Bayesian meta-analytic implementation of the Hui-Walter design, tailored to account for various forms of incomplete data. Posterior inference is handled via a Gibbs sampler.

STK15 (Aurora-A) is a serine/threonine kinase involved in mitotic chromosomal segregation. A genetic variant in STK15 T+91A (resulting in the amino acid substitution F31I) is associated with increased aneuploidy in colon tumors and cell transformation in vitro. Since this polymorphism plays a role in mitotic control-a process critical for all cancer types-we conducted association analyses for risk of cancer development of the colon, breast, prostate, skin, lung and esophagus in 10 independent case-control populations. We carried out a meta-analysis of these 10 case-control studies together with 5 additional published studies for a total of 9549 cases of breast, colon, ovarian, prostate, lung, esophageal and non-melanoma skin cancer and 8326 population or hospital-based controls. Meta-analysis of three colorectal cancer studies showed an increased risk in T+91A homozygotes (OR=1.50; 95% CI of 1.14-1.99). Meta-analysis of four breast cancer studies showed increased risk for T+91A homozygotes (OR=1.35, 95% CI of 1.12-1.64). The results of the multiple cancer type meta-analysis for all 15 studies combined were significant for cancer risk in both homozygotes and heterozygotes. The T+91A heterozygotes show an OR of 1.10 (95% CI of 1.03-1.18, P-value=0.006) and the T+91A homozygotes show an OR of 1.40 (95% CI of 1.22-1.59, P-value<0.001) for cancer risk. These results confirm that the STK15 T+91A variant is a low penetrance cancer susceptibility allele affecting multiple cancer types, and provide genetic evidence from large-scale human population studies that genetic stability at the chromosome level is an important determinant of cancer susceptibility. The data also underline the advantages of comparative association studies involving study populations from different ethnic groups for determination of disease risk.

All available comparative genomic hybridisation (CGH) analyses (n=31, until 12/2003) of human hepatocellular carcinomas (HCCs; n=785) and premalignant dysplastic nodules (DNs; n=30) were compiled and correlated with clinical and histological parameters. The most prominent amplifications of genomic material were present in 1q (57.1%), 8q (46.6%), 6p (22.3%), and 17q (22.2%), while losses were most prevalent in 8p (38%), 16q (35.9%), 4q (34.3%), 17p (32.1%), and 13q (26.2%). Deletions of 4q, 16q, 13q, and 8p positively correlated with hepatitis B virus aetiology, while losses of 8p were more frequently found in hepatitis C virus-negative cases. In poorly differentiated HCCs, 13q and 4q were significantly under-represented. Moreover, gains of 1q were positively correlated with the occurrence of all other high-frequency alterations in HCCs. In DNs, amplifications were most frequently present in 1q and 8q, while deletions occurred in 8p, 17p, 5p, 13q, 14q, and 16q. In conclusion, aetiology and dedifferentiation correlate with specific genomic alterations in human HCCs. Gains of 1q appear to be rather early events that may predispose to further chromosomal abnormalities. Thus, explorative CGH meta-analysis generates novel and testable hypotheses regarding the cause and functional significance of genomic alterations in human HCCs.

Several published epidemiologic studies show increased breast cancer risk for carriers of the Val-allele at codon 655 of the HER2 gene. We conducted additional analyses using data from three studies, including case-control analyses stratified on age and kin-cohort analyses using relatives of cases and controls. The results provide additional evidence that HER2 codon 655 genotype may predispose to early-onset breast cancer.

Variants in the gene encoding the macrophage scavenger receptor 1 (MSR1(4)) protein have been identified in men with prostate cancer, and several small studies have suggested that the 999C>T (R293X) protein-truncating mutation may be associated with an increased risk for this disease.

Cytochrome P450 1B1 (CYP1B1) and catechol-O-methyltransferase (COMT) are important estrogen-metabolizing enzymes and, thus, genetic polymorphisms of these enzymes may affect breast cancer risk. A population-based case-control study was conducted to assess the association of breast cancer risk with CYP1B1 and COMT polymorphisms. A meta-analysis was done to summarize the findings from this and previous studies. Included in this study were 1,135 incident breast cancer cases diagnosed from August 1996 through March 1998 among female residents of Shanghai and 1,235 randomly selected, age frequency-matched controls from the same general population. The common alleles of the CYP1B1 gene were Arg (79.97%) in codon 48, Ala (80.53%) in codon 119, and Leu (86.57%) in codon 432. The Val allele accounted for 72.46% of the total alleles identified in codon 108/158 of the COMT gene. No overall associations of breast cancer risk were found with any of the single nucleotide polymorphisms described above. This finding was supported by a meta-analysis of all previous published studies. No gene-gene interactions were observed between CYP1B1 and COMT genotypes. The associations of breast cancer risk with factors related to endogenous estrogen exposure, such as years of menstruation and body mass index, were not significantly modified by the CYP1B1 and COMT genotypes. We observed, however, that women who carried one copy of the variant allele in CYP1B1 codons 48 or 119 were less likely to have estrogen receptor-positive breast cancer than those who carried two copies of the corresponding wild-type alleles. The results from this study were consistent with those from most previous studies, indicating no major associations of breast cancer risk with CYP1B1 and COMT polymorphisms.

Longevity may depend on a balance between tumor suppression and tissue renewal mechanisms [Campisi, J., 2003. Cancer and ageing: rival demons? Nat. Rev. Cancer 3 (5), 339-349]. Mice with constitutively activated p53 are almost cancer free but their life span is reduced and accompanied by early tissue atrophy [Tyner et al., 2002. p53 mutant mice that display early ageing-associated phenotypes. Nature 415 (6867) 45-53]. Replacement of arginine (Arg) by proline (Pro) at position 72 of human p53 decreases its apoptotic potential [Dumont et al., 2003. The codon 72 polymorphic variants of p53 have markedly different apoptotic potential. Nat. Genet. 33 (3), 357-365] providing a tool to test for a similar trade-off in humans. Using a formal meta-analysis of the published literature we show that carriers of the TP53 codon 72 Pro/Pro genotype have an increased cancer risk compared to Arg/Arg carriers (p<0.05). Next, in a prospective study of 1226 people aged 85 years and over we show that carriers of the Pro/Pro genotype have a 41% increased survival (p = 0.032) despite a 2.54 fold increased (p = 0.007) proportional mortality from cancer. It is suggested that human p53 protect against cancer but at a cost of longevity.

Molecular analyses have demonstrated mutations in the K-ras gene at codon 12 in the majority of pancreatic ductal adenocarcinomas (PDACs). In order to determine whether the K-ras mutation rate increases parallel to the grade of dysplasia in duct lesions, we performed a meta-analysis of the studies published between 1988 and 2003 that provide information on K-ras mutations in hyperplastic and dysplastic duct lesions in the pancreas. The described duct lesions were reclassified according to the nomenclature for pancreatic intraepithelial neoplasia (PanIN), and the molecular methods for detecting K-ras were reviewed. In PanIN lesions from pancreata of patients with PDAC, there was a stepwise increase in K-ras mutations that correlated with the grade of dysplasia of the PanIN lesion. K-ras mutations were found in 36%, 44%, and 87% of PanIN-1a, 1b, and 2-3 lesions, respectively (trend statistic P <.001). Mutation-enriched polymerase chain reaction (PCR) resulted in higher rates of K-ras mutations in PanIN than plain PCR did. The incidence of K-ras mutations in PanIN lesions associated with chronic pancreatitis (CP) or normal pancreas was low (around 10%). In CP, K-ras mutations were only found after a disease duration of 3 years. The correlation of the incidence of K-ras mutations with the grade of dysplasia in PanIN and the occurrence of these mutations in CP with a duration of more than 3 years underlines the importance of this genetic change for the development of PDAC.

The glutathione S-transferase (GST) gene superfamily encodes for enzymes involved in conjugation of electrophilic compounds to glutathione. Several polymorphisms in the GST genes have been implicated as risk factors for prostate cancer. We did a meta-analysis of 11 studies with GSTM1 genotyping (2,063 prostate cancer cases and 2,625 controls), 10 studies with GSTT1 genotyping (1,965 cases and 2,554 controls), and 12 studies with GSTP1 genotyping (2,528 cases and 3,076 controls). The random effects odds ratio was 1.08 [95% confidence interval (95% CI), 0.93-1.25, no significant between-study heterogeneity] for the GSTM1 null versus nondeleted genotype and 0.90 (95% CI, 0.73-1.12; P = 0.03 for heterogeneity) for the GSTT1 null versus nondeleted genotype. Overall, the random effects odds ratio was 1.05 (95% CI, 0.90-1.21; P < 0.01 for heterogeneity) for the GSTP1-Val versus GSTP1-Ile allele. For all three polymorphisms, there was a trend for the presence of an association in the earliest published studies, but this did not seem to be validated in subsequent research. For GSTT1, larger studies gave different results than smaller ones. The meta-analysis shows that these three polymorphisms are unlikely to be major determinants of susceptibility to prostate cancer on a wide population basis.

To clarify the influence of genetic polymorphisms on colorectal cancer.

To study the association of genetic polymorphism of glutathione S-transferase M1 (GSTM1) and its susceptibility to esophageal cancer.

The proto-oncogene RAS, coding for a 21 kDa protein (p21), is mutated in 20% of lung cancer. However, the literature remains controversial on its prognostic significance for survival in lung cancer. We performed a systematic review of the literature with meta-analysis to assess its possible prognostic value on survival. Published studies on lung cancer assessing prognostic value of RAS mutation or p21 overexpression on survival were identified by an electronic search. After a methodological assessment, we estimated individual hazard ratios (HR) estimating RAS protein alteration or RAS mutation effect on survival and combined them using meta-analytic methods. In total, 53 studies were found eligible, with 10 concerning the same cohorts of patients. Among the 43 remaining studies, the revelation method was immunohistochemistry (IHC) in nine and polymerase chain reaction (PCR) in 34. Results in terms of survival were significantly pejorative, significantly favourable, not significant and not conclusive in 9, 1, 31, 2, respectively. In total, 29 studies were evaluable for meta-analysis but we aggregated only the 28 dealing with non-small-cell lung cancer (NSCLC) and not the only one dealing with small-cell-lung cancer (SCLC). The quality scores were not statistically significantly different between studies with or without significant results in terms of survival, allowing us to perform a quantitative aggregation. The combined HR was 1.35 (95% CI: 1.16-1.56), showing a worse survival for NSCLC with KRAS2 mutations or p21 overexpression and, particularly, in adenocarcinomas (ADC) (HR 1.59; 95% CI 1.26-2.02) and in studies using PCR (HR 1.40; 95% CI 1.18-1.65) but not in studies using IHC (HR 1.08; 95% CI 0.86-1.34). RAS appears to be a pejorative prognostic factor in terms of survival in NSCLC globally, in ADC and when it is studied by PCR.

Susceptibility to gastric cancer may be in part attributable to inter-individual variability in metabolic activation or detoxification of carcinogens, and in this context the polymorphic GSTM1 gene has been extensively studied. Seventeen reports detailing a possible association between GSTM1 deletion and gastric cancer have been published so far. In order to examine the risk of gastric cancer associated with GSTM1 null genotype, a meta-analysis of published case-control studies was undertaken using a random effect model. Two studies were excluded because some data were missing in the results. The principal outcome measure was the odds ratio for the risk of gastric cancer. Pooling all the 15 studies identified, the overall odds ratio of gastric cancer risk associated with GSTM1 deficiency was 1.24 (95% confidence interval, 1.00-1.54). Furthermore, by pooling four studies detailing the possible interaction between GSTM1 status smoking habits and gastric cancer risk, an overall estimate of odds ratio of 2.93 (95% confidence interval, 1.56-5.47) for ever smokers with GSTM1 deficiency compared to non-smokers with GSTM1 normal genotype has emerged. These results suggest that GSTM1 status probably has no effect on the risk of gastric cancer per se, but may modulate tobacco-related carcinogenesis of gastric cancer. Greater attention should therefore be paid to the design of future studies: only well designed population-based control studies considering all the possible confounding risk factors and based on a sample size commensurate with the detection of small genotypic risk may allow a more definitive conclusion.

To investigate the relationship between tobacco smoking, drinking and p53 alteration in esophageal carcinoma.

To evaluate the prognostic significance of p53 mutation and P53 protein expression abnormality among esophageal cancer.

Although narrative reviews have suggested an association between (CAG)n and (GGN)n polymorphisms in the AR gene and prostate cancer, it has never been quantified systematically. The purpose of this meta-analysis was to provide relative and absolute quantitative summary estimates with sufficient power.