Computational approaches in cancer multidrug resistance research: Identification of potential biomarkers, drug targets and drug-target interactions.

Like physics in the 19th century, biology and molecular biology in particular, has been fertilized and enhanced like few other scientific fields, by the incorporation of mathematical methods. In the last decades, a whole new scientific field, bioinformatics, has developed with an output of over 30,000 papers a year (Pubmed search using the keyword "bioinformatics"). Huge databases of mass throughput data have been established, with ArrayExpress alone containing more than 2.7 million assays (October 2019). Computational methods have become indispensable tools in molecular biology, particularly in one of the most challenging areas of cancer research, multidrug resistance (MDR). However, confronted with a plethora of different algorithms, approaches, and methods, the average researcher faces key questions: Which methods do exist? Which methods can be used to tackle the aims of a given study? Or, more generally, how do I use computational biology/bioinformatics to bolster my research? The current review is aimed at providing guidance to existing methods with relevance to MDR research. In particular, we provide an overview on: a) the identification of potential biomarkers using expression data; b) the prediction of treatment response by machine learning methods; c) the employment of network approaches to identify gene/protein regulatory networks and potential key players; d) the identification of drug-target interactions; e) the use of bipartite networks to identify multidrug targets; f) the identification of cellular subpopulations with the MDR phenotype; and, finally, g) the use of molecular modeling methods to guide and enhance drug discovery. This review shall serve as a guide through some of the basic concepts useful in MDR research. It shall give the reader some ideas about the possibilities in MDR research by using computational tools, and, finally, it shall provide a short overview of relevant literature.

The oncomiR miR-197 is a novel prognostic indicator for non-small cell lung cancer patients.

BACKGROUND: MicroRNA expression signatures can promote personalised care for non-small cell lung cancer (NSCLC) patients. Our aim was to evaluate the previously unexplored prognostic potential of miR-197, a key oncogenic molecule for NSCLC. METHODS: Total RNA isolation (n=124 NSCLC and n=21 tumour-adjacent normal tissues), was performed using the QIAsymphony SP workstation. The quantity and quality of RNA were assessed by spectrophotometric analysis and an Agilent 2100 bioanalyzer. Polyadenylation and reverse transcription were subsequently carried out. MiR-197 expression levels were measured by qPCR, after quality control (inter-assay CV=7.8%). Internal validation procedures were followed by assigning training and test sets and robust biostatistical analyses were performed, including bootstrap resampling. RESULTS: MiR-197 is associated with larger tumours (P=0.042) and the squamous cell carcinoma histotype (P=0.032). Interestingly, after adjusting for important prognostic indicators, miR-197 expression was identified as a novel independent predictor of unfavourable prognosis for NSCLC patients (HR=1.97, 95% CI=1.10-3.38, P=0.013). We also demonstrate that miR-197 retains its prognostic performance in both early-stage I (P=0.045) and more advanced-stage individuals (P=0.036). CONCLUSIONS: The cost-effective expression analysis of miR-197 could constitute a novel molecular tool for NSCLC management.

RAS/PI3K crosstalk and cetuximab resistance in head and neck squamous cell carcinoma.

PURPOSE: Cetuximab, an antibody directed against the EGF receptor, is an effective clinical therapy for patients with head and neck squamous cell cancer (HNSCC). Despite great clinical promise, intrinsic or acquired cetuximab resistance hinders successful treatment outcomes but little is known about the underlying mechanism. EXPERIMENTAL DESIGN: To study the role of oncogenic HRAS in cetuximab resistance in HNSCC, the frequency of oncogenic HRAS mutations was determined in a cohort of 180 genomic DNAs from head and neck cancer specimens. We also used a combination of cetuximab-resistant cell lines and a transgenic mouse model of RAS-driven oral cancer to identify an oncogenic RAS-specific gene expression signature that promotes cetuximab resistance. RESULTS: Here, we show that activation of RAS signaling leads to persistent extracellular signal-regulated kinase 1/2 signaling and consequently to cetuximab resistance. HRAS depletion in cells containing oncogenic HRAS or PIK3CA restored cetuximab sensitivity. In our study, the gene expression signature of c-MYC, BCL-2, BCL-XL, and cyclin D1 upon activation of MAPK signaling was not altered by cetuximab treatment, suggesting that this signature may have a pivotal role in cetuximab resistance of RAS-activated HNSCC. Finally, a subset of patients with head and neck cancer with oncogenic HRAS mutations was found to exhibit de novo resistance to cetuximab-based therapy. CONCLUSIONS: Collectively, these findings identify a distinct cetuximab resistance mechanism. Oncogenic HRAS in HNSCC promotes activation of ERK signaling, which in turn mediates cetuximab resistance through a specific gene expression signature. Clin Cancer Res; 20(11); 2933-46. ©2014 AACR.

Galectin-1 has potential prognostic significance and is implicated in clear cell renal cell carcinoma progression through the HIF/mTOR signaling axis.

BACKGROUND: Metastatic clear cell renal cell carcinoma (ccRCC) patients have <9% 5-year survival rate, do not respond well to targeted therapy and eventually develop resistance. A better understanding of molecular pathways of RCC metastasis is the basis for the discovery of novel prognostic markers and targeted therapies. METHODS: We investigated the biological impact of galectin-1 (Gal-1) in RCC cell lines by migration and invasion assays. Effect of Gal-1 expression on the mitogen-activated protein kinase pathway was assessed by proteome array. RESULTS: Increased expression of Gal-1 increased cell migration while knocking down Gal-1 expression by siRNA resulted in reduced cellular migration (P<0.001) and invasion (P<0.05). Gal-1 overexpression increased phosphorylation of Akt, mTOR and p70 kinase. Upon hypoxia and increased HIF-1α, Gal-1 increased in a dose-dependent manner. We also found miR-22 overexpression resulted in decreased Gal-1 and HIF-1α. Immunohistochemistry analysis showed that high Gal-1 protein expression was associated with larger size tumor (P=0.034), grades III/IV tumors (P<0.001) and shorter disease-free survival (P=0.0013). Using the Cancer Genome Atlas data set, we found that high Gal-1 mRNA expression was associated with shorter overall survival (41 vs 78 months; P<0.01). CONCLUSIONS: Our data suggest Gal-1 mediates migration and invasion through the HIF-1α-mTOR signaling axis and is a potential prognostic marker and therapeutic target.

The loss of the tumour-suppressor miR-145 results in the shorter disease-free survival of prostate cancer patients.

BACKGROUND: Prostate cancer (PCa) is characterised by great heterogeneity of the disease progression rate. Tumours range from insignificant and not life threatening to high risk for relapse ones. Consequently, a large number of patients undergo unnecessary treatment. miR-145 is a well-documented tumour suppressor and its expression, which is regulated by the p53 pathway, has been found to be decreased in the majority of human malignancies. The aim of our study was to evaluate the clinical utility of miR-145 for the prognostication of PCa. METHODS: Total RNA was isolated from 137 prostate tissue specimens obtained from 73 radical prostatectomy-treated PCa patients and 64 transurethral- or open prostatectomy-treated benign prostate hyperplasia (BPH) patients. Following polyadenylation and reverse transcription, miR-145 levels were determined by quantitative real-time PCR assay, using SNORD48 (RNU48) for normalisation purposes. RESULTS: Downregulated miR-145 expression was found in PCa compared with BPH patients. The reduction of miR-145 expression in PCa was correlated with higher Gleason score, advanced clinical stage, larger tumour diameter and higher prostate-specific antigen (PSA) and follow-up PSA levels. In addition, higher risk for biochemical recurrence and significantly shorter disease-free survival (DFS) was found for the PCa patients expressing lower miR-145. Focusing on 'low- and intermediate-recurrence risk' PCa patients, miR-145 loss was revealed to be a reliable predictor of biochemical relapse and poor DFS independent from Gleason score, clinical stage, PSA and patients' age. CONCLUSION: The loss of the tumour-suppressor miR-145 increases the risk for disease progression and predicts the poor survival of PCa patients.

Kallikrein-related peptidases (KLKs) in gastrointestinal cancer: mechanistic and clinical aspects.

The human tissue kallikrein (KLK1) and kallikrein-related peptidases (KLKs) are secreted serine proteases with diverse expression patterns and physiological roles in different systems, including the digestive system. The aberrant expression of KLKs in gastrointestinal malignancies as well as their implication in carcinogenesis including cell growth regulation, angiogenesis, invasion, and metastasis, has prompted scientists to investigate their potential as cancer biomarkers. Expression of distinct KLKs is associated with various clinic-pathological parameters of patients with gastric, colorectal, pancreatic, hepatic, and esophageal cancer. Moreover, several KLKs possess significant favourable or unfavourable prognostic value in these human malignancies. Identification of novel diagnostic, prognostic and predictive biomarkers will contribute utmost to clinical decision-making, since early diagnosis of gastrointestinal cancer and early detection of recurrence following surgery are critical for the effective treatment of patients and for a positive clinical outcome. The current review provides a brief overview of the functional role of KLKs in gastric, colorectal, pancreatic, hepatic, and esophageal cancer, and describes the current status of KLKs as potential tumour biomarkers in these human malignancies.

Immunohistochemical localization and analysis of kallikrein-related peptidase 7 and 11 expression in paired cancer and benign foci in prostate cancer patients.

Kallikrein-related peptidases 7 and 11 (KLK7/KLK11) share a high degree of structural similarity with PSA (KLK3) and other KLKs. The aim of this study was to evaluate differences in KLK7/ KLK11 expression in paired cancer/benign prostate foci and to determine possible associations with clinicopathological parameters. Seventy archived paraffin-embedded tissue samples obtained from radical prostatectomy were stained for KLK7, KLK11, PSA and PSMA and expression was evaluated semiquantitatively. The results showed statistically significant differences for all studied proteins between BPH and CaP foci. Both KLK7 (P=0.026) and KLK11 (P<0.001) expressions were decreased in prostate cancer cells compared to normal/benign prostate cells. Positive correlations were found for both KLK7 (Rs=0.74, P<0.001) and KLK11 (Rs=0.35, P=0.003) between CaP and BPH. We found a statistically significant upregulation of KLK11 in advanced cases compared to localized ones (P=0.026). For the first time, we report lower expression of KLK11 in CaP compared to BPH and slight upregulation of KLK11 in advanced tumors compared to localized ones. Our observations support the diagnostic potential of KLK7/KLK11 for early prostate cancers but further studies on larger cohorts are needed in order to validate the clinical value of these biomarkers and clarify their biological role in prostate development and tumorigenesis.

Impact of expression differences of kallikrein-related peptidases and of uPA and PAI-1 between primary tumor and omentum metastasis in advanced ovarian cancer.

BACKGROUND: Primary tumor levels of serine proteases of the kallikrein-related peptidases (KLK) family as well as urokinase-type plasminogen activator (uPA) and its inhibitor PAI-1 impact disease course in ovarian cancer. The changes in levels of these factors from primary tumor to omentum metastasis ('level differentials') could thus be associated with metastastic processes. PATIENTS AND METHODS: Protein levels of seven tissue KLK (KLK5-8, 10, 11, 13), uPA, and PAI-1 were determined in extracts of primary tumor tissue and corresponding omentum metastasis of 54 ovarian cancer patients. RESULTS: Higher level differentials of KLK5-8, 10-11, and uPA were associated with residual tumor >10 mm. Residual tumor and larger level differentials of KLK5-7, 10, and uPA were associated with disease progression in the whole cohort. Remarkably, level differentials of KLK5-8 and 10-11 strongly impacted disease progression even in patients with residual tumor mass ≤10 mm; hence, the observed impact of level differentials in KLK5-7 and 10 on disease progression was not simply attributable to their association with surgical success. CONCLUSION: Since they impact both surgical outcome and survival in advanced ovarian cancer, measurement of level differentials could support clinical decisions on surgical and systemic therapy or help in patient selection for novel targeted therapies.

Quantitative expression analysis and prognostic significance of L-DOPA decarboxylase in colorectal adenocarcinoma.

BACKGROUND: L-DOPA decarboxylase (DDC) is an enzyme that catalyses, mainly, the decarboxylation of L-DOPA to dopamine and was found to be involved in many malignancies. The aim of this study was to investigate the mRNA expression levels of the DDC gene and to evaluate its clinical utility in tissues with colorectal adenocarcinoma. METHODS: Total RNA was isolated from colorectal adenocarcinoma tissues of 95 patients. After having tested RNA quality, we prepared cDNA by reverse transcription. Highly sensitive quantitative real-time PCR method for DDC mRNA quantification was developed using the SYBR Green chemistry. GAPDH served as a housekeeping gene. Relative quantification analysis was performed using the comparative C(T) method (2(-DeltaDeltaC(T))). RESULTS: DDC mRNA expression varied remarkably among colorectal tumours examined in this study. High DDC mRNA expression levels were found in well-differentiated and Dukes' stage A and B tumours. Kaplan-Meier survival curves showed that patients with DDC-positive tumours have significantly longer disease-free survival (P=0.009) and overall survival (P=0.027). In Cox regression analysis of the entire cohort of patients, negative DDC proved to be a significant predictor of reduced disease-free (P=0.021) and overall survival (P=0.047). CONCLUSIONS: The results of the study suggest that DDC mRNA expression may be regarded as a novel potential tissue biomarker in colorectal adenocarcinoma.

The use of kallikrein-related peptidases as adjuvant prognostic markers in colorectal cancer.

Several members of the human tissue kallikrein-related peptidase (KLK) family are emerging cancer biomarkers. The aim of this study was to analyse the expression of a panel of KLKs in colorectal cancer and to find out if the multiparametric combination of them can increase the accuracy of prediction of patients survival beyond the traditional clinical information. Nine KLKs (KLK5-8, KLK10, KLK11, KLK13-15) were measured using ELISA assays in cytosolic extracts of 122 colon cancer tissues and their nearby normal mucosa, obtained during surgery. The mean levels of almost all KLKs in tumour tissues were significantly different from their counterparts of normal tissue (P<0.0001). KLK 5, 6, 7, 13, 14 were significantly associated with overall survival in univariate analysis, but after adjusting for age, TNM and differentiation stage, only KLK5 (HR: 1.24 (95% CI: 1.05-1.47)), KLK7 (HR: 1.57 (95% CI: 1.04-2.37)) and KLK14 (HR: 1.43 (95% CI: 1.05-1.94)) remained significant. Addition of a panel of selected KLK markers to clinical parameters gave an increment in AUC of 0.86 beyond the clinical factors at year 1, showing that it can increase the accuracy of prediction of overall survival beyond the traditional clinical information, particularly the short-term (1 year) survival after surgery.

The PRMT1 gene expression pattern in colon cancer.

The methylation of arginine has been implicated in many cellular processes, such as regulation of transcription, mRNA splicing, RNA metabolism and transport. The enzymes responsible for this modification are the protein arginine methyltransferases. The most abundant methyltransferase in human cells is protein arginine methyltransferase 1. Methylation processes appear to interfere in the emergence of several diseases, including cancer. During our study, we examined the expression pattern of protein arginine methyltransferase 1 gene in colon cancer patients. The emerging results showed that the expression of one of the gene variants is associated with statistical significant probability to clinical and histological parameters, such as nodal status and stage. This is a first attempt to acquire an insight on the possible relation of the expression pattern of protein arginine methyltransferase 1 and colon cancer progression.

Extracellular domain of HER2: a useful marker for the initial workup and follow-up of HER2-positive breast cancer.

PURPOSE: The extracellular domain (ECD) of the HER2 receptor is proposed as a real-time marker of HER2-positive breast cancer (BC). In this study, ECD-HER2 levels were compared with standard clinical and pathological prognostic factors. PATIENTS AND METHODS: In 247 consecutive patients (116 with early or localized BC, 116 with advanced or metastatic BC, and 16 with benign mastopathies), serum ECD-HER2 levels were measured. In 116 advanced-disease patients ECD-HER2 status was also studied by immunohistochemistry (IHC) and compared with established clinical and pathological variables. RESULTS: Mean serum ECD-HER2 value was 19.62 ng/ml (median 10.35, range 3-->250). Mean value in benign mastopathies was 9.04 ng/ml, 9.4 ng/ml in early disease and 34.5 ng/ml in advanced disease. No difference between benign mastopathies and early BC was observed, while significant difference between early and advanced BC (p<0.001) was noted. However, in advanced-disease patients a positive correlation of ECD-HER2 with IHC (p=0.002), disease grade (p=0.034) and level II axillary node involvement (p=0.011) was noted, as well as a significant negative correlation with estrogen receptor (ER) and progesterone receptor (PR) (p=0.035 and p=0.011, respectively). CONCLUSION: ECD-HER2 is a reliable marker for breast cancer, as suggested from the existing literature; therefore, its integration in the initial workup and follow-up routine of breast cancer, particularly the HER2-positive, is proposed.

Human tissue kallikrein 7, a novel biomarker for advanced ovarian carcinoma using a novel in situ quantitative method of protein expression.

BACKGROUND: Kallikreins, a subgroup of the serine protease enzyme family, are considered important prognostic biomarkers in cancer. Here, we sought to determine the prognostic value of kallikrein 7 (hk7) in ovarian cancer using a novel method of compartmentalized in situ protein analysis. PATIENTS AND METHODS: A tissue array composed of 150 advanced-stage ovarian cancers, uniformly treated with surgical debulking followed by platinum-paclitaxel (Taxol) combination chemotherapy, was constructed. For evaluation of kallikrein 7 protein expression, we used an immunofluorescence-based method of automated in situ quantitative measurement of protein analysis (AQUA). RESULTS: Mean follow-up time of the cohort was 34.35 months. One hundred and twenty eight of 150 cases had sufficient tissue for AQUA. In univariate survival analysis, low tumor hk7 expression was associated with better outcome for overall survival (OS) and disease-free survival in 3 years (P values 0.032 and 0.037, respectively). In multivariate survival analysis, adjusting for well-characterized prognostic variables, low tumor hk7 expression level was the most significant predictor variable for OS (95% confidence interval 0.125-0.729, P = 0.007). CONCLUSIONS: High tumor hk7 protein expression is associated with inferior patient outcome in ovarian cancer. hk7 may represent a promising prognostic factor in ovarian cancer.

Transcriptional upregulation of human tissue kallikrein 6 in ovarian cancer: clinical and mechanistic aspects.

The human tissue kallikrein family (KLK for protein; KLK for gene) includes 15 members. Twelve kallikreins, including KLK6, are concurrently upregulated in ovarian cancer. However, the mechanism of this phenomenon remains unclear. In this study, we measured KLK6 expression in a large series of ovarian tissue cytosols and examined possible mechanisms of KLK6 up-regulation in ovarian cancer. Using a newly developed enzyme-linked immunosorbent assay (ELISA) with two monoclonal antibodies, we quantified KLK6 expression in ovarian tissue cytosols, and confirmed the upregulation of KLK6 in ovarian cancer and its unfavourable prognostic value. We then examined KLK6 mRNA expression using reverse transcription-polymerase chain reaction and established its good concordance with KLK6 protein expression. This finding suggested that the KLK6 gene is under transcriptional regulation. We then scrutinised a few mechanisms that could explain KLK6 upregulation. The relative abundance of two KLK6 mRNA transcripts was studied; we found the same differential expression pattern in all samples, regardless of KLK6 levels. Genomic mutation screening of all exons and the 5'-flanking region of the KLK6 gene identified two linked single-nucleotide polymorphisms in the 5'-untranslated region, but neither correlated with KLK6 expression. Ovarian cell lines were separately treated with five steroid hormones. None of the treatments produced significant effects on KLK6 expression. We conclude that KLK6 is transcriptionally upregulated in ovarian cancer, but probably not through alternative mRNA transcript expression, genomic mutation, or steroid hormone induction.

mRNA quantification and clinical evaluation of telomerase reverse transcriptase subunit (hTERT) in intracranial tumours of patients in the island of Crete.

Telomerase is a reverse transcriptase that maintains telomeres by adding telomeric TTAGGG repeats to the ends of human chromosomes. The aim of this study was to evaluate quantitatively the mRNA expression of telomerase catalytic subunit (hTERT) in different types of intracranial tumours in relation to their histologic pattern and grade and correlate it with progression-free (PFS) and overall survival (OS) of patients. Human telomerase reverse transcriptase mRNA levels were estimated by the use of real time RT-PCR in 68 samples of intracranial tumours. It revealed statistical correlation between hTERT mRNA expression levels and the grade of the tumours (P<0.001). Patients having negative expression of hTERT mRNA had statistically longer PFS (P=0.031) and OS (P=0.047). Cox univariate regression analysis revealed that hTERT mRNA-positive patients had a high and statistically significant risk of relapse (hazard ratio (HR) of 2.24 and P=0.038). In the Cox multivariate regression model, the levels of hTERT mRNA were adjusted for tumour grade and patients age, and since there was statistically significant relationship between the levels of hTERT mRNA and the grade of the tumours (P=0.003 or P=0.006, respectively), hTERT mRNA levels could not be considered as an independent prognostic factor for PFS or OS.

Quantitative analysis of macrophage inhibitory cytokine-1 (MIC-1) gene expression in human prostatic tissues.

Macrophage inhibitory cytokine-1 (MIC-1) gene is a member of transforming growth factor-beta superfamily and was reported to be highly overexpressed in human prostate cancer using microarray technology. The aim of this study was to evaluate the quantitative expression of MIC-1 in malignant and benign prostate tissues and to associate expression levels with clinicopathological parameters of prostate cancer. Matched (paired) prostatic tissue samples from the cancerous and noncancerous parts of the same prostates were obtained from 66 patients who underwent radical prostatectomy. Quantitative RT-PCR was performed using SYBR Green I on the Roche LightCycler system. Macrophage inhibitory cytokine-1 gene overexpression in cancerous tissues was observed in 88% of cases, compared to noncancerous tissues (P<0.001). The expression level of MIC-1 in cancerous tissues was significantly higher than in noncancerous tissue (P<0.001). Higher expression of MIC-1 gene was significantly associated with higher Gleason score (P=0.004). The expression of the MIC-1 gene in prostate cancer is significantly higher than in noncancerous tissues, especially in more aggressive forms of the disease (Gleason score>5). This is in contrast to prostate-specific antigen that is downregulated in higher-grade tumours. The upregulation of MIC-1 in prostate cancer and in advanced and more aggressive prostatic tumours suggests that MIC-1 protein should be evaluated as a potential diagnostic and prognostic biomarker.

The androgen-regulated gene human kallikrein 15 (KLK15) is an independent and favourable prognostic marker for breast cancer.

Many kallikrein genes were found to be differentially expressed in various malignancies, and prostate specific antigen (encoded by the KLK3 gene) is the best tumour marker for prostate cancer. Prostate specific antigen has recently been shown to be an independent favourable prognostic marker for breast cancer. KLK15 is newly discovered kallikrein gene that is located adjacent to KLK3 on chromosome 19q13.4. KLK15 has 41% similarity to KLK3 and the encoded protein, hK15, can activate pro-prostate specific antigen. We studied the expression of KLK15 by real-time quantitative reverse transcriptase-polymerase chain reaction in 202 tissues from patients with breast carcinoma of various stages, grades and histological types. KLK15 expression was found to be a significant predictor of progression-free survival (hazard ratio of 0.41 and P=0.011) and overall survival (hazard ratio of 0.34 and P=0.009). When all other known confounders were controlled in the multivariate analysis, KLK15 retained its prognostic significance. Higher concentrations of KLK15 mRNA were found more frequently in node negative patients (P=0.042). No association was found between KLK15 expression and any other clinicopathological variable. Further, KLK15 is an independent prognostic factor of progression-free survival and overall survival in the subgroup of patients with lower grade and those with oestrogen receptor and progesterone receptor negative tumours in both univariate and multivariate analysis. KLK15 levels of expression were slightly higher (although not statistically significant) in the oestrogen receptor negative and progesterone receptor negative subgroups of patients. KLK15 is up-regulated by androgens in breast cancer cell lines. Time-course and blocking experiments suggest that this regulation is mediated through the androgen receptor.

Quantitative analysis of human kallikrein gene 14 expression in breast tumours indicates association with poor prognosis.

KLK14 (formerly known as KLK-L6) is a recently identified member of the human kallikrein gene family. This family harbours several genes aberrantly expressed in various cancers as well as established (PSA/hK3, hK2) and potential (hK6, hK10) cancer markers. Similar to other kallikrein genes, KLK14 was found to be regulated by steroid hormones, particularly androgens and progestins, in breast and ovarian cancer cell lines. Preliminary studies indicated that KLK14 is differentially expressed in breast, ovarian, prostatic and testicular tumours. Given the above, we determined the prognostic significance of KLK14 expression in breast cancer. We studied KLK14 expression in 178 histologically confirmed epithelial breast carcinomas by quantitative reverse transcription-polymerase chain reaction and correlated with clinicopathological variables (tumour stage, grade, histotype etc.) and with outcome (disease-free survival and overall survival), monitored over a median of 76 months. KLK14 mRNA levels ranged from 0 to 1,219 arbitrary units in breast cancer tissues, with a mean+/-s.e. of 136+/-22. An optimal cutoff value of 40.5 arbitrary units was selected, to categorise tumours as KLK14-positive or negative. Higher concentrations of KLK14 mRNA were more frequently found in patients with advanced stage (III) disease (P=0.032). No statistically significant association was found between KLK14 and the other clinicopathological variables. KLK14 overexpression was found to be a significant predictor of decreased disease-free survival (hazard ratio of 2.31, P=0.001) and overall survival (hazard ratio of 2.21, P=0.005). Cox multivariate analysis indicated that KLK14 was an independent prognostic indicator of disease-free survival and overall survival. KLK14 also has independent prognostic value in subgroups of patients with a tumour size