Prostate cancer is the leading cause of cancer death in men in economically developed countries 1 and there was an estimated 328 000 cases in Europe in 2008 which makes it the most common form of cancer in men 2. For localized cancer, radical prostatectomy is the most common treatment and it has shown a benefit in cancer specific survival in comparison to watchful waiting 3. Nevertheless, in some cases the cancer will recur with detectable prostate specific antigen (PSA) concentrations in blood, known as biochemical recurrence (BCR) 4. Apart from PSA, no diagnostic or prognostic biomarkers have yet been incorporated into clinical protocols for management and risk stratification of prostate cancer patients despite extensive research efforts 56. Thus, there is still a great need for new biomarkers to predict the course of the disease and differentiate indolent from life threatening cancer.

The RNA-binding motif protein 3, RBM3, is a glycine rich protein with an RNA recognition motif (RRM) capable of binding to both RNA and DNA and it is one of the earliest proteins synthesized in response to cold shock 78. RBM3 is up-regulated in various types of human malignancies 91011 and the role of RBM3 as a putative cancer biomarker was originally unravelled using an antibody-based discovery approach http://www.proteinatlas.org 121314. Since then, RBM3 protein expression, in particular its nuclear localization, has been demonstrated to be associated with a significantly improved survival in several cancer forms, e.g. breast cancer 11, ovarian cancer 15, malignant melanoma 16 and colorectal cancer (Hjelm et al, Proteomics Clinical Applications, in press). In ovarian cancer, RBM3 has also been demonstrated to be an independent factor of good prognosis at the gene expression level 15 and in vitro data point towards a link between RBM3 and cisplatin sensitivity in ovarian cancer cells 15, possibly by regulating several cellular processes involved in maintenance of DNA integrity 17.

In this study, we investigated the prognostic impact of RBM3 expression in prostate cancer by immunohistochemical (IHC) analysis of benign and malignant specimens from 88 patients treated with radical prostatectomy.

The patient cohort consists of 88 patients, between 48-74 years, treated with radical prostatectomy for localized prostate cancer at Skåne University Hospital, Malmö, from 1998 - 2003. Patient and tumour characteristics are shown in Additional File 1. Written consent was obtained from the patients and the local ethics committee at Lund University approved the study. Histopathological, clinical and follow-up data were obtained from the clinical- and pathology records. Vital status and cause of death was obtained from the Swedish Cause of Death Registry up until December 2006. The mean follow up period was 58.5 months after which three patients had died and 85 patients were alive. BCR was defined as PSA > 0.2 ng/ml 18 with a confirmatory level and clinical progression included BCR, death from prostate cancer and adjuvant radiation therapy. During follow up, 14 (15.9%) patients had BCR and 19 (21.6%) patients had clinical progression. Five (5.7%) patients received adjuvant radiotherapy and three (3.4%) patients received adjuvant hormonal therapy.

For tissue microarray (TMA) construction, areas representing cancer and normal prostate tissue were marked on haematoxylin and eosin stained slides, whereby normal tissue was sampled in a noncancerous zone away from the tumour, in most cases the transition zone. Two 1.00 mm cores from areas with invasive cancer and one 1.00 mm core representing benign prostate tissue were sampled from each case and mounted in a new recipient block using a manual arraying device (MTA-1 Beecher Instruments Inc., Sun Prairie, WI, USA). For immunohistochemical analysis, 4 μm TMA-sections were automatically pre-treated using the PT Link system and then stained in an Autostainer Plus (DAKO; Glostrup, Copenhagen, Denmark) with the mouse monoclonal anti-RBM3 antibody AAb030038; Atlas Antibodies AB, Stockholm, Sweden, diluted 1:5000. The specificity of the antibody has been validated previously 15. RBM3 staining was evaluated by two independent observers (LJ and KJ) who were blinded to clinical and outcome data. The fraction of cells expressing RBM3 in the nucleus (NF) were denoted a score from 0 (0-1%), 1 (2-25%), 2 (26-50%), 3 (51-75%) and 4 (> 75%) and the intensity (NI) was scored as 0 (negative), 1 (weak), 2 (moderate) and 3 (strong). A combined nuclear score (NS) of NF × NI, was then constructed as previously described 15. Cytoplasmic staining intensity was graded as 0 (no staining), 1 (weak staining) and 2 (strong staining).

Spearman's rho and Chi-square tests were used for analysis of the correlation between RBM3 expression and relevant clinicopathological characteristics. Kaplan-Meier analysis and log rank test were used to illustrate differences in BCR free time and progression free survival (PFS) according to RBM3 expression. For survival analyses, RBM3 expression (intensity × fraction) was dichotomized into weak vs strong using classification and regression tree analysis (CRT). Cox regression proportional hazards modelling were used to estimate the impact of RBM3 expression on BCR (with RBM3 as a continuous variable) and PFS (with RBM3 as a dichotomized variable) in both uni- and multivariate analysis. All tests were two sided. A p-value of 0.05 was considered significant. All statistical analyses were performed using SPSS version 17 (SPSS Inc, Chicago, IL).

Our results demonstrate that RBM3, while rarely expressed in normal prostatic gland epithelium, is up-regulated in PIN and invasive prostate cancer, and that patients with tumours expressing high nuclear levels of RMB3 have a significantly prolonged time to biochemical recurrence and clinical progression, also after adjustment for conventional prognostic factors. These findings are in line with previous findings in breast cancer 11, ovarian cancer 15, malignant melanoma 16 and colorectal cancer (Hjelm et al, Proteomics Clinical Applications, in press). Thus, there are increasing evidence for RBM3 being a biomarker of good prognosis in multiple cancer forms. The mechanisms behind these findings remain to be elucidated, an undertaking that might be somewhat challenging as current in vitro data point towards a proto-oncogenic role for RBM3 8919 and, hence, do not readily seem to support the clinical situation. However, the common hypothesis that oncogene activation is associated with an aggressive tumour phenotype does not always hold true, as well exemplified by the association between microsatellite instability and good prognosis in colorectal cancer 20. It is evident that RBM3, while sparsely expressed in normal tissue, is up-regulated in most cancer forms and their pre-invasive stages in vivo. In light of the recently proposed association between RBM3 and DNA integrity in ovarian cancer 17, it could be speculated that RBM3 might play an important role in promoting early stages of tumourigenesis by interfering with the anti-cancer barrier provided by various DNA damage checkpoint mechanisms 212223. On the other hand, once a tumour has been established, an attenuated capability of DNA-damage response caused by RBM3 over expression might hinder the pressure for selection of more malignant clones 2122. Further studies are warranted to explore these associations.

In this study, we used a well-validated monoclonal antibody for detection of RBM3 expression 1516. Immunohistochemistry has several advantages compared to other assays in that it can easily be incorporated into clinical protocols and allows for assessment of the subcellular localization of proteins, which might have important prognostic implications. Our results indicate that immunohistochemical assessment of RBM3 expression in formalin-fixed paraffin embedded tumour samples could be a useful tool for prognostication and treatment stratification of prostate cancer patients. Since this was a small study these findings should be further validated in larger studies, preferably tumours from prospective, clinical trials.