Swedish researchers win Best Abstract in Oncology

12 April 2014

A group of Swedish researchers have won the first prize for Best Abstract in Oncology. Lead author Oliver Patschan (Malmö, SE) and colleagues bagged the top prize for their work on the classification of T1 bladder cancer which aims to identify high-risk subtypes using marker genes. Below is the unedited version of the prize-winning abstract:

Abstract Nr: AM14-1750: Molecular classification of T1 urothelial bladder cancer identifies high-risk subtypes

Authors: Patschan, O.1, Sjödahl, G.2, Chebil, G.2, Lövgren, K.2, Lauss, M.2, Gudjonsson, S.1, Kollberg, P.3, Eriksson, P.2, Aine, M.2, Månsson, W.1, Fernö, M.2, Liedberg, F.1, Höglund, M.2, Lund Bladder Cancer Group (LBCG)

1Skåne University Hospital, Dept. of Urology, Malmö, Sweden, 2Skåne University Hospital, Dept. of Oncology, Lund, Sweden, 3Helsingborg Hospital, Dept. of Urology, Helsingborg, Sweden

Introduction & Objectives
A significant proportion of patients with stage T1 urothelial bladder cancer will progress to muscle invasive tumours and need radical surgery, whereas the remaining may show several recurrences without progression, treated with transurethral resections and BCG only. This study was aimed to investigate if molecular subtypes of T1 tumours may be informative with regard to tumour progression.

Materials & Methods
167 primary T1 tumours were characterized by immunohistochemistry (IHC) using eleven marker genes known to be informative of the urothelial molecular subtypes urobasal (Uro), genomically unstable (GU), and SCC-like (SCCL). Subtypes were correlated with clinical parameters multifocality, tumour size, concomitant CIS, invasion depth, lymphovascular invasion, tumour infiltrating CD3 positive cells, with progression to muscle invasion or metastasis as endpoint. Median follow up time was 78 months.

T1 tumours were classified as Uro (32%), GU (58%), and SCCL (10%) using a previously developed classifier. The classification was validated by application of additional IHC markers with known subclass expression patterns. Kaplan-Meier analyses revealed Uro cases to have a low and GU/SCCL to have a high risk for progression. Multifocality, concomitant CIS, invasion depth, lymphovascular invasion, and CD3+ infiltrating lymphocytes were significantly associated with GU/SCCL tumours. In multivariate cox-regression analysis EORTC score (HR 2.36, 95% CI 1.16-4.80, p=0.018), CD3 infiltration (HR 2.32, 95% CI 1.05-5.10, p=0.037), and BCG treatment (HR 0.42, 95% CI 0.18-0.96, p=0.039) were significant independent predictors of progression. High EORTC scores and high CD3 counts are associated with early progression, and almost exclusively observed in GU/SCCL cases.

Urothelial bladder cancer stage T1 may be classified into three different molecular subtypes with distinct clinicopathological properties, the most important being a strong predisposition of GU/SCCL tumours to progress. GU/SCCL tumours with particularly short progression free time correspond to high scoring EORTC or CD3 cases.