Introduction In its wild type form KLF6 (KLF6-WT) acts as a tumour suppressor gene whose expression is lost in a variety of human cancers. Conversely, KLF-6 splice variant 1 (KLF6-SV1) is able to function as an oncogene playing an important role in tumour proliferation and invasion. The aim of this study was to determine how the expression of these two KLF6 variants in patients with colorectal liver metastases (CRLM) correlates with an established prognostic scoring system.

Methods Patients undergoing resection of CRLM over a 2-year period were identified for inclusion within this study. Immunohistochemistry was performed using antibodies directed against KLF6-WT and KLF6-SV1 on formalin fixed paraffin embedded tumour samples and the proportion of positively stained nuclei for each was determined using an automated computerised image analysis system. Clinical data were collected and Fong's Clinical Risk Score (CRS) was calculated as a marker of patients outcome. Spearmans correlation coefficient was used to determine the relationship between nuclear staining of KLF-6 and the CRS. Continuous variables were compared with Mann–Whitney U test. A p value <0.05 was considered significant.

Results 25 patients (male 64 %) were included in this study with a mean age of 69.2 years (range 52–81). The distribution of CRS were as follows; CRS score 0 n=2, CRS score 1 n=6, CRS score 2 n= 3, CRS score 3 n=10, CRS score 4 n=3, CRS score 5 n=1. Nine patients developed recurrence within 1 year of liver resection. All patients remained alive at the end of the study period. There was a positive correlation between KLF6-SV1 and the calculated CRS (Spearman's correlation efficient = 0.481, p=0.015) whereas no such correlation existed with KLF6-WT. In those with a CRS <3 the mean proportion of positive nuclear staining for KLF6-SV1 was 5% as compared to 15% in those with a CRS≥3 (p<0.05)

Conclusion In this preliminary study, high expression of KLF6-SV1 in patients with CRLM correlates with a higher CRS. Further long-term follow-up data are required to determine what affect this has on prognosis and overall patient survival.

Competing interests K Y D Hui: Grant/Research Support from: Cancer Research UK, the Dowager Countess Eleanor Peel Trust, S Robinson: None declared, G Patman: None declared, D Manas: None declared, H Reeves: None declared, S White: None declared.