The clinical use of genetic analyses in colorectal cancer

Abstract: Background: Colorectal cancer (CRC) is a common global disease, with a mortality rate of almost 50%. Prognosis is mainly based on the TNM classification. Surgical interventions have the potential of being curative in patients with stage I-III CRC. Adjuvant treatment with chemotherapy enhances the survival rate, especially in stage III cancer. Chemotherapy does, however, have significant side effects. Therefore, refinement of therapies based on improved prognostic ability on an individual level is essential. One way to achieve this could be to examine the tumours on a molecular level and not just histologically. Recent studies have shown that K-ras mutation is a negative predictor to anti-EGFR (epidermal growth factor receptor) therapy. MicroRNAs were discovered only 2 decades ago but are now the most promising biomarkers in many cancers. In this thesis, the first two papers focus on genetic changes in the tumour and in lymph nodes, in particular looking at the oncogene K-ras. The last two papers focus on microRNAs in tumours and in blood serum. In studies 2, 3 and 4, our findings on the correlation of some molecular changes to prognosis is described. Study I: 17 tumours from CRC patients were divided in to small cubes. DNA was extracted from each biopsy and the occurrence of K-ras mutations, methylation of p16 and MGMT, and loss of heterozygosity (LOH) at 5q, 17p and 18q were analysed. We found that the distribution of methylated p16 and MGMT and LOH at 5q, 17 p and 18q are heterogeneous and present in a large majority of CRC tumours, thereby of limited prognostic value. However, K-ras mutation appears more homogeneously spread, a finding of clinical relevance for the use of biopsies to predict anti-EGFR response. Study II: 99 stage II CRC patients with histologically normal lymph nodes were included. DNA was extracted from lymph nodes, tumours and normal mucosa and the K-ras status was analysed and correlated to prognosis. Of the tumours, 34/99 were identified as positive for K-ras mutations. Of these, 10 patients also expressed K-ras mutations in their lymph nodes. Of the 10 patients with positive lymph nodes, 7 (70%) relapsed and died from the disease within 60 months compared to 8/ 24 (33%) with K-ras negative lymph nodes. Study III: 50 CRC patients were studied. RNA was extracted from the tumours. 5 patients with short and 5 patients with long survival were selected for SYBR-green quantitative PCR-based array to screen for differently expressed microRNAs. From this screening, 6 candidate prognostic microRNAs were validated using TaqMan quantitative PCR in all 50 patients. We found that high expression of miR-185 and low expression of miR-133b correlated to poor survival (p=0.001 and p=0.028, respectively) and metastasis (p=0.007 and p=0.036, respectively) in CRC. Study IV: 16 CRC patients and one with a large adenoma in the colon were included. All patients underwent radical (R0) surgery. Blood serum was collected prior to and 30 days after surgery. 3 microRNAs were analysed with Taqman qPCR (miR-21, miR-133b and miR-185). The serum levels of mir-21 were not affected by radical tumour resection. There was a significant decrease in the level of miR-133b among the patients following surgery, and an overall reduction of miR-185. There was no correlation between intra-individual changes in serum levels pre- and postoperatively to disease outcome, or between baseline levels and the risk of recurrent disease.