Umeå universitets logga

Colorectal cancer is ranked third in worldwide incidence for women and fourth for men representing ͌ 9% of the world cancer or approximately 1 million new cases for 2002 (Parkin et al., 2005). Two thirds of colorectal cancers are located in the colon and one third in the rectum. At diagnosis approximately one third of all patients with colorectal cancer has lymph node positive disease, one third has lymph node-negative disease, and one third has distant metas-tases (Benson et al., 2004). The principal curative treatment for colorectal cancer is surgery. Adjuvant chemotherapy given to lymph node positive colon cancer patients has been shown to increase the survival rate (Haydon, 2003). In rectal cancer patients preoperative irradiation therapy is given to reduce local recurrences and has also been shown to improve survival (Folkesson et al., 2005). Still with these improved treatment modalities only approximately half the number of patients will survive for 5 years. For example, Swedish results for the time period 1995–1999 show a 5-years relative survival of ≈ 57% for both genders (Birgisson et al., 2005).

Tumor stage, based on histopathologi-cal examination of the resected specimen, and perioperative findings predict survival. Relative 5-year survival in Dukes' A (T1-2N0M0, Stage I) is 90–95%, Dukes' B (T3-4N0M0, Stage II) 60–80%, Dukes' C (anyTN1-2M0, Stage III) 40–60% and Dukes' D (anyTN0-2M1, Stage IV) < 5% (Staib et al., 2002). Besides distant metas-tases the most important prognostic indicator is the status of the regional lymphatic field showing presence or absence of tumor cells in regional lymph nodes. Given the importance of correctly identifying Dukes' C patients, i.e., patients with lymph node involvement who are eligible for chemotherapy, we have focused on improving the methods for detecting disseminated tumor cells in regional lymph nodes.