Correlation between iododeoxyuridine and MIB-1 labelling index in gastric carcinoma and adjacent normal gastric tissue

Abstract

Background: In vivo labelling with synthetic thymidine analogues, such as Iododeoxyuridine (IdU) and Bromodeoxyuridine (BrdU), has frequently been used to estimate tumour proliferation. However, this method requires intravenous administration of IdU or BrdU, thymidine analogues that are potential mutagens. Recently, the monoclonal antibody MIB-1 has been developed, recognizing the Ki-67 nuclear antigen, which is associated with cell cycle proliferation and is found throughout the cell cycle (G1, S, G2 and M phases), but not in resting (G0) cells. We studied the correlation between the MIB-1 labelling index (LI) and the IdU labelling index in normal and malignant gastric tissue.

Patients and methods: Twenty patients with gastric cancer received an intravenous injection of IdU (200 mg/m2) before surgery. Specimens were obtained from gastric carcinoma and adjacent normal gastric tissue. The samples were fixed in formalin and immunohistochemical analyses of IdU LI and MIB-1 LI were performed. The LI was defined as the percentage of labelled nuclei of 5000 nuclei counted.

Results: The IdU LI ranged from 3.3% to 18.2% in gastric carcinoma and from 0.5% to 5.6% in adjacent normal gastric mucosa, whereas the MIB-1 LI ranged from 4.2% to 46.0% in gastric cancer and from 1.3% to 25.1% in adjacent normal gastric mucosa. Comparison of IdU LI with MIB-1 LI, using the Spearman rank correlation coefficient test showed a significant correlation between IdU LI and MIB-1 LI in normal gastric tissue (r = 0.63, p < 0.05). However, in gastric carcinoma no significant correlation was found between either proliferation marker (r = 0.07, N.S.).

Conclusion: MIB-1 accurately reflects the in vivo IdU LI in normal gastric tissue, whilst in gastric carcinoma the MIB-1 LI does not seem to be a substitute for the in vivo IdU LI.