Comparison between endoscopic and surgical sampling for the measurement of potential doubling time in colorectal cancer.

In colorectal cancer, tumoral kinetic parameters could influence the therapeutic strategy. The aim of this study was to test the validity of a preoperative assessment of tumoral kinetics on biopsies obtained from endoscopic sampling. After infusion of 250 mg of bromodeoxyuridine (BrdUrd), tumor samples were taken. The following kinetic parameters were evaluated by Begg's method: labeling index (LI) of BrdUrd, duration of the S phase (Ts), and potential doubling time (Tpot). We compared the findings in preoperative endoscopic samples (t1), endoscopic samples (t2), and macrobiopsies of the tumor after surgical resection (t3) to test the effect of time, surgical procedures, and sampling mode on results. Eight diploid and 11 aneuploid colorectal cancers were studied. In diploid tumors, no difference in Tpot, Ts, or LI values was found among the three time intervals. In aneuploid tumors, Tpot value at t1 was lower than t3: 2.9 +/- 1.2 versus 4.5 +/- 1.4 days (P = 0.02). This difference seemed to be independent of sampling type but related instead to the time interval between BrdUrd infusion and biopsy and to the changes due to the surgical condition. On the other hand, LI seemed to be the most stable kinetic parameter. With a cutoff value of 4 days for Tpot and of 10% for LI to differentiate slow and fast growing tumors, the positive predictive values of LI were 90% at t1, 80% at t2, and 50% at t3 in aneuploid tumors.