Rectal epithelial cell kinetics are used as intermediate markers for colorectal cancer and relate to risk. In this study, measures of proliferation using direct immunohistochemistry for proliferating cell nuclear antigen (PCNA) were compared to in vitro labeling by bromodeoxyuridine (BrdUrd) and incubated biopsies that were later stained for PCNA (PCNA-I) in human rectal biopsies. The study group consisted of 20 sets of biopsies from 12 subjects participating in an intervention trial. Fresh nonincubated biopsies were fixed in methacarn and stained immunohistochemically for PCNA (clone 19A2). In parallel biopsies, BrdUrd was incorporated into the DNA of S-phase cells during a 2-h incubation at 37 degrees C under hyperbaric conditions and localized by immunohistochemistry. Additionally, biopsies were incubated under hyperbaric conditions for 2 h at 37 degrees C, fixed in methacarn, and stained for PCNA (PCNA-I). There was a highly significant difference in the labeling index between the three methods (P < 0.01), but there was no significant difference between subjects (P = 0.439). The mean labeling index was 2.3 +/- 0.1% for PCNA, 2.9 +/- 0.1% for PCNA-I, and 4.1 +/- 0.1% for BrdUrd. The proportion of labeled cells in the top two-fifths was significantly higher (P = 0.01) for BrdUrd (5.5 +/- 0.8%) and PCNA-I (6.4 +/- 1.1%) compared to PCNA (3.1 +/- 0.6%), and a significant difference was seen between subjects (P = 0.038). PCNA-I and BrdUrd methods had similar crypt heights with 73.5 +/- 1.8 and 71.2 +/- 1.3 cells/crypt column, respectively, but were significantly shorter (P < 0.001) than PCNA with 83.4 +/- 1.5 cells/crypt column, indicating a loss of cells during organ culture. The simplicity of the PCNA technique, which avoids potential perturbations occurring during organ culture, has considerable appeal as a marker for colorectal cancer risk, but additional studies are needed to correlate PCNA with neoplastic risk.

This content is only available via PDF.