Reproducibility of p53 and Ki-67 immunoquantitation in Barrett's esophagus

OBJECTIVE: To test the reproducibility and time Effectiveness of two immunoquantitation and sampling methods in Barrett's esophagus (BE) mucosa. STUDY DESIGN: Measurements were performed using image cytometry (CAS 200/486) with ''at convenience'' sampling and stereology (QPRODIT 5.2) with both at convenience and systematic random sampling. RESULTS: Quantitation of p53 immunohistochemistry (IHC) by CAS was very reproducible by the same observer (r = .99), but its interobserver reproducibility was lower, and the measurement was time consuming (r = .81, 35 minutes). Moreover, CAS also detected a ''signal'' in the absence of any visually observable brown stain, giving false positive results. Both intraobserver and interobserver reproducibility by QPRODIT 5.2 were good. With QPRODIT, using systematic random sampling, the measurements were more reproducible by different observers and faster (r = .96, 10 minutes) than with QPRODIT with at convenience sampling (r = .91, 20 minutes). Therefore, for assessment of the Ki-67 labelling index (LI) and area percentage (area%) of Ki-67-positive nuclei, we used only QPRODIT with systematic random sampling. While intraobserver reproducibility of both Ki-67 LI and area% was good (r = .96, r = .99), interobserver reproducibility of Ki-67 LI was poorer than area% (r = .72 vs, r = .97). The assessment of Ki-67 LI was time consuming, whereas the measurement of the area% of Ki-67-positive nuclei was very time effective (45 vs. 7 minutes). A strong correlation was found between Ki-67 LI and area% of Ki-67-positive nuclei (r = .92). CONCLUSION: Therefore, we conclude that quantitation of p53 and Ki-67 IHC in BE mucosa can be very reproducible and highly time effective by means of stereology with systematic random sampling.