Correlation of Quantitative Image Analysis Versus Pathologist Estimate of Proliferation Rate in Diffuse Large B-Cell Lymphoma and Corresponding Outcomes
D Chabot-Richards, DA Martin, O Myers, KE Hunt. U. of New Mexico, Albuquerque, NM
Background: Proliferation rates in Diffuse Large B-cell Lymphoma (DLBCL) have been associated with conflicting outcomes in the literature, more often with high proliferation associated with poor prognosis. In the majority of these cases, the proliferation rate was estimated by a pathologist using Ki67 immunohistochemistry (IHC). We hypothesized that a quantitative image analysis (QIA) algorithm would give a more accurate Ki67%, leading to more accurate associations with survival.
Design: Cases of diffuse large B-cell lymphoma (DLBCL) were selected from our database in accordance with WHO criteria. Ki67 percent positivity estimated by the pathologist was recorded from the original report. The same slides used for this assessment were then scanned using the Aperio ImageScope and Ki67 percent positivity was calculated using a computer-based quantitative IHC nuclear algorithm. In addition, chart review was performed and survival time was recorded. Survival time was compared to both pathologist and QIA Ki67% using Wilcoxon and log-rank tests.
Results: We found 60 cases of DLBCL with a Ki67 slide available. The Ki67% estimated by the pathologist from report versus QIA were significantly correlated (p<0.001) but with report values significantly larger than QIA (paired t-test, p=0.021). There was less agreement at lower Ki67%. Comparison of Ki67% versus survival did not show significant association either with pathologist estimate or QIA. However, while not significant, there was a trend of worse survival at higher proliferation rates detected by the pathologist report but not by QIA.
Conclusions: While there is significant correlation between pathologist estimates and QIA of Ki67%, pathologists tend to overestimate the %. Initial data, while not achieving statistical significance, implies a worse survival with higher proliferation rates. Interestingly, our data suggest that the Ki67% assessed by the pathologist may be more closely associated with survival outcome than that identified by QIA. This may indicate that pathologists are better at selecting appropriate areas of the slide. More cases are needed to assess whether this finding would be statistically significant. Due to the good correlation between pathologist estimate and QIA, there is no substantial benefit to using QIA over pathologist estimate at this point in time.
Monday, March 22, 2010 2:15 PM
Platform Session: Section B, Monday Afternoon