Reproducibility of the WHO/IASLC grading system for pre-invasive squamous lesions of the bronchus: a study of inter-observer and intra-observer variation

Abstract

Aims: Although many workers have graded pre-invasive squamous lesions arising in the bronchus, there has been no consensus classification system until the latest edition of the WHO/IASLC histological classification of pulmonary and pleural tumours. Because the value of any such system is dependent on its reproducibility, we have circulated a series of such lesions to a panel of histopathologists to assess interobserver and intra-observer variation when the WHO/IASLC classification was applied.

Methods and results: Colour transparencies of 28 pre-invasive squamous lesions were assessed by six histopathologists (two with a special interest in pulmonary pathology, two generalists and two trainees) on three separate occasions over a period of 3 months, using the criteria of the WHO/IASLC (mild, moderate and severe dysplasia, and in-situ carcinoma). An additional category of metaplasia was added for those cases that showed no dysplasia. Weighted kappa coefficents of agreement (K(w)) were used to evaluate paired observations with a standard quadratic weighting being employed, such that kappa coefficients corresponded to intra-class correlation coefficients. Wilcoxon's sign-ranked test was used to measure the statistical significance of group trends, when comparing kappa values for the three grading systems. Various 3-point systems were also assessed, through combination of the above groups. Intra-observer agreement was substantially better than interobserver variation (mean: 0.71 vs. 0.55). Between the various pathologist groups, inter-observer variation was relatively minor, although intra-observer variation was higher within the trainee pathologist group. Using weighted kappa values, there was no significant difference in either inter-observer or intra-observer agreement between the five point grading system and a 3-point system of metaplasia/mild, moderate and severe/in-situ grades. However, there was a significant increase in variation when a 3-point system of metaplasia/mild, moderate/severe and in-situ carcinoma was used.

Conclusion: This study shows levels of interobserver and intra-observer variation similar to those found in other grading systems in histopathology, with no significant decrease in variability found by abridging the system. The WHO/IASLC system is therefore recommended for future use in both clinical and research fields.