Urothelial neoplasia of the urinary bladder--comparison of interobserver variability for WHO Classification 1972 with WHO/ISUP Consensus Classification 1998.

BACKGROUND

Classification of urothelial bladder tumours is an important factor in the treatment and prognosis of these lesions. Over the years many classifications have been proposed for this purpose. The objective of this study was to classify urothelial neoplasms of the urinary bladder using the latest WHO/ISUP Consensus Classification 1998 and WHO Classification 1972 and compare the two regarding interobserver variability.

METHODS

This study included 100 consecutive biopsy specimens of urothelial neoplasms of the urinary bladder diagnosed at the department of Histopathology, Armed Forces Institute of Pathology, Rawalpindi. These were classified according to WHO Classification 1972 and WHO/ISUP Consensus Classification 1998 by 2 groups of pathologists independently. The tumour categories for WHO classification 1972; papilloma, and transitional cell carcinoma (TCC) grades I, II and III were compared with the WHO/ISUP Consensus Classification entities of papilloma, papillary neoplasm of low malignant potential, low grade and high grade papillary carcinomas. Kappa statistics were used to evaluate interobserver variability. Chi square test was used to calculate significance.

RESULTS

There was agreement on 80 tumours between the two groups of histopathologists when using WHO classification 1972 while there was agreement on 95 tumours using WHO/ISUP consensus classification. The value of Kappa for WHO Classification was 0.68 (good agreement) whereas for WHO/ISUP Consensus Classification it was 0.91 (excellent agreement). The difference between the two systems was statistically significant (p<0,001). Kappa values were less for benign and borderline lesions using both systems.

CONCLUSIONS

WHO/ISUP Consensus Classification 1998 showed less interobserver variability than WHO Classification 1972 in the evaluation of bladder tumours. It was found easier to apply by both groups. There was less agreement on the benign and borderline lesions using both the classifications.