The impact of grading scheme on non-muscle invasive bladder cancer progression: potential utility of hybrid grading schemes

Abstract

Non-muscle invasive bladder cancer (NMIBC) grade is a major determinant of progression risk. The most widely utilised grading systems are the World Health Organization (WHO) 1973 and 2004 schemes. Recent publications suggest the utility of combining both into a four-tier or a hybrid three-tier system, subdividing WHO 2004 high grade into two separate categories while maintaining low grade as a single group. We identified two retrospective cohorts of bladder resections/biopsies of papillary urothelial NMIBC with long term clinical follow-up. The sentinel specimen was assessed for WHO 2004 and 1973 grade, along with pathological stage and carcinoma in situ. Each case was additionally stratified into a hybrid three-tier system (low grade; high grade, grades 2 and 3) and a four-tier system (low grade, grades 1 and 2; high grade, grades 2 and 3). Uni- and multivariable analysis for progression and event free survival (PFS/EFS) were calculated along with the time dependent area under the curve (AUC) for each grading scheme. There were 609 cases (Cohort A, n=343; Cohort B, n=266), including 449 (74%) pTa, 156 pT1 (26%) and four pTx with 338 (56%) low grade (177, grade 1; 161, grade 2) and 271 (44%) high grade (137, grade 2; 134, grade 3). A total of 108 patients progressed (17.7%): 97 high grade, (grade 3, n=59; grade 2, n=38). Multivariable analyses of PFS with the hybrid 3- and 4-tier systems showed higher Harrell's concordance indices (0.851 and 0.853, respectively) than WHO 1973 (0.844) and WHO 2004 (0.846). In both cohorts AUC values were higher (0.77-0.85) for the two hybrid grading systems compared to WHO 1973 or WHO 2004 (0.72-0.82). Similar results were seen on analysis of EFS. The data support the use of a hybrid three-tier or four-tier grading system to improve stratification of NMIBC patients.