Atypical urine cytology and the Johns Hopkins Hospital template: the University of Chicago experience

Impact of implementing the first edition of the Paris system for reporting: A systematic review and meta-analysis

Urine cytology is a noninvasive, widely used diagnostic tool for screening and surveillance of genitourinary tract neoplasms. However, the absence of unified terminology and clear objective morphological criteria limits the clinical benefit of urine cytology. The Paris System for Reporting Urine Cytology (TPS) was developed with the goal of standardizing reporting and improving urine cytology performance in detecting high-grade malignancy (HGM). We aimed to evaluate potential effects of TPS on improving urine cytology diagnostic performance and clinical utility by conducting a systematic review and meta-analysis. We searched six electronic databases to identify cross-sectional and cohort studies written in English assessing the accuracy of urine cytology in detecting genitourinary tract malignancies of patients under surveillance or with clinical suspicion of malignancy from January 2004 to December 2022. We extracted relevant data from eligible studies to calculate relative distribution of cytology diagnostic categories; ratio of atypical to HGM cytology diagnosis; and risk of HGM (ROHGM) and HGM likelihood ratio (HGM-LR) associated with cytology diagnostic categories. We used a generalized linear mixed model with logit transformation to combine proportions and multilevel mixed-effect logistic regression to pool diagnostic accuracy measurements. We performed meta-regression to evaluate any significant difference between TPS and non-TPS cohorts. We included 64 studies for 99,796 combined total cytology samples, across 31 TPS and 49 non-TPS cohorts. Pooled relative distribution [95% confidence interval (CI)] of negative for high-grade urothelial carcinoma (NHGUC)/negative for malignancy (NM); atypical urothelial cells (AUC); suspicious for high-grade urothelial carcinoma (SHGUC)/suspicious for malignancy (SM); low-grade urothelial neoplasm (LGUN); and HGM categories among satisfactory cytology cases were 83.8% (80.3%-86.9%), 8.0% (6.0%-10.6%), 2.2% (1.4%-3.3%), 0.01% (0.0%-0.1%), and 4.2% (3.2%-5.5%) in TPS versus 80.8% (76.8-2.7%), 11.3% (8.6%-14.7%), 1.8% (1.2%-2.7%), 0.01% (0.0%-0.1%), and 3.3% (2.5%-4.3%) in non-TPS cohorts. Adopting TPS classification resulted in a significant increase in the frequency of NHGUC and a reduction in AUC cytology diagnoses, respectively. The AUC/HGM ratio in TPS cohort was 2.0, which showed a statistically significant difference from the atypical/HGM ratio of 4.1 in non-TPS cohort (p-value: 0.01). Moreover, the summary rate (95% CI) of LGUN called AUC on cytology significantly decreased to 20.8% (14.9%-28.3%) in the TPS compared with 34.1% (26.4%-42.8%) in non-TPS cohorts. The pooled ROHGM (95% CI) was 20.4% (6.2%-50.0%) in nondiagnostic (NDX), 15.5% (9.6%-24.2%) in NHGUC, 40.2% (30.9%-50.2%) in AUC, 80.8% (72.9%-86.8%) in SHGUC, 15.1% (5.7%-34.3%) in LGUN, and 91.4% (87.3%-94.3%) in HGM categories in TPS studies. NHGUC, AUC, SHGUC, and HGM categories were associated with HGM-LR (95% CI) of 0.2 (0.1-0.3), 0.9 (0.6-1.3), 6.9 (2.4-19.9), and 16.8 (8.3-33.8). Our results suggest that TPS 1.0 has reduced the relative frequency of AUC diagnosis, AUC/HGM ratio, and the frequency of LGUNs diagnosed as AUC on cytology. Adopting this classification has improved the clinical utility of SHGUC and HGM cytology diagnoses in ruling in high-grade lesions. However, an NHGUC diagnosis does not reliably rule out the presence of a high-grade lesion.

Examining longitudinal markers of bladder cancer recurrence through a semiautonomous machine learning system for quantifying specimen atypia from urine cytology

BACKGROUND

Urine cytology is generally considered the primary approach for screening for recurrence of bladder cancer. However, it is currently unclear how best to use cytological examinations for assessment and early detection of recurrence, beyond identifying a positive finding that requires more invasive methods to confirm recurrence and decide on therapeutic options. Because screening programs are frequent, and can be burdensome, finding quantitative means to reduce this burden for patients, cytopathologists, and urologists is an important endeavor and can improve both the efficiency and reliability of findings. Additionally, identifying ways to risk-stratify patients is crucial for improving quality of life while reducing the risk of future recurrence or progression of the cancer.

METHODS

In this study, a computational machine learning tool, AutoParis-X, was leveraged to extract imaging features from urine cytology examinations longitudinally to study the predictive potential of urine cytology for assessing recurrence risk. This study examined how the significance of imaging predictors changes over time before and after surgery to determine which predictors and time periods are most relevant for assessing recurrence risk.

RESULTS

Results indicate that imaging predictors extracted using AutoParis-X can predict recurrence as well or better than traditional cytological/histological assessments alone and that the predictiveness of these features is variable across time, with key differences in overall specimen atypia identified immediately before tumor recurrence.

CONCLUSIONS

Further research will clarify how computational methods can be effectively used in high-volume screening programs to improve recurrence detection and complement traditional modes of assessment.