Under-reporting of carcinoma in situ in patients with high-grade papillary non–muscle-invasive bladder cancer

Abstract

Background

Concomitant carcinoma in situ (CIS) with high-grade papillary urothelial carcinoma impacts patient risk stratification and treatment. However, on identification of a high-grade papillary lesion, pathologists may be disinclined to search for possible CIS. Several newer agents for bacillus Calmette-Guérin–unresponsive non–muscle-invasive bladder cancer (NMIBC) are indicated specifically for patients with CIS ± papillary tumors; hence failure to detect/report CIS may limit treatment options.

Objective

To assess CIS frequency in patients with high-grade Ta/T1 NMIBC.

Methods

In this retrospective cohort study, transurethral resection of bladder tumor biopsy specimens with an accompanying diagnosis of high-grade papillary NMIBC without CIS were identified in the PathNet database (2023–2024) and re-reviewed by a single pathologist for the presence of concomitant CIS.

Results

In total, 316 cases of high-grade papillary urothelial carcinoma (254 Ta, 62 T1 disease) were identified for re-review. Mean patient age was 73 (range 27–93) years; 81.0% of patients were male. On re-review, CIS was identified in 80 cases (25.3%, 95% CI 20.8%–30.4%). CIS frequency was higher in T1 disease (19/62 cases [30.6%]) vs Ta disease (61/254 cases [24.0%]). Among the 8 pathologists who performed the initial review, CIS-positivity rates in their individual biopsy samples varied from 15.0% (6/40) to 66.7% (2/3).

Conclusion

This study suggests that ∼25% of patients with high-grade papillary urothelial carcinoma may have unreported CIS. Close communication between the urologist and pathologist is crucial in ensuring specific exclusion of CIS in all bladder specimens, since a diagnosis of concomitant CIS may impact patient management.

Introduction

Approximately 70% to 80% of cases of urothelial carcinomas are detected at the stage of non–muscle-invasive bladder cancer (NMIBC) ¹ . To help guide treatment decisions, NMIBC is risk-stratified based on clinical and pathologic factors, including tumor size, stage, and grade, and the absence/presence of urothelial carcinoma in situ (CIS) ² . According to the AUA/SUO risk stratification system, high-risk NMIBC includes multifocal or recurrent high-grade Ta disease, high-grade Ta tumors >3 cm in diameter, high-grade stage T1 disease, and CIS ³ . Among these, CIS is particularly concerning due to its strong association with progression to muscle-invasive disease and increased mortality risk ⁴ . If left untreated, approximately 50% of CIS cases progress to muscle-invasive disease within five years, and even with treatment, 30%–40% progress within ten years ⁵ . As such, the presence of CIS impacts risk stratification, treatment selection, and, in the US, insurance coverage. According to the NCCN bladder cancer guidelines, therapies such as intravenous pembrolizumab and intravesical nadofaragene firadenovec carry a level 2A recommendation for use in bacillus Calmette-Guérin (BCG) ̶ unresponsive high-grade NMIBC when CIS is present (with or without papillary disease), but only a level 2B recommendation for management of BCG- unresponsive high-grade NMIBC without CIS ⁶ . Failure to detect or document concomitant CIS leads to a risk of undertreatment due to inaccurate risk stratification and can restrict patient access to potentially beneficial therapies because of health plan coverage limitations.

Inter-observer variability among pathologists in the diagnosis of bladder tumors remains a longstanding concern despite improvements in standardization of tissue sample handling and reporting, and quality control ¹ . This variability may stem in part from challenges in accurately identifying CIS, which can be subtle and difficult to distinguish from other histologically atypical flat urothelial lesions, including urothelial hyperplasia, atypia and dysplasia ⁷ . Additional contributing factors include the number and quality of transurethral resection of bladder tumor (TURBT) specimens provided to the laboratory, the degree of collaboration between the urologist and pathologist, and the thoroughness of the biopsy examination^8–10.

The present study investigates whether re-review of NMIBC cases originally diagnosed as high-grade Ta or T1 papillary NMIBC without CIS may reveal previously undetected concomitant CIS.

Methods

This retrospective analysis was based on a re-examination of TURBT pathology specimens submitted during the period November 2023 – November 2024 to PathNet (Little Rock, AR), a community-based multidisciplinary pathology laboratory, from urology clinics across the continental US (predominantly the Midwest, West and South census regions). Bladder specimens identified in the PathNet database with the descriptor ‘high-grade papillary urothelial carcinoma without concomitant CIS’ were selected for re-review.

The original pathology review was conducted at the PathNet laboratory by a total of 8 pathologists, comprising fellowship-trained general, soft tissue, and gynecological pathologists. For the pathology re-review, a single pathologist experienced in reading genitourinary specimens (who had previously performed some of the original evaluations) was assigned to assess the selected bladder specimens for the presence of concomitant urothelial CIS in addition to the initially diagnosed high-grade papillary urothelial carcinoma. During re-review, any lesions in immediate proximity to the papillary neoplasm (‘shoulder lesions’) were excluded from consideration. Patient demographic data (where available) and diagnostic data were collected from the PathNet database, including age, sex, number of TURBT samples, original diagnostic stage, ICD-10 code, and the name of the reading pathologist. Demographic data were not available for all patients, as specimens were submitted from a diverse group of practices without standardized data collection protocols.

Primary outcomes of interest included the incidence of CIS detection, stratified by initial diagnostic stage and by the original diagnosing pathologist.

Results

Cohort demographics and clinical features

Screening of the PathNet database record for the period November 2023 – November 2024 identified a total of 316 patients who were diagnosed with high-grade papillary NMIBC without concomitant CIS, of whom 254 patients had Ta stage disease (80.4%) and 62 patients had T1 stage disease (19.6%) (Table 1). ICD-10 codes variously indicated malignant neoplasm of the bladder (C67.x), neoplasm of unspecified/uncertain behavior of the bladder (D49.4/D41.4), or hematuria (R31.x), but only 2 cases of CIS of the bladder (D09.0) were referenced – neither of which were confirmed on pathology re-review. The diagnoses were made by eight different pathologists.

Of the patients with available demographic data (n = 190), 155 (81.6%) were male and 35 (18.4%) were female, and their mean age was 73 (range: 27–93) years. For the full analysis dataset (N = 316), the mean number of TURBT tissue samples submitted per patient was 1.3 (range 1–7).

Rate of CIS on re-evaluation

On pathology re-review, CIS was identified in 80 of 316 (25.3%, 95% CI 20.8%–30.4%) cases. The rate of CIS detection on re-review was numerically higher in patients initially diagnosed with T1 NMIBC (19/62, 30.6%) compared with those with Ta NMIBC (61/254, 24.0%); however, the study was not designed to determine differences in CIS rate between disease stages. CIS detection rates on re-review varied depending on which of the eight pathologists had performed the initial review and diagnosis, ranging from 15.0% (6/40 cases) to 66.7% (2/3 cases) (Table 2).

Table 1.

Demographic and clinical characteristics of the analysis population at initial review.

	Analysis population N = 316
Mean age,a years	73
Male,a n (%)	158 (81.0)
Mean tissue samples per patient, n (range)	1.3 (1–7)
Ta disease on initial review, n (%)	254 (80.4)
T1 disease on initial review, n (%)	62 (19.6)

^a

Mean age and proportion of male patients were calculated for the subset of patients with available data on these parameters (n = 190).

Table 2.

CIS positivity rates by initial diagnosing pathologist at re-review.

	Number of patients evaluated during initial review	Number of patients identified with concomitant CIS on re-review	CIS detection rate on re-review (%)
Pathologist 1	95	28	29.5
Pathologist 2	11	3	27.3
Pathologist 3	3	2	66.7
Pathologist 4	12	4	33.3
Pathologist 5	42	15	35.7
Pathologist 6	79	15	19.0
Pathologist 7	34	7	20.6
Pathologist 8	40	6	15.0
TOTAL	316	80	25.3

Discussion

The findings of this study suggest that CIS may be underreported in patients initially diagnosed with high-grade papillary-only NMIBC. Specifically, approximately one-quarter of patients were found to have CIS that was not identified during initial pathology evaluation, a finding that may have directly impacted treatment decisions and patient clinical outcomes. This underscores the broader issue of inter-observer variability in urological pathology reporting ¹ . Consistent with our findings, previous studies have indicated that re-review of TURBT specimens leads to diagnostic revision in approximately one-quarter to one-third of cases^11,12. This is further compounded by intra-observer diagnostic discrepancies: in our study, the re-reviewing pathologist, who had also performed some of the original evaluations, revealed previously unidentified cases of CIS.

Given that bladder cancer treatment decisions are heavily reliant upon pathology findings, any change in diagnostic interpretation can significantly alter patient management ¹³ . Pathologists evaluating TURBT specimens should be aware of the clinical importance of identifying concomitant CIS in patients diagnosed with papillary NMIBC. This is particularly pertinent for general pathologists who may lack specific expertise in uropathology ¹¹ . One of the persistent challenges in bladder pathology is distinguishing true CIS from lateral flat spread or “shoulder lesions” of high-grade papillary urothelial carcinoma^7,14. A recent survey of pathologists and urologists revealed substantial variation in the interpretation of atypical flat urothelial lesions that may exhibit early papillary features ¹⁵ .

These results emphasize the critical need for robust communication between urologists and pathologists to optimize bladder cancer diagnosis, risk stratification, and management. Key elements of this collaboration include high-quality sample collection and documentation, adequate depth of resection to include muscle, precise localization of biopsy sites, and provision of relevant clinical history with clearly stated diagnostic questions for the reviewing pathologist ¹⁶ . To avoid potential diagnostic confusion between CIS and papillary shoulder lesions, it is important that the urologist takes the time to biopsy and label geographically distinct areas of bladder mucosa that are not immediately adjacent to the papillary tumor. Furthermore, it is incumbent upon the urologist to request that the reviewing pathologist specifically rules out the presence of concomitant CIS when returning a diagnosis of papillary urothelial carcinoma. Pathologists in turn can further facilitate communication by using standardized protocols and synoptic reporting templates. A drawback of the electronic health record (EHR) programs used by pathology laboratories is that in some situations there is no option of entering a secondary diagnosis of CIS after a primary diagnosis of papillary urothelial carcinoma, and in other situations the process of entering a secondary diagnosis is cumbersome and time-consuming. Direct dialogue with the urologist may help ensure that the pathologist addresses the specific clinical concerns, rather than relying solely on EHR-based diagnostic codes.

As a real-world study, there are several limitations to acknowledge. Patient demographic and clinical information were incomplete due to variability in data collection across the referring practices. The absence of cytology and other biologically relevant data limited the ability to stratify risk for disease upgrading. Additionally, the pathology re-review was unblinded and conducted by a pathologist who had participated in the initial evaluations, thereby introducing potential bias.

Conclusion

Approximately one-quarter of high-grade Ta/T1 NMIBC specimens may also have undetected CIS. Given that the presence of concomitant CIS can influence therapeutic eligibility, insurance coverage, and treatment outcomes in patients with NMIBC, close collaboration between pathologists and urologists is essential to ensure accurate diagnoses. Review of bladder cancer specimens by those with genitourinary pathology training should be considered when available. Methodical examination of all available bladder tissue fragments is critical to ensure that CIS is accurately identified or excluded, thereby supporting optimal treatment planning and patient outcomes.