Pitfalls of frozen section diagnosis in ureter margin evaluation of plasmacytoid urothelial carcinoma of urinary bladder

Abstract

Background

Plasmacytoid urothelial carcinoma (PUC) is a rare and aggressive subtype that often presents at advanced stages with poor prognosis. This study investigated tumor invasion to better understand tumor behavior and potentially to improve management strategies by comparing the clinicopathologic characteristics of PUC with positive ureter resection margin (+ URM) with PUC with negative URMs (-URM).

Methods

This retrospective analysis used pathology reports from 2017 to 2023 for cases diagnosed with PUC during radical cystectomy (RC). All applicable H&E slides of RC specimens were reviewed. Cases with a plasmacytoid component greater than 25% in the RC specimens were analyzed. Frozen section analyses (FSAs) and permanent section analyses (PSAs) of ureter resection margins were performed.

Results

Fifteen patients with a plasmacytoid component greater than 25% in their RC specimens were identified. Compared with -URM PUC cases, +URM PUC cases were located more frequently at the trigone or bladder neck, and all + URM cases exhibited ureter orifice involvement. Among 6 PSA-positive cases, three (50%) cases showed discrepancies with FSA. Three + URM cases exhibited PUC tumor cells along the submucosa and muscularis propria layer, and the 3 remaining cases showed PUC tumor cells along the adventitia. We observed a consistent adventitia invasion in all the discordant cases, with sectioning errors and misinterpretation identified as the primary causal factors.

Conclusion

To the best of our knowledge, this is the first study to demonstrate two separate patterns of tumor infiltration along the ureter and to discuss the significance of comparing FSA with PSA in PUC. The significance of comprehensive management strategies for PUC patients, including a thorough evaluation of ureteral margins and accurate interpretation of periureteral fat tissue, is highlighted. Large, well-designed studies are needed to strengthen the evidence and to establish optimal management strategies for patients with PUC.

Introduction

Plasmacytoid urothelial carcinoma (PUC) is a rare subtype that accounts for 1–3% of invasive UC [1,2,3]. PUC typically presents as advanced stage disease and is associated with an advanced tumor stage at radical cystectomy (RC), a high rate of lymph node (LN) metastasis, and poor prognosis [4,5,6]. Several studies have identified a notable preference for peritoneal spread and a unique spreading pattern along the fascial plane as the distinguishing characteristics of PUC [5,6,7].

Frozen section analyses (FSAs) on the ureters during RC help to identify malignant tissue and enables additional resection to decrease the recurrence risk [8]. On average, patients with the plasmacytoid subtype undergo 2.3 intraoperative surgical frozen margin resections for each affected ureter before negative margins are obtained, which is twice the number typical in conventional UC [5]. Furthermore, PUC had an 8.6 times higher risk of a positive ureter resection margin (+ URM) in a permanent section analysis (PSA) than was found in conventional UC; even compared with the micropapillary subtype, PUC demonstrated a URM positivity rate about twice as high [5].

In this study, we compare the clinicopathologic characteristics of PUC with + URM with those of PUC with negative URM (-URM). Based on that assessment, we investigated tumor invasion to better understand tumor behavior from a pathological point of view and to explore improved management options for patients with the plasmacytoid subtype.

Materials and methods

Patient selection

This study was approved by the Institutional Review Board at Ewha Womans University Medical Center (2024-03-006). We retrospectively searched our pathology reports from 2017 to 2023 and identified cases diagnosed with PUC during RC. Two genitourinary pathologists (J.M.K and S.P) reviewed all applicable H&E slides of RC specimens. Tumors were graded according to the 2022 World Health Organization Classification of the urinary and male genital tumors [9] and were staged according to the TNM system from the American Joint Committee on Cancer 8th Edition Cancer Staging Manual [10]. Cases with a plasmacytoid component greater than 25% in the RC specimens were included in the current analyses. Patient age, sex, tumor location, pathologic tumor stage, LN status, proportion of plasmacytoid component, neoadjuvant chemotherapy, and clinical outcome were recorded.

FSA and URM assessment

RC was performed following the standard protocol, with routine dissection of pelvic LNs, including the bilateral internal iliac, external iliac, and obturator LNs. The right and left URMs were sent for FSA prior to reconstruction of the ureterointestinal anastomosis for the neobladder. The URMs were snap-frozen, cut at 5 μm, and stained with a rapid H&E protocol. Two genitourinary pathologists performed histological assessments of those specimens to determine the presence of neoplastic changes. The URMs were trimmed, embedded in paraffin, and step-sectioned for the PSAs, which represent the final URMs. Pan cytokeratin (Novacastra, Buffalo Grove, IL) immunostaining was performed as needed to highlight PUC tumor cells.

Statistical analysis

Clinicopathologic variables were compared between groups using the Chi-square and two-sample t tests. Statistical analyses were performed using SPSS^® software version 20 (SPSS Inc., Chicago, IL). A p-value < 0.05 was considered statistically significant.

Results

Clinicopathologic characteristics in the -URM and + URM groups

Fifteen patients with a plasmacytoid component greater than 25% in their RC specimens were identified during the study period. All 15 patients were male and presented at an advanced tumor stage (≥ pT2) during RC. Nine of the 15 (60%) cases showed -URM, and 6 (40%) showed + URM. Notably, the + URM PUC cases were located more frequently at the trigone or bladder neck than the -URM PUC cases (p = 0.005), and all + URM cases exhibited ureter orifice involvement (p = 0.003). The + URM PUC patients displayed a higher proportion of plasmacytoid component, along with increased incidences of LN involvement, distant metastasis, and mortality. However, those parameters did not exhibit significant differences from those of the -URM PUC patients. The clinicopathologic characteristics of the two groups are summarized in Table 1.

Table 1 The clinicopathologic characteristics of the negative ureter resection margin patients and the positive ureter resection margin patients

Concordance between FSA and PSA

URM PUC involvement was confirmed in 6 of the 15 specimens by PSA (40%). The FSA was positive in 3 of the 15 specimens. The FSA was negative in all 9 of the PSA-negative specimens, but 3 (50%) of the 6 PSA-positive cases had FSA-negative results. Upon examining the URM involvement pattern in the 6 + URM cases, 3 cases exhibited PUC tumor cells along the submucosa and muscularis propria layer (Fig. 1a, c), and the remaining 3 cases showed PUC tumor cells along the adventitia (Fig. 1b, d). These patterns extended into the longitudinal section of the ureter in the RC specimens (Fig. 2). Notably, in all 3 of the discordant cases between FSA and PSA, only the adventitia invasion of tumor was observed. Upon investigating the cause of the discordant cases, we found that 2 cases were due to sectioning errors (Fig. 3), and 1 case was due to misinterpretation by the surgical pathologist.

Fig. 1

Two involvement patterns of plasmacytoid urothelial carcinoma in the ureter resection margin. a Tumor cells are exhibited along the submucosa and the muscularis propria layer (arrows) (H&E, x4). b Tumor cells are noted along the adventitia (arrows) (H&E, x4). c Magnified (H&E, x10) of a, pan cytokeratin staining highlights tumor cells, inset (H&E, x40). d Magnified (H&E, x10) of b, pan cytokeratin staining highlights tumor cells, inset (H&E, x40)

Fig. 2

a Illustrated anatomy of the ureterovesical junction. b Plasmacytoid urothelial carcinoma tumor cells extend into the muscularis propria of the ureter in radical cystectomy specimen (arrows) (H&E, x4). c Plasmacytoid urothelial carcinoma tumor cells extend into the adventitia of the ureter in radical cystectomy specimen (arrows) (H&E, x4). d Magnified (H&E, x10) of b, pan cytokeratin staining highlights tumor cells, inset (H&E, x20). e Magnified (H&E, x10) of c, pan cytokeratin staining highlights tumor cells, inset (H&E, x20)

Fig. 3

Representative H&E slides from a discordant case. a Periureteral fat tissue loss is observed in the frozen section of the ureter resection margin (arrows) (H&E, x4). b Tumor cells are noted in periureteral fat tissue in the permanent section of the ureter resection margin (arrows) (H&E, x4). c Magnified (H&E, x10)

Discussion

To the best of our knowledge, we are the first to compare FSA and PSA results from the URM of PUC cases and to demonstrate two separate patterns of tumor infiltration. In this study, we found a notable 50% discrepancy rate between FSA and PSA results for the URM. Moreover, we observed a consistent periureteral fat infiltrative pattern (adventitia invasion) in all the discordant cases, with sectioning errors and misinterpretation identified as the primary causal factors.

Submission of URM for intraoperative pathology consultation is common for bladder cancer patients undergoing RC [11]. However, studies about the significance of FSA of the URM in PUC are rare. Kaimakliotis et al. [5] reported that a 32% URM-positive status in PUC, which aligns closely with our results of a 40% URM positivity rate. In the study of Kaimakliotis et al. [5], the URM positivity rate in PUC was 8.6 times higher than that in conventional UC (2.3%). Furthermore, PUC showed a much higher URM positivity rate than that (17.9%) in micropapillary UC (MPC), another aggressive subtype [5]. Notably, all + URM in conventional UC and MPC involved only carcinoma in situ (CIS) on the endoluminal URM, whereas 5 of the 8 + URM cases in PUC showed tumor cell invasion without concurrent CIS [5]. In accordance with the literature, all + URM cases in our cohort exhibited only invasion of PUC tumor cells. These findings are consistent with previous studies indicating that PUC has a propensity to manifest disease on peritoneal surfaces, leading to common + URM and surgical upstaging [6, 7]. These results suggest a significant likelihood that residual tumor will remain at the operative site even after radical surgery. These observations are associated with the independent effect of plasmacytoid subtype histology on survival outcomes. It remains unknown whether + URM in PUC is independently associated with disease-specific or overall survival, but our findings in this study and those in the literature imply that a thorough assessment of intraoperative URMs is essential in PUC.

We found that + URM cases in PUC showed 2 distinct patterns of tumor infiltration. In our cohort, PUC tumor cells infiltrate through the submucosa and muscularis propria layer of the distal ureter, or tumor cells were noted along the adventitia over a longitudinal section of the ureter in the RC specimen. Kaimakliotis et al. [5] also reported that all + URM involving invasive tumor cells in PUC showed an infiltrative pattern of the subserosal and adventitia ureteral plane, and PUC tumor cells were also observed along the subserosal surface and ureteral adventitia on multiple, sequential surgical excisions. These findings explain that PUC tumor cells that penetrate beyond the detrusor muscle could potentially move along the bladder serosa beneath the peritoneal reflection or the endopelvic fascia. Once within Waldeyer’s space, they can be tracked along the distal ureteral subserosa and adventitia [5]. On the other hand, the mechanism of intramuscular invasion was elucidated by the study of Beunk et al. [12]. They explained that smooth muscle tumor invasion in the bladder involves multiple steps, beginning with epithelial cancer cells infiltrating the connective tissue that separates the epithelial and muscle layers, followed by muscle invasion. Here, bladder tumor cells exhibit chain-like invasion into smooth muscle tissue, showing characteristics of jammed or collective migration aligned with the linear endomysium [12]. They also reported that muscle invasion in cancer patients is frequently correlated with destructive growth and a worse prognosis [12]. Anatomically, the inner smooth muscle layer of the ureter merges with that of the contralateral ureter to form the superficial trigone [13]. The distal ureter is accompanied by a separate external smooth muscle layer (Waldeyer sheath) to the bladder [13]. As it traverses the bladder wall, the Waldeyer sheath merges with detrusor muscle fibers to form the deep trigone of the bladder, which is enclosed by the 2 ureter orifices [13]. These distinctive anatomical structures between the ureter and bladder can provide additional support for the 2 PUC tumor cell infiltration patterns highlighted in this study, as well as the observation that these 2 infiltration patterns align with the longitudinal ureter muscle. These explanations are in line with the findings in our cohort that PUC with + URM was significantly more common in the trigone or bladder neck, and all PUC cases with + URM showed tumor involvement in the ureteral orifice, tended to exhibit higher tumor and LN stages, and showed slightly more progressive behavior. The aggressiveness of PUC can be explained by the loss of E-cadherin. Several studies have demonstrated that decreased expression of E-cadherin, a membrane connective tissue protein, is associated with more aggressive pathology, loss of cell differentiation, increased cell invasion, and poor prognosis in UC [14,15,16]. Loss of membranous E-cadherin expression and its nuclear accumulation are highly associated with PUC [14, 15] and could allow PUC tumor cells to easily invade surrounding tissues, ultimately leading to an advanced stage and rapid progression.

In this study, discordance between FSA and PSA was observed in 3 of the 6 (50%) + URM cases in the histopathological examination of the permanent URM tissues. Interestingly, all the discordant cases exhibited the periureteral fat infiltrative pattern (adventitia invasion), and the causes of the discordance between FSA and PSA were sectioning errors and misinterpretation by the pathologist of tumor cells as inflammatory cells. These findings indicate that a thorough evaluation of the ureteral margin should be conducted, with a focus on the periureteral fat tissue. The freezing and thawing processes used for frozen sectioning can cause tissue damage, with fatty tissues being especially vulnerable, potentially resulting in tissue loss and sectioning challenges. When preparing frozen sections, it is important to include periureteral fat tissue in the embedding process. Additionally, if the periureteral fat tissue is not visible on the slide or in any suspicious cases, additional frozen section examinations should be conducted. Pathologists should also pay close attention to the periureteral fat tissue when interpreting frozen slides. Ultimately, this thorough assessment of intraoperative URM should be applicable to all UCs that include PUC. In addition to this, it is recommended that pathologists include detailed information on PUC in every pathological report, as lack of awareness regarding the plasmacytoid subtype during FSA appears to pose greater challenges. Given that PUC tumor cells show morphological resemblance to plasma cells, it can be suggested that including detailed information about PUC in pathological reports may assist pathologists in minimizing the risk of misinterpretation during FSA.

This study has several limitations. It had a retrospective design that could not avoid inevitable limitations such as selection bias. The small sample size caused by the rarity of this entity and our focus on PUC with a plasmacytoid subtype component greater than 25%, and inclusion of patients who received neoadjuvant chemotherapy could affect the overall data quality. Also, statistically significant differences were not found in recurrence and survival between + URM patients and -URM patients, although + URM patients showed slightly more progressive behavior in the present study. Larger cohorts and further well-designed multicenter studies are needed to clarify the clinical significance of frozen assessment in PUC and to establish strong recommendations. Despite those shortcomings, this study offers valuable insights based on currently available data. These findings can aid clinicians and pathologists in shared decision-making discussions about the management and prognosis of patients with PUC.

Conclusion

To the best of our knowledge, this is the first study to demonstrate 2 separate patterns of tumor infiltration along the ureter and to discuss the significance of FSA compared with PSA in PUC. Based on an understanding of the two unique infiltration patterns of PUC, we emphasize the importance of thoroughly embedding and sectioning periureteral fat tissue during the frozen sectioning process and accurately evaluating periureteral fat tissue during frozen slide interpretation. Large prospective studies are warranted to clarify the importance of a comprehensive management strategy in patients with PUC.