Clinicopathologic analysis of patients undergoing repeat transurethral resection of bladder tumour following an initial diagnosis of urothelial carcinoma with lamina propria invasion and variant/divergent histology

Introduction

Urothelial carcinoma (UCa) of the bladder is the sixth most common malignancy in the USA and is associated with significant morbidity and mortality.1 Depth of invasion on bladder biopsy/transurethral resection (TUR) is one of the most important prognostic indicators, and a subset of patients diagnosed with UCa with lamina propria (LP) invasion (pT1) is at increased risk for disease recurrence and progression.2–4 The ability to identify patients with UCa with LP invasion who are at risk for recurrence and progression based on clinicopathologic features thus holds prognostic significance. UCa with variant/divergent histology occur in 10%–33% of cases and are reported to have worse clinicopathologic outcomes, including advanced tumour stage, extravesical disease, lymph node metastasis and positive margins.4–6 Other studies, however, have failed to find significant prognostic differences between variant/divergent histologies and conventional UCa.7–11 Guidelines regarding clinical management for pT1 UCa with variant/divergent histology therefore remain broad, making patients with pT1 UCa with variant/divergent histology particularly challenging to manage. The clinicopathologic assessment of the impact of pT1 UCa cases exclusively focusing on those with variant/divergent histology has not been well characterised in the pathology literature. Here, we describe the clinicopathologic characteristics of 95 patients diagnosed with pT1 UCa on biopsy/TUR with at least one variant/divergent histology identified.

Materials and methods

A search was made through our urologic pathology files and expert consult cases of the senior author for biopsy/TUR cases with a diagnosis of UCa with LP invasion and variant/divergent histology from 2014 to 2020. Regarding variant/divergent histologies, we did a search for cases that had the terms squamous, glandular (including enteric, intestinal, villoglandular and mucinous), micropapillary, clear cell/lipid rich, diffuse/signet ring/plasmacytoid, nested, sarcomatoid, poorly differentiated/anaplastic, small cell, lymphoepithelioma-like, osteoclast-like giant cells and tumour giant cells. Cases that did not have true glandular differentiation with just gland-like lumina were excluded. Cases with two or more variant/divergent histologies identified on initial biopsy were included in the analysis of each separate variant/divergent subtype and the ‘multiple’ category. Patients with a prior diagnosis of UCa with muscularis propria (MP) invasion or upper tract UCa were excluded. Clinicopathologic characteristics were summarised and compared using z-score test for two populations with statistical testing performed at a 0.05 significance level.

Results

Ninety-five patients with at least one biopsy/TUR of UCa with LP invasion and variant/divergent histology were identified (table 1). Sixty patients had biopsies/TURs performed at our institution and 35 had biopsies/TURs performed at outside hospitals with slides reviewed in consultation. Three (3.1%) patients had prior reports of variant/divergent histology which were diagnosed by outside pathologists with slides unavailable for review at our institution. There were 66 (69.5%) males and 29 (30.5%) females with a mean age of 72 years (range: 46–92 years). Initial variant/divergent histologies identified were: glandular (35.8%), squamous (23.2%), micropapillary (20%), clear cell/lipid rich (12.6%), diffuse/signet ring/plasmacytoid (10.5%), nested (9.5%), sarcomatoid (6.3%), poorly differentiated/anaplastic (4.2%), small cell (2.1%), lymphoepithelioma-like (2.1%), osteoclast-like giant cells (1.1%) and tumour giant cells (1.1%), figure 1A–D. Two or more variant/divergent histologies were identified in 18.9% of these cases.

Table 1

Histopathologic features and urothelial carcinoma persistence rates on repeat biopsy/transurethral resection (TUR) following variant/divergent histology identification.

Figure 1Figure 1Figure 1

Urothelial carcinoma (UCa) with variant/divergent histologies. (A) UCa with clear cell features (H&E, 10× magnification). (B) Plasmacytoid UCa (H&E, 20× magnification). (C) UCa with divergent (glandular) differentiation (H&E, 10× magnification). (D) Micropapillary UCa (H&E, 10× magnification).

Twenty-one (22.1%) patients had a prior diagnosis of UCa before a variant/divergent histology was identified. Stage of UCa in the patients was categorised as follows: 5 with non-invasive high-grade UCa (pTa), 3 with UCa in situ (CIS) only and 13 with pT1 UCa. The new variant/divergent histologies identified in patients with prior UCa were: glandular (37.5%), squamous (25%), micropapillary (12.5%), plasmacytoid (12.5%), nested (12.5%), clear cell (8.3%) small cell (4.2%) and poorly differentiated (4.2%).

Forty-nine patients underwent at least one repeat biopsy/TUR, with median time to repeat biopsy/TUR 9 weeks after initial variant identification (range: 2–121 weeks). In these patients, the rate of disease persistence was 75.5% and categorised as follows: low-grade pTa (4.1%), high-grade pTa (8.2%), pTis (8.2%), high-grade pT1 (49%) and high-grade pT2 (6.1%) (table 1). Variant/divergent histology was identified in 40.5% of biopsies/TURs with recurrence and occurred only in cases with invasive carcinoma UCa (pT1 or pT2). When present on a repeat biopsy/TUR, variant/divergent histology concordance with the original diagnosis was high at 93.3%.

Sixty-three patients with available clinical history had no prior diagnosis of pT1 UCa. In these cases, 44.4% of tumours were unifocal, 42.9% of tumours were multifocal and 12.7% of tumours were non-specified (table 2). The rate of micropapillary variant was significantly higher in multifocal tumours compared with unifocal tumours (37% vs 7.1%; p=0.007). In contrast, the rate of cases with squamous differentiation was higher in unifocal tumours compared with multifocal tumours (32.1% vs 7.4%; p=0.02).

Table 2

Tumour focality on first pT1 biopsy/transurethral resection

Carcinoma was present in a diverticulum of three patients. One patient was managed with surveillance and Bacillus Calmette Guerin (BCG). The other two patients underwent partial cystectomies, with one having pT1 high-grade UCa and the other pT3 high-grade UCa on final pathology.

Treatment information following the initial identification of variant/divergent histology in patients with clinical follow-up is summarised in table 3. Treatment strategies included: intravesical therapy with surveillance cystoscopy, radiation and systemic chemotherapy and upfront cystectomy. Eight patients (12.7%) elected to not pursue additional urologic therapy following diagnosis. Notably, following standard staging work-up with cross axial imaging at the time of biopsy, 12 patients (19%) had clinically worse disease and were referred for appropriate therapy. Of the 26 patients initially treated conservatively with surveillance cystoscopy and intravesical therapy, 17 (65.4%) experienced recurrence and 6 (23.1%) progressed to at least T2 disease. Average time from initial identification of a variant/divergent histology to recurrence was 60 weeks. The variant/divergent histology in patients with pathologic progression was: one glandular, one squamous, one poorly differentiated, one plasmacytoid, one glandular/plasmacytoid/micropapillary and one squamous/clear cell.

Table 3

Frequency of treatment modalities following variant/divergent histology identification

Overall, 20 patients underwent cystectomy (16 radical and 4 partial) with lymph node dissection performed in 18 patients (90%). Three patients (15%) underwent neoadjuvant chemotherapy prior to resection. Average time from variant identification to cystectomy was 30.3 weeks (range: 2–142 weeks). Only one patient had a diagnosis of pT2 disease prior to cystectomy. Sixteen patients had UCa identified at resection (80%) categorised as follows: two pTis (10%), eight pT1 (40%), two pT2 (10%), three pT3 (15%) and one pT4 (5%) (table 4). Of these 16 patients, 7 had a component of variant/divergent histology identified (43.8%), which when present, had a high concordance with the original variant/divergent histology identified at biopsy (85.7%).

Table 4

Histopathologic features of patients who had cystectomy

Discussion

Depth of invasion of UCa on bladder biopsy/TUR is one of the most important prognostic factors, and a subset of patients diagnosed with UCa with LP invasion (pT1) is at increased risk for disease recurrence and progression. This current study is the first to our knowledge to specifically analyse the clinicopathologic features of patients diagnosed with pT1 UCa and variant/divergent histology in the pathology literature.

UCa with LP invasion has a reported total recurrence rate of 44.5% and progression rates of 17.1%.12 Even when the original tumour is completely resected, residual disease occurs in 33.8% of patients with progression rates of 4%–11.8%.13 Due to this clinical risk, early repeat biopsy/TUR and continued surveillance are recommended for patients following a diagnosis of pT1 UCa.2 3 The ability to identify the patients at increased risk for recurrence and progression thus carries prognostic importance. Variant/divergent histology has been recognised as an important factor associated with higher rates of recurrence and worse clinical outcomes.4 14–19 The presence of micropapillary histology, for instance, is reported to have a direct association with pathologic grade, supporting a biologic basis for a more aggressive clinical behaviour.17 The recurrence rate in our cohort was similarly high, with 75% of patients who underwent a subsequent biopsy/TUR having persistent/recurrent UCa. Half of these patients had persistence of pT1 disease and 6.1% of patients were upstaged to pT2. Intravesical therapy had minimal impact on recurrence rates as 65.4% of patients experienced at least one recurrence after completing at least one induction treatment course. Even more concerning, almost a quarter of patients who received a course of intravesical therapy still progressed to pT2 disease, confirming variant histology as a high-risk feature in patients with pT1 disease. Interestingly, variant/divergent histology was identified in less than half of repeat biopsies/TURs with recurrence and only in patients with invasive disease. Possible explanations for the decreased presence of variant/divergent histology on follow-up include the known heterogeneity inherent in UCa as well as the possible role of intravesical therapy. Further work on the characterisation of the morphologic impact intravesical therapy can have on variant/divergent histology and detection will be important.

Almost a quarter of patients had an initial diagnosis of UCa prior to the first identification of a variant/divergent histology, with 13.7% of these patients having a prior diagnosis of conventional UCa with pT1 disease. This highlights the importance of continued surveillance in patients with UCa, especially for patients with pT1 disease as detection of variant/divergent histologies may change clinical treatment options and strategies. Interestingly, in patients without a prior diagnosis of pT1 disease, increased rate of micropapillary histology was observed in multifocal tumours, while squamous histology was increased in unifocal tumours. The consequence of adequate tumour sampling at the time of biopsy/TUR cannot be overemphasised, especially when multiple lesions are present.

Radical cystectomy is indicated in selected high-risk patients with non-muscle invasive disease, with most frequent reasons being BCG refractory disease and large tumour burden.2 20 Cystectomy has additionally been advocated for patients with variant/divergent histology given reports of increased extravesical disease and lymph node metastasis.21 Indeed, only 1 of the 20 patients in our cohort had pT2 disease prior to cystectomy and 13 of the 20 patients proceeded directly to cystectomy without any intervening period of intravesical therapy and surveillance. The clinical significance of variant/divergent histology and early cystectomy remains uncertain as conflicting results regarding survival outcomes persist.7 9 16 21–23 The sensitivity of TUR of bladder tumour (TURBT) to detect variant histology is reported to be 39%, while the sensitivity of cystectomy to detect previously noted variant histology is reported to be 49%, though this is not well established.24 25 Of the 20 patients in this study who underwent cystectomy following a diagnosis of variant/divergent histology on biopsy/TUR, only 35% had residual variant/divergent histology identified on cystectomy. When only accounting for patients with UCa detected at cystectomy, the sensitivity of cystectomy to detect a previously diagnosed variant/divergent histology was 43.8% in our cohort. Interestingly, though concordance was high when a variant/divergent histology was identified at cystectomy, one patient with discordant variant histology not only had plasmacytoid variant of UCa present on both original biopsy/TUR and repeat biopsy/TUR but also had nested variant of UCa identified in the cystectomy specimen. Determining the clinical impact of having variable and/or persistent variant/divergent histology at the time of cystectomy is therefore a potential area for further investigation and risk stratification.

Partial cystectomy without lymph node dissection was performed on two patients with diverticular disease. One patient with glandular differentiation had pT1 disease on final pathology and one patient with squamous differentiation had pT3 disease on final pathology. Further characterisation of patients presenting with diverticular UCa with variant histology will be necessary to determine the optimal clinical management for these unique cases. Lymph node metastasis was identified in one patient at the time of cystectomy and occurred in association with micropapillary histology with final pathologic stage categorised as pT1N2.

Limitations of our study include the retrospective, single-institutional design of the analysis and the limited sample size and variable clinical follow-up.

In conclusion, the majority of patients in our single-institutional cohort with pT1 disease and variant/divergent histology detected on TUR had persistent UCa. Our findings support the fact that variant/divergent histology is a critical risk factor for UCa recurrence and highlight the critical role of repeat bladder sampling, especially in patients who initially had variant/divergent histology, even in the absence of MP invasion.

Take home messages

The subset of patients with urothelial carcinoma (UCa) of the bladder and lamina propria (LP) invasion in biopsies/transurethral resections (TURs) who are at significant risk for recurrence and progression to UCa with muscularis propria (MP) invasion have not been well characterised.

We identified 95 patients with at least one biopsy/TUR of UCa with LP invasion and variant/divergent histology.

The rate of micropapillary UCa was significantly higher in multifocal tumours compared with unifocal tumours.

This study highlights the critical role of repeat biopsy/TUR especially in patients who have variant/divergent histology, even in the absence of MP invasion.

Data availability statement

All data relevant to the study are included in the article. All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statementsPatient consent for publicationEthics approval

This is not a clinical trial. This study was completed following the guidelines of and with approval from the authors’ institutional review board.

留言 (0)

沒有登入
gif