Introduction
For better communication between pathologists and urologists, the first edition of The Paris System for Reporting Urinary Cytology (TPSRUC, The Paris System [TPS] 1.0) was published in 2016 to standardize the reporting system.1 After 5–6 years of practice based on the criteria of TPS 1.0 and prospective studies, the second edition (TPS 2.0) was published in 2022.2,3 TPS highlights the need to focus on accurately identifying high-grade urothelial carcinoma (HGUC). After adopting TPS, there was an increase in false-negative rates and risk of malignancy (ROM) for negative for high-grade urothelial carcinoma (NHGUC).4 Although the specificity and negative predictive value of TPS remains high (98.0–98.6% and 68.0–94.8%, respectively), the sensitivity and positive predictive value remain low (37.5–46.0% and 70.6–96.0%, respectively).5,6
TPS 2.0 comprises of six diagnostic categories: non-diagnostic, NHGUC, atypical urothelial cells (AUC), suspicious for high-grade urothelial carcinoma (SHGUC), HGUC, and other malignancies. Low-grade urothelial neoplasm (LGUN) was moved under the NHGUC category. An approach to diagnosis in urinary cytology was outlined in TPS 2.0 (Fig. 1).2 Furthermore, three chapters were added in TPS 2.0: cytology of the upper urinary tract, risk of high-grade malignancies (ROHM), and the history of urine cytology: the long and winding road to Paris 2.0. The main goal of TPS was to lower the rates for indeterminate categories, particularly “atypia”, and have the highest specificity and sensitivity for HGUC. In each chapter and diagnosis criteria, the further research direction was detailed, followed by sign-out samples, including formats and notes, making the clinical practice of urine cytology more uniform and efficient. Furthermore, TPS 2.0 further illustrates how to define the N/C ratio, which is almost the critical diagnostic criteria, and reduces the confusion after introducing TPS 1.0. The present study summarizes the updates for TPS 2.0, and the corresponding clinical practice.
Pathogenesis of urothelial carcinoma
Molecular pathways of the neoplastic transformation of urothelium
In TPS 2.0, the Cancer Genome Atlas was incorporated into the pathogenesis of low-grade (LG) and HGUC. A number of genomic alterations were identified in the pathogenesis of urothelial carcinoma (UC).7–9FGFR and RAS alterations were frequently identified in low-grade papillary carcinoma (LGPUC), FGFR and TP53 were frequently identified in HGUC,10 and TP53 and RB were frequently identified in carcinoma in situ (CIS). However, few genomic alterations were frequently identified in muscle invasive HGUC, such as the RTK/RAS/PI3K pathway (72%), FGFR3 alterations, ERBB2 enrichment, the histone modification pathway (89%), the SWI/SNF complex pathway (64%), and the TP53/RB pathway (93%).11
Adequacy of urine specimens
Adequacy criteria
Any urine specimen with abnormalities (atypical, suspicious, and positive for malignancy);
Voided urine: >25–30 ml (25 ml with ThinPrep and 30 ml with SurePath preparation);12
Instrumented urine urothelial cells: >20 cells/10 high-power fields (HPFs), satisfactory; 10–20 cells/10 HPFs, satisfactory but limited by low cellularity; <10 cells/10 HPFs, unsatisfactory/non-diagnostic.13
Unsatisfactory cases account for 0.3% of voided urine specimens.6
The less-than-optimal adequacy category14
Urine specimens that meet every adequacy criterion, except for urothelial cellularity, were assigned as “less-than-optimal” adequacy. It remains uncertain to assign such cases as adequate or non-diagnostic/unsatisfactory. More studies are needed.
Negative for high-grade urothelial carcinoma
A urinary specimen can be possibly categorized as NHGUC when it is adequate and lacks any cytomorphologic findings for HGUC. NHGUC cases accounted for 90.5% of voided urine specimens.6
Diagnostic criteria for NHGUC15,16
Benign urothelial cells
Benign urothelial cells were classified into superficial (umbrella), intermediate (“parabasal-like”), and deep (basal) urothelial cells. Superficial urothelial cells are large cells with one or multiple, large, centrally located round-to-oval nuclei, smooth nuclear membranes, abundant cytoplasm, cytoplasmic vacuoles for some cells, and low nuclear/cytoplasmic (N/C) ratios (<0.5) (Fig. 2a, b). Intermediate urothelial cells are intermediate in size, with round-oval nuclei, smooth nuclear membranes, and bland chromatin, but with less cytoplasm than superficial urothelial cells, and low N/C ratios (<0.5) (Fig. 2a). Deep urothelial cells are small cells with round nuclei, containing evenly distributed fine granular chromatin, scant cytoplasm, and high N/C ratios (Fig. 2b). Urothelial cells may present as single cells, in nests/sheets, or evenly in papillary-like fragments (especially in instrumented urinary specimens), without fibrovascular cores and cytologic atypia (Fig. 2c).
Squamous epithelial cells
Squamous cells are possibly from the urethra, squamous metaplasia, or contamination. The presence of benign squamous cells, including its parakeratotic or anucleate forms, is not specifically required for reporting (Fig. 2d).
Glandular cells
Glandular cells may be observed in urinary specimens, such as renal tubule cells observed in renal tubule casts (Fig. 3a), urachal remnants, endometriosis (Fig. 3b), cystitis cystica/glandularis (Fig. 3c), intestinal metaplasia, and prostatic cells with pigments (Fig. 3d). These would present as single cells or nests of columnar cells with small nuclei, and a vacuolated or granular cytoplasm.
Degenerative changes
The nuclei of degenerative cells are shrunken (condensed) and dark, with mildly irregular nuclear membranes. The cytoplasm may contain granules, and possibly large eosinophilic and/or cyanophilic intracytoplasmic inclusions (Melamed-Wolinska bodies, Fig. 2e). N/C ratios are low (<0.5).
Urothelial tissue fragments
Benign-appearing vs. atypical urothelial tissue fragments were clarified in TPS 2.0. Cohesive groups of urothelial cells arranged in honeycomb, without cytologic atypia, for HGUC were named as benign urothelial tissue fragments (BUTFs, Fig. 2f). Some mild atypia was allowed for this category. BUTFs may be observed in voided urine, and are always present in instrumented urinary specimens. Any cytological atypia observed from tissue fragments should be classified as atypical urothelial tissue fragments (AUTFs) under the AUC category (Fig. 2g).
Low-grade urothelial neoplasia
In TPS 2.0, LGUN was re-categorized under NHGUC. The diagnosis of LGUN was rendered only when the tissue fragments contained innocuous cells and fibrovascular cores (Fig. 2h). An explanatory note may be needed. The differential of LGUN includes urothelial papilloma, urothelial proliferation of unknown malignant potential (UPUMP), papillary urothelial neoplasm of low malignant potential (PUNLMP), and LGPUC.
Urolithiasis (“Stone atypia”)
Urolithiasis can induce benign urothelial cells to form sheets or three-dimensional clusters. These cells may present with reactive changes, possibly mild hyperchromasia, low N/C ratios, and smooth nuclear membranes (Fig. 2i).
Urothelial change characteristics of infectious processes
Bacterial infection is the most common urinary tract infection. Acute inflammation (abundant neutrophils), bacteria, fungi, and reactive changes can be observed in acute bacterial infection. Viral cytopathic effects may be observed in infections with polyomavirus (Fig. 4a),17–19 herpes simplex virus (HSV, usually type II, but also type I), cytomegalovirus (CMV), and human papillomavirus (HPV) (Fig. 4b). Among these, polyomavirus is the most frequent. Parasites are extremely rare in the United States.
Urothelial changes associated with treatment effects
Radiation causes cytomegaly, nucleomegaly, multinucleation, nuclear and cytoplasmic vacuoles, finely granular/smudgy chromatin, and cytoplasmic polychromasia. N/C ratios are preserved and low (Fig. 4c). Bacillus Calmette-Guerin (BCG) causes granulomatous inflammation (Fig. 4d, e).20 Chemotherapy with mitomycin and thiotepa causes nuclear enlargement, multinucleation, and hyperchromasia of superficial cells (Fig. 4f).21 Seminal vesicle cells may be observed in urine specimens (Fig. 3d), especially after prostatic massage and digital rectal examination, and often together with the presence of spermatozoa. Furthermore, seminal vesicle cells have hyperchromatic, degenerated and smudgy nuclei, and golden-brown lipofuscin pigments in the cytoplasm.22 Abnormal DNA content can also be present.23
Bladder diversion specimens
After cystectomy, the ileal conduit, Indiana pouch, or neobladder is used to replace the bladder to store urine. Urine specimens mainly contain degenerated glandular cells in a background of necrotic debris, mucin, bacteria, and inflammatory cells.24 Urothelial cells and squamous cells may be present.
Ancillary testing
UroVysion may be helpful in the diagnosis.
Category performance2
The NHGUC category has a ROHM of 8.7–36.7%, a false-negative rate of 3.3%, and a negative predictive value (NPV) of 96.7%.
Atypical urothelial cells
A specimen that contains urothelial cells with mild-to-moderate cytologic (not architectural) changes for HGUC is categorized as AUC.25 AUC cases accounted for 5.6% of voided urine specimens.6
Diagnostic criteria for AUC
The diagnostic criteria for AUC includes the major criterion and one minor criterion, which applies for all urine specimens, regardless of the preparation method (Fig. 5a).
Major criterion: Increase in N/C ratio of ≥0.5, but <0.7, due to nuclear enlargement;26,27
Minor criteria:28
Nuclear hyperchromasia;
Irregular nuclear membranes (chromatinic rim or nuclear contour);
Irregular, coarse and clumped chromatin.29,30
Clarification of issues unresolved by TPS 1.0 regarding atypia
Urothelial degeneration presents with irregular cytoplasmic borders, cytoplasmic vacuolization, clumped, hazy, smudged, or indistinct chromatin, interrupted chromatin rim, or variable thickness and irregular contours, but rarely possesses sharp angles that are often observed in malignancies. The AUC category includes specimens, in which due to poor preservation and degeneration, the nature and degree of change in urothelial cells cannot be well-characterized, and there are concerns on HGUC.
Atypical squamous cells (ASCs) might be observed in urine specimens, in which cytomorphology would fall short of the definitive diagnosis of squamous cell carcinoma. It appears to be inappropriate to categorize ASCs in “atypical urothelial cells”.
With the N/C ratio as the criterion, the maintenance of the TPS criteria for upper tract samples and all preparation types were discussed.31
Ancillary testing
FISH/UroVysion may be helpful in the diagnosis.
Suspicious for high-grade urothelial carcinoma
This diagnosis is restrictively used for cases of abnormal urothelial cells that quantitatively fall short of the definitive diagnosis of HGUC.32,33 SHGUC cases accounted for 1.6% of voided urine specimens.6
Diagnostic criteria for SHGUC2
The diagnostic criteria for SHGUC includes the major criteria, and two of three features (Fig. 5b). The same criteria applies for all types of specimens.
A cut-off range of at least 5–10 abnormal cells for lower tract (LT) specimens and ≥10 cells for upper tract (UT) specimens is recommended for HGUC.34
Major criteria: Increase in N/C ratio to ≥0.7 due to nuclear enlargement;
Two of these three features:
Moderate-to-severe hyperchromasia;
Irregular clumpy chromatin;
Irregular nuclear membrane.
Clinical management recommendation.35,36
Patients with a diagnosis of SHGUC should be investigated, in order to determine the presence of HGUC, and manage this based on the patient’s history, clinical setting, and cytologic findings.
High-grade urothelial carcinoma
The primary purpose of TPS is to have the highest positive predictive value for HGUC. HGUC cases accounted for 1.9% of voided urine specimens.6
Diagnostic criteria for HGUC2
At least 5–10 malignant cells are required for the diagnosis of HGUC for LT specimens, and ≥10 cells are required for UT specimens.
The N/C ratio of ≥0.7 due to nuclear enlargement is the most restrictive and recommended benchmark for diagnosing HGUC (Fig. 5c).37 However, a spectrum of N/C ratios was noted for HGUC, which ranged from <0.5 to >0.7;
Moderate-to-severe hyperchromasia (Fig. 5c);
Irregular nuclear membranes (Fig. 5c);
Coarse/clumped chromatin (Fig. 5c);
Additional cytomorphologic features:
Cellular pleomorphism, variation in size and shape, such as round, oval, elongated, and spindle;
Dense or vacuolated cytoplasm;
Prominent nucleoli (Fig. 5d);
Mitoses;
Necrosis (Fig. 5d, f);
Inflammation (Fig. 5f);
Hypochromasia (Fig. 5d).
Variances in the cytomorphology of HGUC38
N/C ratio
HGUC cells present with a wide range of N/C ratios, which range from <0.5 to >0.7. There are sufficient cells with N/C ratios of ≥0.7 and other cytologic features to interpret with confidence (Fig. 5e).
Hypochromasia in HGUC is rare, and a potential pitfall when diagnosing HGUC. Except for hypochromasia, these tumor cells should meet other criteria for HGUC (Fig. 5d).39,40
Degenerative changes
Degenerative changes are commonly identified in voided urine specimens, and these can be observed in HGUC. The features of these degenerative changes include loss of cytoplasm, “blown up” nuclei resulting in increased N/C ratio, irregular nuclear membrane due to dehydration, and dense nuclei due to pyknosis (Fig. 5f). Degenerative cells should not be evaluated for diagnosis. An extremely dark chromatin (a combination of hyperchromasia and coarse/clumped chromatin) is a well-recognized feature of HGUC, especially in degenerated HGUC cells. These cells are an independent predictor of malignancy, and should be counted as diagnosis criteria for HGUC (Fig. 5g).
Important histologic variants of HGUC in urine cytology
A total of 14 histologic variants of HGUC were recognized in the 2022 World Health Organization (WHO) classification of tumors of the urinary system.41 It remains extremely challenging to diagnose some HGUC variants by urine cytology. The TPS 2.0 provides some images,and summarizes the cytomorphologic features of some subtypes.
Cytologic features of some variants
HGUC with squamous differentiation
In addition to HGUC cells, keratinized and/or non-keratinized (with intercellular bridges) squamous tumor cells would be present (Fig. 6a).42
HGUC with glandular differentiation
In addition to HGUC cells, malignant columnar glandular cells would be present, and present in a glandular formation. These tumor cells would have a vacuolated or mucin-filled cytoplasm, and hyperchromatic nuclei that contains coarse chromatin and nuclear membrane irregularities (Fig. 6b). The diagnosis of adenocarcinoma of the urinary tract can only be determined by extensive examination through biopsy or cystectomy tissue. If no definitive HGUC cells are identified in a urine specimen, it may be better to sign out as carcinoma with glandular differentiation, deferring the biopsy or cystectomy, and adding a note to provide the differentials.
The nested variant of urothelial carcinoma
These tumor cells present in nests/three-dimensional clusters. Furthermore, these tumor cells would have medium-sized, round-to-polygonal shapes, present with slightly increased N/C ratios, have HUGC nuclear features, and have a moderate-to-abundant granular-to-dense cytoplasm with distinct cell borders (Fig. 5d). Since its cytology features may overlap with reactive changes and other HGUC subtypes, it is nearly impossible to make a definitive diagnosis of nested variant on urinary cytology samples.
Micropapillary urothelial carcinoma
These tumor cells present in a micropapillary architecture without fibrovascular cores, but with peripheral or eccentric hyperchromatic nuclei, and cytoplasmic vacuoles. Furthermore, the N/C ratios would be elevated (Fig. 6c).43 Other cytomorphologic features would overlap with HUGC.
Plasmacytoid urothelial carcinoma
These tumor cells are present as single cells, and have eccentrically placed, enlarged, hyperchromatic nuclei with irregular nuclear membranes, coarse chromatin and inconspicuous nucleoli, and an abundant and thick cytoplasm (Fig. 6d). The N/C ratio may be less than 0.7.44
Fibrovascular cores
Fibrovascular cores observed in HGUC are considered for the first time in TPS 2.0. Papillary tissue fragments can be observed in papilloma, LGPUC and HGPUC (Fig. 5h). Papillary-like clusters can also be observed in NHGUC, in both voided and instrumented urine (much more common). The diagnostic significance of papillary or papillary-like clusters entirely depends on the presence or absence of cytologic features for HGUC.
Effects of cytopreparation on the cytomorphology of HGUC
Alcohol fixation used in preparation clumps cells to form three-dimensional clusters, mimicking the papillary structure. Furthermore, alcohol fixation shrinks cells, and increases the N/C ratio. This allows the nuclei of alcohol-fixed, Papanicolaou-stained cells to be better preserved and visualized for evaluatingnuclear features that are critical for the diagnosis. Regardless of what kind of specimen preparation technique is used, the TPS criteria can be applied confidently.45
Cytopathology of the upper urinary tract
Definition of upper tract urothelial carcinoma (UTUC)
UTUC includes upper tract HGUC (UTHGUC), CIS, and LGUC (UTLGUC) arising from the ureter or renal pelvis. TPS 2.0 focuses on detecting UTHGUC (CIS and HGUC).
UTUC prevalence
Synchronous UTUC and urothelial carcinomas of the bladder (UCB) can be observed in 17% of cases. Patients with primary UTUC develop recurrences in the bladder (22–47% of cases), which is likely due to the cancerization effect or intravesicular seeding, multifocal lesions, and the contralateral upper urinary tract (UUT) (2–6% of cases). Genomic mutations in FGFR3, KDM6A, and CCND1 are reported to be significantly associated with higher risk of subsequent UCB, and TP53 mutations are associated with lower risk.
Criteria for upper tract urothelial carcinomas2
The TPS cytomorphological criteria for UTHGUC are identical to those for LT specimens. For voided urine specimens, the origin (UT vs. LT) of the HGUC cannot be determined. For instrumented specimens, although the TPS cytomorphologic criteria for HGUC has been applied, a study reported that the cytologic features concerning HGUC included increased N/C ratios, hyperchromasia, and irregular nuclear membranes, as well as pleomorphism, mitoses and atypical degeneration. However, the difference was not statistically different between UTHGUC and UTLGUC. Coarse chromatin and bizarre single cells can be more frequently observed in UTHGUC when compared to UTLGUC.46
It remains challenging to make a diagnosis of UTLGNU, since the cytomorphologic features of UTLGUN tissue fragments overlaps with those observed in benign/reactive tissue fragments. The presence of urothelial tissue fragments with fibrovascular cores would be consistent with the clinical impression of LGUN, provided that the diagnostic features for high-grade lesions are not observed in the specimen.
Performance of urinary cytology for upper tract urothelial carcinoma
Urinary cytology has variable sensitivity (19–100%), specificity (86–100%), and positive predictive value (92–100%) for the diagnosis of UTHGUC.47–49 The overall sensitivity for UTLGUN is significantly lower, when compared to that for HGUC, with reported sensitivities of 46% for UTLGUN prior to TPS vs. 0–34% (average 17%) after the implementation of TPS.50–52
Clinical management of upper tract urothelial carcinoma
Ideally, a diagnosis of HGUC in a selective cytology specimen should be confirmed through tissue diagnosis before any definitive treatment.51 The decision to proceed with the nephroureterectomy solely based on the UT cytology result (HGUC or SHGUC), without an endoscopically or radiographically visible lesion, is generally not recommended. These patients should be closely monitored, to identify the source of the abnormal cells, and provide treatment accordingly, in a timely fashion.
Special consideration for collection, preparation and adequacy
UUT cytology is better than VU, since the usefulness of VU specimens remains limited. Tissue biopsy with uroscopy should be performed. All blocks prepared from instrumented specimens may contribute to the diagnosis of both HGUC and LGUN. More than 20 urothelial cells per 10 HPFs are required for the statement of adequacy for UT specimens.
Ancillary testing for upper tract urothelial carcinoma
FISH/UroVysion combined with a cytology morphology study would improve the diagnostic accuracy of UTUC.
Non-urothelial malignancies (NUM) and other miscellaneous lesions
Non-urothelial malignancies are rare in urine specimens, accounting for 0.1%.
Primary epithelial malignancies
Primary squamous cell carcinoma
Primary squamous cell carcinoma is characterized by exclusive squamous differentiation, without other associated malignant components.
Cytologic criteria:2
Keratinizing squamous cell carcinoma: Pleomorphic large squamous cells that present as single cells, or in nests or three-dimensional clusters, and have polygonal, spindle, or bizarre (such as with a tadpole, etc.) shapes, with moderate-to-abundant keratinized/orangeophilic cytoplasm, well-defined sharp borders, and markedly atypical hyperchromatic/dark-ink nuclei. Squamous pearls and “cell in cell” patterns may be observed (Fig. 5i);
Non-keratinizing squamous cell carcinoma: Specimens that contain single and clusters of malignant squamous cells, with intercellular bridges, prominent nucleoli, and a scant-to-moderate dense cytoplasm;
Anucleated squamous cells (“ghost cells”), small atypical parakeratotic cells, necrosis and neutrophils may be observed in the background.
Atypical squamous cells
Atypical squamous cells are rarely observed in urine cytology (0.3–0.9%), and are possibly associated with bladder or cervical cancer in up to 20–30% of cases.53 Atypical squamous cells can also be observed in HPV infection-associated lesions of the bladder and urethra.
Cytologic criteria:
Atypical squamous cells with large and hyperchromatic nuclei, abnormal nuclear or cytoplasmic shapes, and densely orangeophilic cytoplasm;
Atypical squamous cells that quantitatively and/or quantitatively fall short of the diagnosis of squamous cell carcinoma.
Primary adenocarcinoma
Primary adenocarcinoma is characterized by exclusive glandular differentiation, without the association with other malignant elements.
Enteric adenocarcinoma
These tumor cells present with a similar morphology to colonic adenocarcinoma, clusters of columnar glandular cells arranged in glandular architectures, eccentric hyperchromatic nuclei, and moderate, delicate, or vacuolated cytoplasm, in the background of abundant necrosis. Signet-ring cells may be observed.
Mucinous adenocarcinoma
The atypical columnar cells would be arranged in rounded, three-dimensional clusters, and have small-to-moderate amounts of delicate or vacuolated cytoplasm, and medium-sized nuclei with a distinct nucleoli, in the background of abundant mucin (Fig. 7a, b). The tumor cells may be positive for CK7 and/ or CK20 (Fig. 7c).
Clear cell adenocarcinoma
The large tumor cells present in clusters with occasional hobnail configurations, which have abundant vacuolated cytoplasm and centrally located nuclei that contain prominent nucleoli.
Notes
Since adenocarcinomas in urine cytology do not generally present with distinctive cytomorphologic features to allow for the determination of the subtype, it would be most appropriate to report these cases as adenocarcinoma, NOS, and draft a comment with differentials. Immunostains may be helpful for the definitive diagnosis in some cases, in order to distinguish primary adenocarcinoma from metastatic adenocarcinoma and subtyping adenocarcinoma.
Primary neuroendocrine tumors
Primary neuroendocrine tumors include well-differentiated neuroendocrine tumors (WDNETs), neuroendocrine carcinomas (NECs), including small-cell carcinoma (SmCC), large-cell NEC (LCNEC), and paraganglioma.
Small-cell carcinoma
SmCC is rare that often associated with HGUC or other subtypes of carcinoma.
Cytologic criteria:2
Small-to-medium-sized tumor cells may be arranged in single, loose, or tight clusters, occasionally in linear pattern or rosette formation (Fig. 7d);
The tumor cells contain round/oval nuclei with granular or stippled hyperchromatic chromatin, and present with nuclear molding, scanty cytoplasm, and high N/C ratios. Mitoses, apoptosis, necrosis, crushing artifact, and hemorrhage may be observed (Fig. 7e);
The tumor cells are immunoreactive to neuroendocrine markers, synaptophysin (Fig. 7f), chromogranin A, CD56 and INSM1, and possibly immunoreactive to TTF1.
Paraganglioma
Paraganglioma arises from the embryonic nests of chromaffin cells in the sympathetic plexus of the detrusor muscle. Patients would commonly present with hypertension, headache, and hematuria. These tumor cells are rarely detected in urine.
Cytologic features:
Large epithelioid cells that present in single or loose cohesive clusters. Tumor cells with round-to-oval nuclei, fine granular chromatin and inconspicuous nucleoli, and a moderate amount of cytoplasm;54
The differential diagnosis would include HGUC, plasmacytoid UC, lymphoma, melanoma, NET, and sarcoma.
Secondary epithelial malignancies
Metastatic carcinomas from the breast, gastrointestinal tract, gynecologic origin, liver, lungs, kidneys and prostate (Fig. 7g) are rarely detected in urine cytology. The cytomorphology of these metastatic carcinomas is similar to the primary tumor, and may mimic primary urothelial or non-urothelial malignancies. Immunostains performed on cell blocks (Fig. 7g, i), e.g., NKX3.1 for confirmation of prostatic primary or slides may help confirm the origin of the carcinoma.
Non-epithelial malignancies
Non-epithelial malignancies, such as sarcoma, melanoma and lymphoma, account for less than 0.5% of all bladder tumors.55
Sarcoma
Sarcoma with smooth muscle, skeletal muscle, endothelial, or no specific differentiation can arise in the urinary tract.56,57 The diagnosis of sarcoma by urine cytology is seldom made.
Cytologic criteria:
Pleomorphic spindle cells present as single cells, or in sheets or clusters, are pleomorphic, have hyperchromatic nuclei that possibly contains prominent nucleoli, and have a moderate amount of cytoplasm with ill-defined cytoplasmic borders;
The nuclei of leiomyosarcoma would have typical oval-to-spindle shapes with blunt ends;58
When high-grade or pleomorphic tumor cells, especially those with spindle shapes, are present in urine specimen, the differential should include primary or metastatic sarcoma, primary or metastatic HG carcinoma with sarcomatoid differentiation, melanoma, and lymphoma;
Immunostains and molecular studies would be helpful for the diagnosis.
Malignant melanoma
Melanoma is rarely detected in urine cytology59–61
Cytologic criteria:
Atypical tumor cells present in single cells;
The tumor cells have an epithelioid, plasmacytoid, or spindle in shape, contain large, pleomorphic nuclei that contain prominent/macro-nucleoli and occasional intranuclear pseudoinclusions, and have moderate-to-abundant cytoplasm;
Intracytoplasmic, dark, finely dusty brown-to-black melanin pigments may be observed in tumor cells and macrophages;59,62
The tumor cells are immunoreactive to SOX10, Melan A, HMB45 and S100, but negative for cytokeratins.
Lymphoma
The most common primary lymphoma is low-grade, extranodal marginal zone B-cell lymphoma of mucosa-associated lymphoid tissue (MALT) origin. This is commonly associated with chronic cystitis, and affects adults older than 60. The most common high-grade lymphoma is diffuse large B-cell lymphoma (DLBCL).63
Cytologic criteria:
Atypical lymphoid cells with nuclear membrane irregularity are the general features of lymphoma. DLBCL is characterized by the presence of atypical large lymphoid cells, with large round-to-oval nuclei, dense chromatin, irregular nuclear membrane, prominent nucleoli, and moderate cytoplasm (Fig. 7j, k);
The differential diagnosis of high-grade lymphoma includes HGUC (especially plasmacytoid HGUC), SmCC, melanoma, and sarcoma;
Immunohistochemistry (IHC) and flow cytometry to detect lymphoid markers, such as Pax5 positivity in DLBCL (Fig. 7l), FISH, and molecular studies can help to confirm the diagnosis.
Primary plasmacytoma
Primary plasmacytoma is rare.64
Cytologic criteria:
Neoplastic cells present as single cells, and have eccentric nuclei that contain peripheral clumping chromatin, binucleation, and prominent nucleoli, and dense basophilic cytoplasm with perinuclear hof;
The differential diagnosis includes HGUC, especially plasmacytoid UC, lymphoma, melanoma, NET, and sarcoma;
IHC and flow cytometry to detect CD138, CD38, CD56, kappa and lambda light chain are useful for the diagnosis.
Nephrogenic adenoma (NA)
NA is a common benign lesion that most commonly involves the urinary bladder. NA presents in various growth patterns, including tubular, cystic, tubulocystic, papillary, and flat. These tumor cells range from cuboidal to low columnar cells, with eosinophilic cytoplasm.65 Clear cell change, degenerative atypia, and a hobnail appearance can occur.
Cytologic criteria:
The tumor cells are arranged in small groups or papillae, or present as isolated cells, have polygonal or columnar shapes, and moderate vacuolated cytoplasm, and present with no or mild nuclear atypia, with slightly increased N/C ratios, nuclear hyperchromasia, and prominent nucleoli;
The tumor cells are diffusely positive for Pax-8, Pax-2 and AMACR;
The differential diagnosis includes reactive changes, HGUC, clear cell adenocarcinoma, and low-grade RCC.
Ancillary studies in urinary cytology
Merely few tests have been approved by the US Food and Drug Administration (FDA) for diagnostic application, including UroVysion® fluorescence in situ hybridization (U-FISH), BTA™, and NMP22™. Furthermore, few tests have been studied and proposed, including tests for tumor-associated antigens, polymerase chain reaction (PCR)-based tests for mutations, and next-generation sequencing (NGS) to detect mutations or epigenetic changes in UC.
UroVysion FISH66
Papanicolaou-stained slides prepared from any urine sample can be used for the test.67 U-FISH contains four single-stranded DNA probes, three chromosome enumeration probes (CEPs) that target the pericentromeric regions of chromosomes 2, 7, and 17, and another locus-specific identifier (LSI) probe that targets the 9p21 locus. After staining, the slides are analyzed using an epi-fluorescence microscope equipped with a 100 watt mercury lamp or LED light source, and appropriate filters to detect multicolor fluorescent signals,2 and the magnification ranges from 600× to 1,000×. The number of signals for all four probes should be counted and recorded as abnormal when there is a gain (≥2 signals) for two or more chromosomes 3 (red), 7 (green), and 17 (aqua), or when there is a loss of both copies of 9p21 (gold), as suggested by the manufacturer. The test is considered positive when ≥4 of the 25 analyzed cells present with gains for two or more chromosomes, or ≥12 of these 25 cells have zero 9p21 signals (Fig. 8).
The imaging and automation of the UroVysion FISH analysis saves time, improves productivity, quality control, and the archiving of images, and increases accuracy.68 Automated imaging systems consist of an automated scanning microscope coupled with software for image analysis. Target FISH is performed on Papanicolaou-stained monolayer preparations that have been pre-scanned by the imaging system, with atypical urothelial cells selected by the operator.
UroVysion FISH has a variable sensitivity of 89–100%, and a specificity of 60–100%.66,69 A new category, “urothelial neoplasia diagnosed by U-FISH, not otherwise specified (UNF-NOS),” was recently proposed. However, there have been concerns on its cost-effectiveness. Furthermore, there is a possibility of false-positive results in the UroVysion FISH analysis.70
FDA-approved liquid-based tests
Two versions of the bladder tumor antigen (BTA) test were approved by the FDA: quantitative ELISA test (BTA) TRAK™ and qualitative point-of-care test BTA stat. BTA stat and BTA TRAK™ have similar overall sensitivity (64% vs. 65%) and specificity (77% vs. 74%), respectively.71 These tests have higher sensitivity, when compared to urine cytology. However, similar to NMP22, BTA assays suffer from a higher false-positive rate in patients with benign urinary tract diseases, limiting its use in clinical practice.72
The nuclear matrix protein (NMP22) test was approved by the FDA. However, this test should not be performed on voided urine samples obtained within five days after the instrumentation of the urinary bladder. False-positive results are common in patients with benign bladder conditions.73
Other liquid-based biomarkers
Various potential ancillary tests have emerged based on protein detection, microRNAs, gene expression profiling, epigenetic changes, and PCR-based detection.
Ancillary tests based on next-generation sequencing technology
NGS is a technique for acquiring the DNA sequence from each starting molecule of DNA in a sample. Various different mutations can be present in different cases of UC, permitting the simultaneous identification of multiple mutations.74 Mutations, epigenetic alterations, and even copy number changes can be detected by NGS. Epigenetic changes (DNA methylation) can be detected using chemical treatment (bisulfite) to convert methylated cytosines to a nucleotide that is ultimately sequenced as thymidine. Then, the locations of these methylated cytosines can be computationally deduced. Copy number changes (which can represent aneuploidy) are deduced based on the number of individual reads. The commercially available NGS tests are uCAPP-Seq,75 AssureMDx,76 and UroSeek,77,78 with a sensitivity of 83–93%, and a specificity of 86–99%.
Risk of high-grade malignancy15
The new term, ROHM, rather than ROM, was introduced in TPS 2.0. After applying TPS 1.0, various studies have been conducted to calculate the ROHM of each category, including five studies that retrospectively reclassified the previous diagnosis using TPS 1.0, and six studies that applied the TPS 1.0 criteria for the new urine cytology diagnosis.2,6,32,33,35,79–86 The diagnosis criteria for LGUC in urine cytology remains very strict in TPS, resulting in significantly fewer diagnosed cases of LGUC. The calculation of ROHM of LGUC remains limited by the case number, which ranges from 0% to 54.1%.2,24,32,35,38,79,81–89
The weighted overall ROHM is calculated based on the sample size and standard deviation in each category (Table 1), presenting the overall updated performance of TPS 1.0.
Table 1Estimated risk of high-grade malignancy (ROHM) for each category of The Paris System for Reporting Urinary Cytology2
| TPS category | Risk of high-grade malignancy |
|---|
| ND | 0–16% |
| NHGUC | 8–24% |
| LGUN | 0–44% |
| AUC | 24–53% |
| SHGUB | 59–94% |
| HGUC/malignant | 76–100% |
Clinical management
The implementation of TPS has standardized the diagnostic criteria, thereby limiting the interobserver variability in cytologic interpretation, creating a more reproducible diagnostic tool, and improving the communication among clinicians and cytopathologists.
Management of the unsatisfactory specimen2
The management of unsatisfactory specimens should be left at the clinician’s discretion. The reasons for the unsatisfactory specimen should be investigated by cytopathologists and clinicians, and a repeat sample may be obtained depending on the risk for a significant lesion.
Management of negative for high-grade urothelial carcinoma2
The American Urological Association (AUA)/Society of Urodynamics, Female Pelvic Medicine, and Urogenital Reconstruction (SUFU) proposed new guidelines that do not recommend the use of cytology in the initial evaluation of microscopic hematuria.90 These guidelines recommend a patient-centered approach to diagnostic testing for microhematuria, based on the risk for urothelial malignancy. The AUA/Society of Urologic Oncology (SUO) proposed new guidelines that recommend a risk-stratified approach in the surveillance of non-muscle-invasive bladder cancer.91 Urine cytology or other urinary biomarkers are not recommended for patients with a history of low-risk cancer, and a normal cystoscopy. Urine cytology is essential in the surveillance of patients for recurrences following therapy, since these patients remain at risk for recurrences in the remnant urothelium (upper tracts and urethra).92–94 The role of additional molecular testing, such as UroVysion FISH and other urinary biomarkers (including ImmunoCyt and Cxbladder), remains to be determined. A patient with a diagnosis of NHGUC may continue the routine surveillance at intervals that commensurate with the risk of recurrence.
Management of atypical urothelial cells
The implementation of TPS has decreased the rate of AUC diagnoses.36,95 The workup for AUC should be individualized based on the patient’s risk assessment, and this may prompt a thorough evaluation. Patients with hematuria or persistent irritative voiding symptoms still require a thorough evaluation, with upper tract imaging and cystoscopy. Patients with known risk factors for urothelial carcinoma and atypical cytology should receive further investigation, in order to rule out any malignancies. For patients with a prior history of urothelial malignancy, the extent of the work-up would be dependent on the clinical suspicion of recurrent disease.96 The role of additional molecular testing, such as UroVysion FISH and other urinary biomarkers, remains to be determined.93,97,98 Some centers have instituted reflex UroVysion FISH testing for the adjudication of AUC diagnoses. A positive FISH assay would be managed similar to a suspicious diagnosis, while a negative FISH test would be expectantly followed.99–102
Management of suspicious and positive for high-grade urothelial carcinoma
From a practical standpoint, the clinical management of “SHGUC” is identical to the HGUC diagnosis, requiring an active investigation to identify the source of the suspicious or positive cells by utilizing regular cystourethroscopy103,104 (fluorescence cystoscopy, narrowband imaging, and directed/random bladder biopsies), and evaluate the prostatic urethra and upper tracts. The upper tract can be evaluated through imaging studies, computerized tomography (CT) urography, magnetic resonance (MR) urography, and ultrasound with retrograde pyelography.
Management of low-grade urothelial neoplasms
Most bladder cancers present as low-grade non-invasive papillary tumors, and diagnosing LGUC based on cytology can be challenging, with a relatively low sensitivity. Transurethral resection allows for the establishment of the histologic diagnosis, and is therapeutic for most solitary low-grade tumors. The AUA and European Association of Urology (EAU) recommends a risk-adapted surveillance protocol for non-muscle-invasive bladder cancer.91,105 Routine surveillance cystoscopies may be performed at regular intervals.106 The decision to provide adjuvant intravesical therapy (chemotherapy or immunotherapy) is based on the risk of recurrence and progression.107 The stratification of patients into low-, intermediate-, and high-risk groups is essential for deciding the appropriate use of adjuvant intravesical chemotherapy (mitomycin, doxorubicin, or gemcitabine) or BCG instillation. Surgical removal of the bladder should be considered in case of BCG-unresponsive tumors, or NMIBCs with the highest risk of progression.106
Management of non-urothelial tumors
It may be difficult to distinguish primary non-urothelial malignancy from urothelial carcinoma with divergent histologic differentiation in urine cytology, which most often requires histology and ancillary studies on biopsy or resection specimens. The complete resection of all visible tumors, when appropriate, is recommended. A multidisciplinary approach, which employs surgery, systemic chemotherapy, and radiation on an individualized basis, should be considered. The standard treatment for primary squamous cell carcinoma of the bladder is radical cystectomy, followed by radiotherapy, with concurrent chemotherapy for unresectable or residual tumors.108 Urachal carcinoma is treated through the wide local excision of the umbilicus and urachal remnant, with cystectomy and pelvic lymphadenectomy.109 Primary adenocarcinoma is treated by partial cystectomy.110 Neuroendocrine tumors should be managed using a multimodality approach, with neoadjuvant systemic chemotherapy, followed by radical cystectomy. The combination chemotherapy with definitive radiotherapy may be considered.111,112
Conclusions
The second edition of the Paris System for Reporting Urine Cytopathology further clarifies and defines the diagnostic criteria of each category. LGUN was re-categorized in NHGUC, and a new chapter, Cytology of the Upper Urinary Tract, was added. According to the feedback after the publication of TPS 1.0, the updated ROM and management suggestions in each category provides clarification for the communication among pathologists, radiologists and urologists. The primary purposes of TPS are to decrease the atypical category, and increase the specificity of the HGUC diagnosis from urine cytology. The measurement of the nuclear area may improve the accuracy of TPS. Future studies are needed, in order to further define the utilization of artificial intelligence (AI) and molecular testing in urine cytology, and refine AI, as well as the molecular basis of HGUC and biomarkers for AUS cases.
Abbreviations
- AUC:
atypical urothelial cells
- HGUC:
high-grade urothelial carcinoma
- LGUN:
low-grade urothelial neoplasm
- ND/U:
nondiagnostic/unsatisfactory
- NHGUC:
negative for high-grade urothelial carcinoma
- SHGUC:
suspicious for high-grade urothelial carcinoma
- TPS:
The Paris System for Reporting Urine Cytology
Declarations
Funding
None.
Conflict of interest
The manuscript was submitted during Dr. Xiaoqi Lin’s term from June 2021 to June 2023 serving as an editorial board member of Journal of Clinical and Translational Pathology. The authors have no other conflict of interests to declare.
Authors’ contributions
Dr. Fei Chen wrote the draft of the manuscript, and Dr. Xiaoqi Lin conducted the final review and editing, and provided the figures with legends. Both authors significantly contributed to the study, and approved the final manuscript.