Section 2 -

Urine Cytology – A Daily Approach Warick Delprado Douglass Hanly Moir Pathology Sydney, Australia

"Urine Cytology" has previously been referred to as difficult and poorly reproducible. Some of the causes of this have been: the confusion in classification and terminology, and a lack of understanding of the pathobiology of urothelial neoplasms, perceived degenerative changes in cells caused by the so-called toxic environment of urine resulting in reports along the lines of "atypical degenerate cells present" which drive clinicians to distraction, the more glamorous areas of cytology, and associated disinterest in urine, overshadowing the study and teaching of urine cytology, and urologists occasionally not finding any abnormality despite urine cytology being specific and definite, resulting in a lack of confidence in reports, despite the fact that that it is known that some carcinomas can take a long time to become clinical. Some of the problems in the daily practice of urine cytology can be seen in the following case. A 70 year old female presented with haematuria. Investigation and subsequent bladder biopsy showed a non-invasive low grade papillary urothelial carcinoma. Over the next nine months repeat cystoscopies showed a PUNLMP and later a non-invasive papillary urothelial carcinoma with associated carcinoma in situ. There was no invasion. She was treated with several courses of BCG. Twelve months after presentation, biopsy showed chronic inflammation and atypical urothelium. At 15 months, urine cytology (x3) was "malignant – high grade". Subsequent biopsy showed low grade papillary urothelial carcinoma. Six months later, repeat cytology was "highly suspicious of malignancy". Single bladder biopsy showed PUNLMP. Six months later, (two years after initial presentation) an enlarged para-aortic lymph node biopsied showed metastatic high grade carcinoma, consistent with urothelial carcinoma. This case demonstrates the major problems that can be encounted relating to classification, terminology and communication. PUNLMP was considered by the clinician as an appropriate finding after "malignant" and "highly suspicious of malignancy" cytology. The cytology was actually pointing to the initial high grade carcinoma in situ which progressed to invasive malignancy. This confusion can lead to lack of faith in urine cytology by clinicians and discomfort with performing urine cytology by cytologists and cytopathologists. So how do cytopathologists and cytologists overcome these problems? This involves quality maintainence and improvement. The best way is to: find out what you are doing, by reviewing practices and reporting, then find out what you should be doing by reviewing diagnostic criteria, and finally doing what needs to be done by setting in place an approach that allows you to report accurately and reliably on a daily basis. What are we doing? It is important to understand or measure the reports being issued by a department. The purpose is to ensure that all cytopathologists and cytologists are using the same terminology and criteria. A review such as this can be done in two ways. Firstly, a series of cases can be reviewed and rescreened to ensure no repeated errors are occurring. Ideally, this should be done with cases that have cytological and pathological correlation. Secondly, a formal case review can be carried out with a fixed number of known cases circulated to everyone within a department. Errors or differences in opinion should be analysed and discussed. What should we be doing? Standard reporting terminology and guidelines should be agreed upon within a department. This should start with ensuring that everyone is using the same terminology. Firstly, all reports should use the current WHO (2004) classification of urothelial neoplasia. Secondly, criteria based upon published series should be uniformly used in all reports. What do we need to do? A simple approach to urine cytology can be defined to take most of the heartache from reporting urine. This involves attitudinal approach as well as an approach to each slide. An Approach to Diagnosis in Urinary Cytology. 1. Don't panic 2. Be aware of the pecularities of urothelial neoplasia histopathology This is covered in length in Dr Murphy's presentation. Based upon disagreement between pathologists, and misconceptions of the pathology of urothelial neoplasia, urine cytology has been made, or at least been made to seem, harder than it really is. 3. Remember basic cytology Traditional criteria of malignancy still apply in urine cytology and allow recognition of nearly all high grade lesions. Difficulties relate to low grade lesions, that are less clinically significant. 4. Apply standard terminology Non gynaecological cytology is traditionally reported in four categories – negative, atypical, suspicious and malignant. With some modification, these can be used in urine cytology. Use of the term "atypical" is controversial and can be problematic. However, provided terminology is clearly defined and explained to clinicians, the categories give good information and allow clear clinical recommendations. The use of "dysplasia" is useful in the context of PUNLMP. Recommended reporting is: Negative (includes reactive) Atypical (not normal, but low likelihood of malignancy) – can be reviewed and followed Suspicious (not normal, with high likelihood of malignancy) - investigate Malignant – low grade or high grade carcinoma Dysplasia – "significantly abnormal cells present with features short of malignancy and consistent with flat dysplasia or PUNLMP" 5. Look for patterns Pattern recognition is an important aid in sharpening diagnostic accuracy. However, it is not the diagnosis. Patterns draw attention to the possibility of malignant cells, which then must be specifically identified. Diagnosis of malignancy eventually depends only on cellular criteria of malignancy. Structure Traditionally, we have been taught, and have taught, that cellular tissue aggregates and papillary fragments should be largely ignored when reporting urine cytology. They are often related to instrumentation. Conventional wisdom was that one should note the presence of fragments, but look to individual cells for a diagnosis. Whilst this is largely correct, the shape of cellular aggregates does give additional information that assists in making a diagnosis. There are three important patterns: "Tissue fragments" - with curled, or balled shape. These result from shed or displaced "intact" sheets of cells from a surface. They have rounded or curled- up margins. There is often a surface along some of the margins. They usually occur as part of a reactive pattern or as a result of instrumentation, particularly as part of a bladder wash. However, they can occur in neoplasia, mainly transitional cell carcinoma in situ, and are more pronounced in bladder washes. (C or J shape) "Papillary fragments" - branched, nodular crowding and overlapping These result from being shed from the margins of papillary neoplasms. They are characterized by poor surface definition and irregular, curled-up margins with nodular areas of overlapping cells. They are seen in papillary neoplasia, both low and high grade. (E shape) "Cellular aggregates" - dissociating These result from shedding of dissociating cells in loose clusters. These cells show loss of cohesion and separation (falling apart), particularly at the edges. No surface is identified. This pattern appears to be accentuated in bladder washes, and as such may be partly artefactual. It is seen in high grade papillary neoplasia and transitional cell carcinoma in situ. Background The presence of debris, inflammation or cellular material can alert to the possibility malignancy. However, malignant cells must still be identified for a diagnosis. "Apoptotic Debris Cells" Result from incipient necrosis of superficial urothelial cells in tumours. They are small nuclei with hyperchromatic (dense) uniform chromatin, less than that of a leucocyte in diameter and with variable cytoplasm. This pattern of cells is seen rarely in PUNLMP, but commonly in papillary tumours, both low and high grade. It can also be seen in TCC in-situ. The pattern results from nuclear shrinkage as part of cell loss from superficial layers. However, it may also be seen following intra-vesicle therapy associated with extensive tumour necrosis. These cells should be recognized and used as a marker for, but are not diagnostic of malignancy. A search should be made for classical malignant cells. If these are absent, the appearances should be called suspicious rather than "atypical degenerate". "Inflammation" A difficult and time consuming presentation, characterised by an overwhelming acute inflammatory cellular pattern that masks individual abnormal cells. "Squamous debris" Numerous squamous cells can mask individual malignant cells 6. Know the clinical history Whilst it is true that smears can be interpreted in isolation, patient history affects the presentation on the slide. Is this an initial presentation or a recurrence? Has there been previous therapy and what type? Is the specimen a urine or a bladder wash? 7. Look at the slide Urine cytology should not be left till the end of the day after more important cytology has been reported. Report as part of the routine work, taking your time and don't forget the high power lens to search between the background. Overall, there is a sequence of questions to the approach to urine cytology: Clinical New / Recurrence (? review previous) / Post treatment Specimen Urine / Bladder Wash / Catheter specimen / Ileal Conduit Patterns Background Normal / Inflammation / Diathesis / Apoptosis Other Bacterial overgrowth / Crystals / Sperm (caution) Tissue fragments Nil / number - ? urine / wash Types Curled / balled (?reactive) [J] Branched / papillary [E] Dissociating (insitu) Cells Number / Populations (? <2 or >2) Apoptotic / Squamous (?metaplasia) Nuclei Size (N/C ratio) / Position (central / eccentric) Shape - Uniform / irregular ? Smooth / Folds / Grooved / Sharp edges / angles Chromatin Normal / Uniform / Granular / Irregular (? Quadrants) Nucleoli Size / Shape / Smooth / Angular Cytoplasm N/C Ratio Vacuoles / solid / squamous References: "WHO Blue Book": Eble JN, Sauter G, Epstein JI, Sesterhenn IA. WHO/IARC Classification of Tumours. Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs. IARC Press. Lyon 2004 Montironi R, Lopez-Beltran A, The 2004 WHO Classification of Bladder Tumours: A Summary and Commentary. International Journal Surgical Pathology 2005.13(2):143-153. Murphy WM. Urinary Cytopathology. ASCP Press 2000 Murphy MW, Soloway MS, Jukkola AF et al. Urinary Cytology and bladder cancer. The cellular features of transitional cell neoplasms. Cancer 1984. 53:1555-1565 Foster CS, Ross JS. Pathology of the Urinary Bladder. Saunders 2004 Allen, DJ, Challacombe, B, Clovis, JS, Chandra, A, Dasgupta, P and Popert, R. Urine cytology: appropriate usage maximizes sensitivity and reduces cost. Cytopathology 16:3, 139-42 (2005) Nabi, G, Greene, DR and O'Donnell, M. How important is urinary cytology in the diagnosis of urological malignancies? Eur Urol 43:6, 632-6 (2003) Whisnant, RE, Bastacky, SI and Ohori, NP. Cytologic diagnosis of low-grade papillary urothelial neoplasms (low malignant potential and low-grade carcinoma) in the context of the 1998 WHO/ISUP classification. Diagn Cytopathol 28:4, 186-90 (2003) Curry, JL and Wojcik, EM. The effects of the current World Health Organization/International Society of Urologic Pathologists bladder neoplasm classification system on urine cytology results. Cancer 96:3, 140-5 (2002) Owens, CL and Ali, SZ Atypical squamous cells in exfoliative urinary cytology: clinicopathologic correlates. Diagn Cytopathol 33:6, 394-8 (2005) Deshpande, V and McKee, GTAnalysis of atypical urine cytology in a tertiary care center. Cancer. 2005 Dec 25;105(6):468-75 Raitanen, MP, Aine, RA, Kaasinen, ES, Liukkonen, TJ, Kylmälä, TM, Huhtala, H, Tammela, TL and Finnbladder Group. Suspicious urine cytology (class III) in patients with bladder cancer: should it be considered as negative or positive? Scand J Urol Nephrol 36:3, 213-7 (2002) Renshaw, AA. Subclassifying atypical urinary cytology specimens. Cancer 90:4, 222-9 (2000) Nasuti, JF, Fleisher, SR and Gupta, PK. Significance of tissue fragments in voided urine specimens. Acta Cytol 45:2, 147-52 (2001) Mack, D and Frick, J. Diagnostic problems of urine cytology on initial follow-up after intravesical immunotherapy with Calmette-Guérin bacillus for superficial bladder cancer. Urol Int 52:4, 204-7 (1994)