[358] Cytology for the Diagnosis and Surveillance of Upper Urothelial Tract Carcinoma – A Single Cancer Center Experience

Jasreman Dhillon, Philippe E Spiess, Wade J Sexton, Julio M Pow-Sang, Barbara A Centeno. Moffitt Cancer Center, Tampa, FL

Background: Upper urinary tract (UUT) urothelial carcinoma (UC) is difficult to sample with routine biopsy devices due to the confined anatomy of the renal pelvis and ureter. This study highlights the merits of cytology for diagnosis of UUT UC.
Design: Patients with UUT cytology samples collected at our institution from 01/2009 to 12/2009 were identified from the LIS under an IRB approved protocol. Results of UUT brushings, washings, biopsies, and resection specimens were analyzed. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) was calculated for cytology samples. The final diagnosis was established based on biopsy/resection specimens and clinical follow up.
Results: Thirty-four cases were retrieved (13 brushings, 31 washings) from 31 patients (22 males, 9 females). 13 patients were under surveillance for a prior UC. Table 1 shows the results for all sample types compared to the final diagnosis. Brushings were diagnostic in 12 of 13 cases (Table 2). Washings were diagnostic in all 31 cases (Table 2). For the purpose of statistical analysis, benign and atypical diagnoses were considered as negative and suspicious and malignant diagnoses were considered as positive. In our study, washings (sensitivity = 71%, specificity = 90%, PPV = 94% and NPV = 60%) yielded better results as compared to brushings (sensitivity = 60%, specificity = 50%, PPV = 86%, NPV = 20%). Calculi caused two false positives (1 brushing, 1 washing). Two HG UUTUC had only urine cytology samples which were negative. Seventeen patients had biopsies, and sixteen patients underwent nephroureterctomy/ureterectomy. Clinical follow-up was obtained on all 31 patients.

Table 1. Distribution of Diagnoses rendered by different modalities
DiagnosisNumberUnsatisfactoryBenignLG UCHG UCCISUC NOS
Final Diagnosis3401051612

Table 2. Comparison of Brushings and Washings to Final Diagnoses
Diagnostic ModalityBenignLG UCHG UCUC, NOS
Brushings - Benign/Atypical (n = 5)1130
Brushings - Positive/Suspicious (n = 7)1060
Washings - Benign/Atypical (n = 15)9330
Washings - Positive/Suspicious (n = 16)11122

Conclusions: These results confirm that biopsies are performed less often than washings. Washings performed better than brushings for the diagnosis of UUTUC. Urine cytology alone is not effective at diagnosing UUTUC.
Category: Cytopathology

Tuesday, March 5, 2013 11:00 AM

Proffered Papers: Section F, Tuesday Morning


Close Window