The use of nuclear morphometry in predicting recurrence of transitional cell carcinoma

Despite apparently adequate surgical resection, metastases develop in about 50% of the patients who undergo radical cystectomy for invasive bladder cancer and they die of disease within 2 years of surgery. Recent reports have shown that combination chemotherapy with regimens including cisplatin results in significant improvement in time to progression and overall survival for patients with transitional cell carcinoma. We have used quantitative nuclear morphometry in an effort to predict which individual patients with muscle invasive transitional cell carcinoma are at increased risk for recurrence and in need of adjuvant therapy. A total of 14 patients with deeply invasive tumors without known metastases (tumors, nodes and metastasis stage P3A-B, N-, M0) who had not received perioperative radiation or systemic chemotherapy was studied. Of the patients 7 (group 1) had no evidence of disease with a mean followup of 41 months and 7 (group 2) had recurrent transitional cell carcinoma at a mean followup of 17 months. Both groups were similar with respect to age and could not be distinguished by standard pathological analysis of stage (p = 0.66) or grade (p = 0.99). A total of 150 tumor nuclei from each patient was digitized and analyzed by a high resolution quantitative morphometric imaging system, developed in this laboratory, to determine if nuclear shape descriptors could separate these 2 clinically distinct groups. A multivariate analysis combining 3 independent nuclear shape descriptors (average nuclear area, minimum feret-diameter ratio and kurtosis of feret-diameter ratio) was highly significant (p = 0.003) and separated patients in group 1 from group 2 without overlap. These encouraging results suggest that nuclear morphometry may be valuable in determining which patients are at increased risk of disease recurrence following radical cystectomy and, therefore, who should receive adjuvant chemotherapy.