Comparison of quantitative and classic prognosticators in urinary bladder carcinoma. A multivariate analysis of DNA flow cytometric, nuclear morphometric and clinicopathological features.

The prognostic value of nuclear morphometry and DNA flow cytometry of paraffin embedded material of 58 patients with primary and untreated transitional cell carcinoma of the bladder was compared with that of histological grade (WHO-system), tumour stage (TNM-classification), tumour size, multiplicity and ulceration. Small nuclear size (mean nuclear area less than or equal to 95 microns2) (n = 25) and DNA diploidy (n = 28) indicated a favourable outcome (5-year survival 95.8% and 92.2%); large nuclei (mean nuclear area greater than 95 microns2) (n = 33) and DNA aneuploidy (n = 30) indicated a worse prognosis (5-year survival 61.4% and 62.5%) (Mantel-Cox; p = 0.002 and p = 0.007). The quantitative techniques had the advantage over subjective histological grading that distinguishment of an intermediate patient group (WHO-system: grade 2; n = 32) with heterogeneous outcome (5-year survival 78%) was avoided. Multivariate analysis showed tumour stage as the most important prognosticator of survival. Neither the quantitative techniques, nor the other classic features added significantly to the prediction. The additional value of the quantitative techniques was however shown in superficial carcinoma (TNM-classification: stage Ta and T1; n = 37); large nuclei (mean nuclear area greater than 95 microns2) (n = 15) and aneuploid DNA peaks (n = 13) were associated with progressive recurrent tumour (n = 7) (Mantel-Cox: p = 0.03 and p = 0.0004). The quantitative methods thus indicate which patients are at risk for progression and may enable more appropriate treatment at an earlier stage of disease.