Combinations of urine-based tumour markers in bladder cancer surveillance

OBJECTIVE: The aim of the study was to investigate whether combinations of urine-based tumour markers including urinary cytology (Cytology or Cyt) increase the sensitivity in the detection of bladder cancer recurrence. MATERIAL AND METHODS: Urinary cytology, NMP22, UroVysion (FISH) and ImmunoCyt (uCyt+) were determined in 221 patients during the follow-up of non-muscle-invasive transitional cell carcinoma (NMI TCC) before cystoscopy (n = 49) or with the suspicion of TCC recurrence before transurethral resection of the bladder (n = 173). For all markers alone as well as in all possible combinations (multimarker panels, MPs) sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy were evaluated. MPs were considered positive if at least one marker was positive. RESULTS: No malignancy was found in 108 patients, whereas recurrent TCC was confirmed in 113 patients. Sensitivity and specificity for Cytology were 84% and 62%, for NMP22 68% and 49%, for FISH 72% and 63%, and for uCyt+ 73% and 62%, respectively. The NPV was below 80% for all markers alone. Combinations of two and three markers increased the sensitivity as well as the NPV to over 90 and 80%, by reducing specificity to an average of 44% and 35%, respectively. The most sensitive combinations were NMP22, uCyt+ together with Cytology and FISH, and uCyt+ together with NMP22 (sensitivity for both combinations 98%). There was no further improvement when all four markers were combined. CONCLUSIONS: Combinations of tumour markers increased the sensitivity and NPV in the detection of recurrence of NMI TCC. A stepwise approach of tumour marker determination may be used to reduce the frequency of follow-up cystoscopies at a reasonable risk.