Current and possible future diagnostic methods for upper tract urothelial carcinoma

Sammanfattning: Background: Urothelial carcinoma is a type of cancer originating from the mucus membrane of the urinary tract. It most commonly occurs in the bladder but may also occur in the upper urinary tract, then called UTUC. UTUC is mainly detected in sexagenarians and older individuals. The gold standard of treatment has been nephroureterectomy (RNU), but this is a major surgery that carries the risk of significant peri- and postoperative morbidity. In addition, the associated decrease in kidney function affects whether these patients can receive adjuvant chemotherapy. Kidney saving surgery (KSS), such as focal laser ablation via ureterorenoscopy (URS), is increasingly recommended for selected patients, as several studies have reported similar disease-specific survival (DSS) outcomes in patients with low-risk UTUC, irrespective of the surgical method used (KSS vs. RNU). KSS has a significantly lower perioperative morbidity rate but a higher recurrence rate, so these patients require vigilant monitoring and follow-up. For more personalized treatment, it is crucial to distinguish patients with aggressive UTUC who require radical surgery and adjuvant treatment from those with nonaggressive disease who can safely benefit from KSS. This thesis comprises 4 studies on the current and possible future diagnostic methods for UTUC. Aim: The overall aim of these studies was to improve the diagnostic work-up, to aid treatment choices and, thus, to improve the survival of patients with UTUC. Study I aimed to assess the diagnostic accuracy of radiographic and endoscopic methods; study II aimed to evaluate the samples acquired during URS; study III aimed to investigate whether 3D imaging could be used in the diagnostic work-up; and study IV aimed to determine whether gene mutations in the tumour could be correlated to tumour stage, grade and long-term prognosis. Patients and methods: The studies are based on a prospective cohort of patients referred for diagnostic work-up or treatment of UTUC during the period 2005-2012. Study I also included patients without UTUC who were subject to investigation. Histopathological and cytological assessments were used as reference standards. Studies II-IV included only patients with UTUC, and RNU specimens were used as a reference standard. The statistical methods used were binary classification tests in studies I-II, descriptive statistics in study III, and principal component analysis, hierarchical clustering and analysis of variance in study IV. Results: Study I showed that multiphase CT urography (MCTU) had superior diagnostic accuracy compared to other imaging modalities and that MCTU and URS had different strengths. None of the methods were 100% accurate, emphasizing the importance of sample collection during URS. Study II found that the cytology results of in situ barbotage and histopathology each separately identified almost all cancers but were not always correct in grading the tumour. In addition, there was a significant correlation between tumour grade and ploidy in G1 and G3 tumours, aiding in the interpretation of ambiguous samples. In study III, 3D imaging could differentiate between superficial low-grade and invasive high- grade UTUC among 4 samples. Study IV showed that the mutational patterns in the tumour correlated with tumour stage, grade and long-term prognosis. Conclusion: None of the current diagnostic methods are 100% accurate; they all have different strengths and weaknesses. Our results showed that MCTU should be regarded as the preferred imaging modality (unlike at the time of the study) and that the diagnostic accuracy of cytology could be greatly improved if analysed in focal barbotage. The diagnostic accuracy can most likely be improved by using a combination of these modalities. New diagnostic methods, such as analysis of tumour gene mutations and 3D imaging, may add important information to the diagnostic process.