Short Communication For reprint orders, please contact: reprints@futuremedicine.com Restricted mean survival time as outcome measure in advanced urothelial bladder cancer: analysis of 4 clinical studies Melania Rivano* ,1 , Luca Cancanelli 2 , Lorenzo Di Spazio 3 , Marco Chiumente 4 , Daniele Mengato 5 & Andrea Messori 6 1 Clinical Oncology Pharmacy Department, A. Businco Hospital, 09121 Cagliari, Italy 2 Hospital Pharmacy Department, Azienda ULSS 2 Marca Trevigiana, 31033 Castelfranco Veneto, Italy 3 Hospital Pharmacy Department, S.Chiara Hospital, 38122 Trento, Italy 4 Scientifc Direction, Italian Society for Clinical Pharmacy & Therapeutics, Milan, Italy 5 Hospital Pharmacy Department, Bolzano General Hospital, 39100 Bolzano, Italy 6 HTA Unit, Regional Health Service, Florence, Italy *Author for correspondence: melania.rivano@gmail.com Background: The purpose of this study was to assess the effectiveness of immune checkpoint inhibitors (ICIs) in advanced urothelial carcinoma. Materials & methods: We selected seven cohorts of patients pub- lished in four clinical trials. The restricted mean survival time (RMST) was used to analyze survival curves, perform the comparisons and rank the treatments based on their effectiveness. The performance of RMST was compared with that of a network meta-analysis. Results: Three ICIs, vinfunine and best standard care, given as second line, were examined. ICIs signifcantly improved overall survival compared with best stan- dard care. However, the survival gain was quite limited (up to 2.27 months). Post hoc pairwise comparisons were calculated. Conclusion: Our results summarized the effcacy of these treatments and confrmed the good methodological performance of RMST. First draft submitted: 29 May 2020; Accepted for publication: 22 October 2020; Published online: 5 November 2020 Keywords: advanced urothelial bladder cancer • hazard ratio • immune checkpoint inhibitors • Kaplan–Meier curves • milestone • post hoc pairwise comparisons • ranking • restricted mean survival time, NETMA • survival curves Advanced urothelial carcinoma (UC) has a poor prognosis and expected survival is generally less than 12 months [1]. First-line cisplatin-based chemotherapy can improve overall survival (OS) [2,3] but most patients have disease progression and substantial toxicities [4,5]. Vinflunine and taxanes are commonly used as second-line single-agent chemotherapy, but these drugs do not achieve any substantial survival benefit [6,7]. In the past few years, checkpoint inhibitors have revolutionized the treatment of metastatic UC. Monoclonal antibodies against PD-1 receptor and its ligands (PD-L1 and PD-L2) have shown survival benefits and manageable safety profile in patients with disease progression [8]. Rassy and coworkers have performed a network meta-analysis (NETMA) to determine which of the approved immune checkpoint inhibitors (ICIs) is the most effective in metastatic urothelial bladder cancer [9]. The NETMA was conducted according to the current methodological standards of this technique; treatments were ranked in terms of effectiveness according to their values of hazard ratio (HR) [10]. In general, in a clinical study where patients are exposed to the risk of experiencing a negative event over time (time-to-event design), the Kaplan–Meier algorithm has the purpose of combining the so-called censored patients (i.e., patients with termination of their follow-up without event occurrence) with the patients experiencing the event in order to construct the graph of the time-to-event curve (which in oncology is generally an OS curve or a progression-free survival curve). A survival curve can be interpreted from a standalone perspective or in the framework of a comparative design (which typically takes place when two survival curves are generated within a randomized controlled trial). In the first case (standalone analysis), the median is traditionally the most common parameter; the median, however, has an important disadvantage in that it cannot be computed when too few events Immunotherapy (Epub ahead of print) ISSN 1750-743X 10.2217/imt-2020-0160 C  2020 Future Medicine Ltd