Stesura Seveso Archivio Italiano di Urologia e Andrologia 2023; 95, 1 ORIGINAL PAPER established, yet the majority will not be diagnosed with urological cancer and the cause will be attributed to tran- sient benign physiological conditions, including UTI, UL or BPE. The extent of investigation and timing of this investigation in not well defined. Taking in consideration the finitude of available resources, it would be important to prioritize the patients with more risk of suffering from urological cancer, to quickly diag- nose, evaluate and treat. There are known risks factors to urological cancer, such as smoking history, exposure to occupational chemical and dyes or pelvic radiation, but there is no algorithm available that allow the physicians to estimate an approximate risk of bladder malignancy. With our study we intend to help physicians assessing the likelihood of urologic malignancy based in general clinic, laboratory, and image data. The aim of the study was to identify predictive factors of bladder cancer among patients presenting with hema- turia. PATIENTS AND METHODS Patients eligible for this study were referred to urology appointment by the General Practitioner (GP) or from de emergency department because of hematuria, between January 1, 2017 and December 31, 2019. Patients otherwise asymptomatic who were referred due to incidental imaging findings suspicious of bladder can- cer or upper tract urothelial carcinoma were excluded. All patients had cystoscopy and upper tract imaging. Main outcome of interest was bladder cancer diagnosis, defined as presence of urothelial carcinoma in pathologi- cal study after transurethral bladder resection (TURB) according to TNM WHOW tumor classification and European Association of Urology risk classification. Variables such as demographics, imaging and clinical fac- tors were evaluated (including age, gender, smoker sta- tus, anticoagulation or antiaggregating drug use, previous pelvic irradiation, number of hematuria episodes, pres- ence of lower urinary tract symptoms (LUTS), urine culture positivity, back pain, history of urolithiasis, fever, ultra- sound and cystoscopy results). The demographic and clinical features are shown in Table 1. Statistical analysis Descriptive statistics were calculated for all patients Introduction: The presence of blood in the urine should be promptly investigated to rule out urological malignancies, bladder cancer being the most frequent. Given its frequency among general population and the lack of unlimited health resources in an era of cost-effec- tiveness, it is important to prioritize patients with higher risk of malignancy. Objectives: To identify predictive factors of bladder cancer among patients presenting with hematuria. Patients and Methods: We retrospectively reviewed 296 cases referred to our department for hematuria. We evaluated differ- ent demographic, clinical and ultrasound features to uncover possible associations with diagnosis of bladder cancer in those patients, to estimate the individual risk of being diagnosed with bladder cancer during the investigation of hematuria. Results: A total of 296 patients were studied for hematuria between January 1, 2017 and December 31, 2019, 23.6% of those having ultimately bladder cancer confirmed after transurethral resection. Older age, male gender (OR 2.727, p = 0.069), a history of smoking (OR 3.84, p < 0.05), recurrent hematuria (OR 3.396, p < 0.05) and positive ultrasound exam for bladder cancer (OR 30.423, p < 0.05) were identified as predictors of bladder cancer in patients with hematuria. Conclusions: This study suggests that it is possible to reliably estimate the risk of bladder cancer in patients with hematuria, using clinical and imaging data to help defining who should be investigated first and in whom the investigation could be post- poned. KEY WORDS: Bladder cancer; Hematuria; Smoking; Ultrasound; Male gender. Submitted 20 November 2022; Accepted 22 November 2022 INTRODUCTION Hematuria is defined as presence of blood in urine. It can either be microscopic (only detected in urinalysis and with variable definition among regions) or macroscopic. Hematuria is one of the most frequent causes of referral to emergency department or urology appointment. There are many etiologies, most being benign, like urinary tract infection (UTI), benign prostatic enlargement (BPe) or urolithiasis (UL) but the existence of a subjacent urologi- cal cancer, most often bladder cancer, must be dismissed. The need to investigate almost all patients who present with macroscopic and microscopic hematuria is well Predicting bladder cancer risk in patients with hematuria. A single-centre retrospective study Roberto Jarimba 1, 2, Vasco Quaresma 1, João Pedroso Lima 1, 2, Miguel Eliseu 1, 2, Edgar Tavares da Silva 1, 2, Pedro Moreira 1, Arnaldo Figueiredo 1, 2 1 Urology and Renal Transplantation Department, Centro Hospitalar e Universitário de Coimbra, Coimbra, Portugal; 2 Faculty of Medicine, University of Coimbra, Portugal. DOI: 10.4081/aiua.2023.11026 Summary Archivio Italiano di Urologia e Andrologia 2023; 95, 1 R. Jarimba, V. Quaresma, J. Pedroso Lima, M. Eliseu, E. Tavares da Silva, P. Moreira, A. Figueiredo included in the present study. The crude association between each individual categorical covariate and bladder cancer diagnosis was accessed by Chi-square and binary logistic regression. All analysis were conducted using IBM SPSS statistics ver- sion 26. All comparisons were made using 2-sided tests, with p < 0.05 considered statistically significant. RESULTS A total of 296 patients were studied for hematuria between January 1, 2017 and December 31, 2019. Overall, 23.6% of those patients were diagnosed with bladder cancer after TURB results. Clinical and imaging predictors were found to be statisti- cally significative. Some are well known risk factors for bladder cancer: patients diagnosed with bladder cancer were older (73.74 vs 66.8 years, p < 0.05); men had roughly 2.5 times (OR: 2.519, p < 0.05) more risk of being diagnosed with bladder cancer; patients ever exposed to tobacco had 3.8 times (OR 3.852, p < 0.05) more risk of bladder cancer. The number of hematuria episodes and urine microbiol- ogy seem to have a predictive value for the diagnosis of bladder cancer, in univariate analysis. Patients with mul- tiple episodes of hematuria have a higher risk of being diagnosed with bladder cancer (OR: 2.093, p < 0.05) vs patients with a single episode. An identification of bacte- ria in the urine microbiology was inversely correlated with bladder cancer diagnosis (OR: 0.737, p < 0.05); a total of 29 patients had a positive urine culture, none of those were found to have bladder cancer. Ultrasound evaluation had a sensitivity and specificity for bladder cancer of 71% and 87%, respectively. While there was a bladder cancer suspicion at ultrasound exam- ination in 70% of the patients with subsequent diagnosis of bladder cancer, only 16% of patients diagnosed with bladder cancer had a normal ultrasound evaluation. Those patients were older (OR: 1.084, p = 0.011) and had more frequently a smoking history (OR: 4.503, p = 0.048). The second most common finding in the ultra- sound exam among patients diagnosed with bladder can- cer was unilateral hydronephrosis, found in 6.6% of those patients. In the multivariate analysis, using a binary logistic regres- sion, age, tobacco exposure, multiple episodes of hema- turia and a positive ultrasound were found to be correlat- ed with risk of bladder cancer. The calculated model showed an accuracy of 86.5%, with a sensitivity of 64.9% and a specificity of 93.1%. Twelve (15.4%) out of 78 patients with suspected blad- der cancer at cystoscopy had negative histology for blad- der cancer after TURB. Among patients studied for microscopic hematuria (10.6% of the total), only 6% were diagnosed with blad- der cancer. All patients with microscopic hematuria diag- nosed with bladder cancer were male older than 40 years of age and had ultrasound positive for bladder. Only 1.4% and 0.3% of the patients were diagnosed with upper tract urothelial carcinoma and renal cell carcino- ma, respectively. Among patients with a diagnosis other than bladder can- cer, 25% had prostatic bleeding, 15.9% UTI-related hematuria and 8.1% were diagnosed with urolithiasis. Results are shown in Table 2. Table 1. Demographic and clinical features of patients referred because of hematuria. Total (n = 296) Non-BC (n = 226) BC (n = 70) Age (years) 68.45 ± 15.94 66.81 ± 15.92 73.74 ± 14.92 Gender Female 95 (32.1%) 84 (37.2%) 11 (15.7%) Male 201 (67.9%) 142 (62.8%) 59 (84.3%) Tobacco Never user 225 (76%) 184 (81.8%) 41 (58.6%) Ever user 48 (16.2%) 28 (12.4%) 20 (28.6%) Not known 23 (7.8%) 14 (6.2%) 9 (12.9%) Hematuria episodes > 1x 147 (49.7%) 103 (45.6%) 44 (62.9%) 1x 118 (39.9%) 98 (43.4%) 20 (28.6%) Microscopic hematuria 31 (10.5%) 25 (11.1%) 6 (8.6%) Urine culture Negative 216 (73.2%) 161 (71.6%) 55 (78.6%) Positive 29 (9.8%) 29 (12.9%) 0 (0%) Not known 50 (16.9%) 35 (15.6%) 15 (21.4%) Ultrasound Normal 126 (42.9%) 115 (42.9%) 11 (15.9%) Suspicious 76 (25.9%) 27 (12%) 49 (71%) Urolithiasis 31 (10.5%) 28 (12.4%) 2 (4.3%) Renal mass 12 (4.1%) 12 (5.3%) 0 Bladder wall thickness 8 (2.7%) 8 (3.6%) 0 Hydronephrosis 13 (4.4%) 9 (4%) 4 (5.8%) Prostate enlargement 18 (61%) 18 (8%) 0 Vesical blood clot 1 (1.7%) 1 (0.4%) 0 Bladder stone 3 (1%) 3 (1.3%) 0 Suspicion of UTUC 1 (0.3%) 1 (0.4%) 0 Anticoagulation/anti aggregation No 180 (60.8%) 135 (59.7%) 45 (64.3%) Yes 116 (39.2%) 91 (40.3%) 25 (35.7%) Pelvic radiation No 281 (95.3%) 215 (95.6%) 66 (94.3%) Yes 14 (4.7%) 10 (4.4%) 4 (5.7%) UTI suspicion No 224 (75.7%) 167 (73.9%) 57 (81.4%) Yes 72 (24.3%) 59 (26.1%) 13 (18.6%) Back pain No 249 (84.7%) 187 (83.1%) 62 (89.9%) Yes 45 (15.3%) 38 (16.9%) 7 (10.1%) Fever No 286 (96.6%) 216 (95.6%) 70 (100%) Yes 10 (3.4%) 10 (4.4%) 0 Previous LUTS No 183 (61.8%) 139 (61.5%) 44 (62.9%) Yes 104 (35.1%) 79 (35%) 25 (35.7%) Not known 9 (3%) 8 (3.5%) 1 (1.4%) Urolithiasis history No 268 (90.5%) 201 (88.9%) 67 (95.7%) Yes 28 (9.5%) 25 (11.1%) 3 (4.3%) Cystoscopy Normal 130 (43.9%) 130 (57.5%) 0 Dubious 25 (8.4%) 21 (9.3%) 4 (5.7%) Suspicious 78 (26.4%) 12 (5.3%) 78 (94.3%) Prostate enlargement 50 (16.9%) 50 (22.1%) 0 Bladder Trabeculation 6 (2%) 6 (2.7%) 0 Urethral stenosis 3 (1%) 3 (1.3%) 0 Bladder stone 3 (1%) 3 (1.3%) 0 UTI: Urinary tract infection; LUTS: Lower urinary tract symptoms. Archivio Italiano di Urologia e Andrologia 2023; 95, 1 Bladder cancer and hematuria DISCUSSION Our study, as others (1), highlights the need for investigat- ing almost every patient presenting with hematuria. In our cohort, the overall probability of being diagnosed with bladder cancer throughout the investigation of hematuria was 23.6%. NICE states that a signal or a symptom associ- ated with ≥ 3% risk of cancer should prompt referral for diagnostic test (2) and many patients want that investiga- tion be made for a symptom associated ≥ 1% risk of cancer (3). American Urology Association (AUA) recommends that all patients with visible hematuria and patients with micro- scopic hematuria (≥ 3 red blood cells/high-power field), aged ≥ 35 years, should be investigated (4). In an era of relative lack of health resources when com- pared with demands, it is of the most importance to pri- oritize patients regarding diagnostic procedures (i.e. cys- toscopy, CT scan). This way, patients with higher proba- bility of bladder cancer can be prioritized to receive said tests, and patients with lower probability being safely postponed (but studied nonetheless). We were not able to define thresholds for strictly risk- based categories, but the data allow us to roughly esti- mate the risk of bladder cancer in a single patient. Some predictors for bladder cancer were uncovered, as age, gender, tobacco exposure history, number of episodes of visible hematuria and ultrasound evaluation. Older patients, male (OR 2.843, p = 0.061), a history of smok- ing (OR 3.852, p < 0.05), with recurrent hematuria (OR 3.471, p < 0.05) and positive ultrasound exam for blad- der cancer (OR 31.663, p < 0.05) are at highest risk and should be investigated promptly. Univariate analysis showed that negative urine culture was a risk factor for bladder cancer, but not in multivariate analysis. None of patients with a positive urine culture at the moment of hematuria had a subsequent diagnosis of bladder cancer. In our study, ultrasound had sensibility of 71% and specificity of 87% for bladder cancer. It can be a good screening test but cannot be an alternative to cystoscopy, because 16% of patients diagnosed with bladder cancer had a normal ultrasound evaluation. In patients investigated for non-visible hematuria the diagnostic rate of bladder cancer was 8.6% and all of them had a positive ultrasound. According to our results and in agreement with National Board of Health and Welfare of Sweden (5), it seems plausible that patients with nonvisible hematuria with a negative ultrasound evalua- tion are at low risk of bladder cancer and the investiga- tion can be postponed. We uncovered some new clinical predictors for bladder cancer in patients with hematuria, like recurrent hema- turia and a negative urine culture. Our study has some limitations. It is a retrospective study with a relatively small number of patients that limits the statistically power of the analysis. Urinary cytology was not included in co-variates, because, in our department most patients collect a bladder washing during cys- toscopy procedure. The incidence of bladder cancer in our cohort is probably higher than its real incidence in overall patients with hematuria because it represents the detection rate in a secondary care setting. This study suggests that it is possible to reliably estimate the risk of bladder cancer in patients with hematuria, using clinical and imaging data to manage available healthcare resources without compromising the standard of care. REFERENCES 1. Tan WS, Feber A, Sarpong R, et al. Who Should Be Investigated for Haematuria? Results of a Contemporary Prospective Observational Study of 3556 Patients. Eur Urol. 2018; 74:10-4. 2. National Collaborating Centre for Cancer (UK). Suspected Cancer: Recognition and Referral. London: National Institute for Health and Care Excellence (NICE); 2015 Jun. 3. Banks J, Hollinghurst S, Bigwood L, et al. Preferences for cancer investigation: a vignette-based study of primary-care attendees. Lancet Oncol. 2014; 15:232-40. 4. Davis R, Jones JS, Barocas DA, et al. American Urological Association. Diagnosis, evaluation and follow-up of asymptomatic microhematuria (AMH) in adults: AUA guideline. J Urol. 2012; 188 (6 Suppl):2473-81. 5. Malmström P-U. Time to abandon testing for microscopic haema- turia in adults? BMJ. 2003; 326:813-5. Table 2. Univariate and multivariate analysis for predictors of bladder cancer in patients with hematuria. Predictors of bladder cancer for 265 visible hematuria referrals Univariate Multivariate OR CI OR CI Age (years) - - 1.047 * 1.012-1.080 Gender (0 = f, 1 = m) 2.519 * 1.236-5.131 2.843 0.952-8.489 Ever smoker 2.876 * 1.435-5.764 3.852 * 1.301-11.405 Hematuria episodes (0 = 1, 1 = > 1x) 2.093 * 1.153-3.801 3.471 * 1.435-8.395 Urine culture (0 = neg, 1 = pos) 0.737 * 0.677-0.802 0.496 0.698-2.100 Ultrasound suspicion 24.425 * 11.483-51.955 31.633 * 12.867-77.772 Test �2 d� P Overall model evaluation R square Nagelkerke 0.545 Goodness-of-fit test Hosmer & Lemeshow 8.531 8 0.383 Uf: female; m: male; neg: negative; pos: positive. * p < 0.05. Correspondence Roberto Jarimba, MD (Corresponding Author) robertojarimba@chuc.min-saude.pt Vasco Quaresma, MD vpdquaresma@gmail.com Miguel Eliseu, MD mgl.nobre@gmail.com João Pedroso Lima, MD Joaopedrosolima@gmail.com Edgar Tavares da Silva, MD edsilva.elv@gmail.com Pedro Moreira, MD pedronetomoreira@gmail.com Arnaldo Figueiredo, MD, PhD ajcfigueiredo@gmail.com Serviço de Urologia, Centro Hospitalar e Universitário de Coimbra Rua Professor Mota Pinto 3004-561, Coimbra (Portugal) Conflict of interest: The authors declare no potential conflict of interest.