1
|
Menold HS, Wieland VLS, Haney CM, Uysal D, Wessels F, Cacciamani GC, Michel MS, Seide S, Kowalewski KF. Machine learning enables automated screening for systematic reviews and meta-analysis in urology. World J Urol 2024; 42:396. [PMID: 38985296 DOI: 10.1007/s00345-024-05078-y] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/08/2024] [Accepted: 05/23/2024] [Indexed: 07/11/2024] Open
Abstract
PURPOSE To investigate and implement semiautomated screening for meta-analyses (MA) in urology under consideration of class imbalance. METHODS Machine learning algorithms were trained on data from three MA with detailed information of the screening process. Different methods to account for class imbalance (Sampling (up- and downsampling, weighting and cost-sensitive learning), thresholding) were implemented in different machine learning (ML) algorithms (Random Forest, Logistic Regression with Elastic Net Regularization, Support Vector Machines). Models were optimized for sensitivity. Besides metrics such as specificity, receiver operating curves, total missed studies, and work saved over sampling were calculated. RESULTS During training, models trained after downsampling achieved the best results consistently among all algorithms. Computing time ranged between 251 and 5834 s. However, when evaluated on the final test data set, the weighting approach performed best. In addition, thresholding helped to improve results as compared to the standard of 0.5. However, due to heterogeneity of results no clear recommendation can be made for a universal sample size. Misses of relevant studies were 0 for the optimized models except for one review. CONCLUSION It will be necessary to design a holistic methodology that implements the presented methods in a practical manner, but also takes into account other algorithms and the most sophisticated methods for text preprocessing. In addition, the different methods of a cost-sensitive learning approach can be the subject of further investigations.
Collapse
Affiliation(s)
- H S Menold
- Department of Urology and Urological Surgery, University Medical Center Mannheim, University of Heidelberg, Theodor-Kutzer-Ufer 1-3, 68167, Mannheim, Germany
| | - V L S Wieland
- Department of Urology and Urological Surgery, University Medical Center Mannheim, University of Heidelberg, Theodor-Kutzer-Ufer 1-3, 68167, Mannheim, Germany
| | - C M Haney
- Department of Urology, University of Leipzig, Leipzig, Germany
- Intelligent Systems and Robotics in Urology (ISRU), DKFZ Hector Cancer Institute at the University Medical Center Mannheim, Mannheim, Germany
- German Cancer Research Center (DKFZ), Heidelberg, Germany
| | - D Uysal
- Department of Urology and Urological Surgery, University Medical Center Mannheim, University of Heidelberg, Theodor-Kutzer-Ufer 1-3, 68167, Mannheim, Germany
| | - F Wessels
- Department of Urology and Urological Surgery, University Medical Center Mannheim, University of Heidelberg, Theodor-Kutzer-Ufer 1-3, 68167, Mannheim, Germany
| | - G C Cacciamani
- USC Institute of Urology, University of Southern California, ©, Los Angeles, CA, USA
| | - M S Michel
- Department of Urology and Urological Surgery, University Medical Center Mannheim, University of Heidelberg, Theodor-Kutzer-Ufer 1-3, 68167, Mannheim, Germany
| | - S Seide
- Böhringer Ingelheim, Ingelheim am Rhein,, Germany
| | - K F Kowalewski
- Department of Urology and Urological Surgery, University Medical Center Mannheim, University of Heidelberg, Theodor-Kutzer-Ufer 1-3, 68167, Mannheim, Germany.
- Intelligent Systems and Robotics in Urology (ISRU), DKFZ Hector Cancer Institute at the University Medical Center Mannheim, Mannheim, Germany.
- German Cancer Research Center (DKFZ), Heidelberg, Germany.
| |
Collapse
|
2
|
Xu J, Bai J, Pan H, Zhou Z. Impact of perioperative blood transfusion on prognosis after nephrectomy in patients with renal cell carcinoma: A meta-analysis and systematic review. Medicine (Baltimore) 2024; 103:e37646. [PMID: 38579099 PMCID: PMC10994588 DOI: 10.1097/md.0000000000037646] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 07/23/2023] [Accepted: 02/27/2024] [Indexed: 04/07/2024] Open
Abstract
BACKGROUND Perioperative blood transfusion (PBT) has been associated with worse prognosis in several malignancies. For renal cell carcinoma (RCC), the effect of PBT is still debated. OBJECTIVE To evaluate the impact of PBT on prognosis after nephrectomy in patients with RCC. METHODS This study is A systematic review and meta-analysis of published article data (PRISMA protocol) for literature related to PBT and RCC through extensive search of EMBASE, Medline via PubMed, Web of Science and Cochrane Library, language limited to English, with no time constraint until May 20, 2022. We pooled the results of multivariable cox regression analyses from each study, with subgroup analyses by dose and timing of transfusion. All analyses were done using Stata14. RESULTS A total of 12 studies involving 27,683 participants were included. Our meta-analysis pooled the results of multivariable cox regression analysis in each study, showing that PBT is associated with higher overall Mortality (OM; hazard ratio [HR] = 1.34, 1.23-1.44), cancer-specific mortality (CSM; HR = 1.35, 1.20-1.51), and disease recurrence (HR = 1.54, 1.18-1.89). when only patients with nonmetastatic RCC were included, PBT was still associated with higher OM (HR = 1.29, 1.11-1.47) and disease recurrence (HR = 1.58, 1.18-1.98), but the association with CSM (HR = 1.26, 0.99-1.52) was not statistically significant. In subgroup analysis by transfusion dose, small (1-2) units of PBT were not associated with CSM (HR = 1.84, 0.95-2.73), but large (≥3) units were associated with higher CSM (HR = 2.98, 1.74-4.22) and disease recurrence (HR = 1.99, 1.31-2.67). Each additional unit of PBT resulted in a higher CSM (HR = 1.07, 1.04-1.10). In subgroup analysis by transfusion timing, intraoperative transfusion was associated with higher CSM and disease recurrence, but postoperative transfusion was not. CONCLUSIONS PBT is associated with higher OM, CSM and disease recurrence. This adverse effect seems to be particularly significant in high-dose intraoperative transfusion. It is necessary to limit the overuse of PBT, especially high-dose intraoperative transfusion, in order to improve the prognosis of patients undergoing nephrectomy for RCC.
Collapse
Affiliation(s)
- Jiangnan Xu
- Department of Urology, Yancheng First Hospital, Affiliated Hospital of Nanjing University Medical School, The First people’s Hospital of Yancheng, Yancheng, China
| | - Jinming Bai
- Department of Urology, Yancheng First Hospital, Affiliated Hospital of Nanjing University Medical School, The First people’s Hospital of Yancheng, Yancheng, China
| | - Huixing Pan
- Department of Urology, Yancheng First Hospital, Affiliated Hospital of Nanjing University Medical School, The First people’s Hospital of Yancheng, Yancheng, China
| | - Zhengdong Zhou
- Department of Urology, Yancheng First Hospital, Affiliated Hospital of Nanjing University Medical School, The First people’s Hospital of Yancheng, Yancheng, China
| |
Collapse
|
3
|
Ebner B, Fleckenstein F, Volz Y, Eismann L, Hermans J, Buchner A, Enzinger B, Weinhold P, Wichmann C, Stief CG, Humpe A, Pyrgidis N, Schulz GB. Oncological impact of perioperative blood transfusion in bladder cancer patients undergoing radical cystectomy: Do we need to consider storage time of blood units, donor age, or gender matching? Transfusion 2024; 64:29-38. [PMID: 38053445 DOI: 10.1111/trf.17618] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/12/2023] [Revised: 10/08/2023] [Accepted: 10/18/2023] [Indexed: 12/07/2023]
Abstract
BACKGROUND The oncological impact of perioperative blood transfusions (PBTs) of patients undergoing radical cystectomy (RC) because of bladder cancer (BCa) has been a controversial topic discussed in recent years. The main cause for the contradictory findings of existing studies might be the missing consideration of the storage time of red blood cell units (BUs), donor age, and gender matching. STUDY DESIGN AND METHODS We retrospectively analyzed BCa patients who underwent RC in our department between 2004 and 2021. We excluded patients receiving BUs before RC, >10 BUs, or RC in a palliative setting. We assessed the effect of blood donor characteristics and storage time on overall survival (OS) and cancer-specific survival (CSS) through univariate and multivariable Cox regression analysis. We also performed a propensity score matching with patients who received BUs and patients who did not on a 1:1 ratio. RESULTS We screened 1692 patients and included 676 patients for the propensity score matching. In the multivariable analysis, PBT was independently associated with worse OS and CSS (p < .001). Postoperative transfusions were associated with better OS (p = .004) and CSS (p = .008) compared to intraoperative or mixed transfusions. However, there was no influence of blood donor age, storage time, or gender matching on prognosis. DISCUSSION In our study of BCa patients undergoing RC, we demonstrate that PBT, especially if administered intraoperatively, is an independent risk factor for a worse prognosis. However, storage time, donor age, or gender matching did not negatively affect oncological outcomes. Therefore, the specific selection of blood products does not promise any benefits.
Collapse
Affiliation(s)
- Benedikt Ebner
- Department of Urology, University Hospital, LMU Munich, Munich, Germany
| | | | - Yannic Volz
- Department of Urology, University Hospital, LMU Munich, Munich, Germany
| | - Lennert Eismann
- Department of Urology, University Hospital, LMU Munich, Munich, Germany
| | - Julian Hermans
- Department of Urology, University Hospital, LMU Munich, Munich, Germany
| | - Alexander Buchner
- Department of Urology, University Hospital, LMU Munich, Munich, Germany
| | - Benazir Enzinger
- Department of Urology, University Hospital, LMU Munich, Munich, Germany
| | - Philipp Weinhold
- Department of Urology, University Hospital, LMU Munich, Munich, Germany
| | - Christian Wichmann
- Division of Transfusion Medicine, Cell Therapeutics and Haemostaseology, University Hospital, LMU Munich, Munich, Germany
| | | | - Andreas Humpe
- Division of Transfusion Medicine, Cell Therapeutics and Haemostaseology, University Hospital, LMU Munich, Munich, Germany
| | - Nikolaos Pyrgidis
- Department of Urology, University Hospital, LMU Munich, Munich, Germany
| | | |
Collapse
|
4
|
Chu Z, Cushway T, Wong M, Lim KX, Peh WM, Ng CT, Lim WY, Ong SGK, Tey TT, Foo FJ, Koh FH. Incidence and predictors of hypophosphataemia after ferric carboxymaltose use-A 3-year experience from a single institution in Singapore. Br J Haematol 2023; 202:1199-1204. [PMID: 37455143 DOI: 10.1111/bjh.18979] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/17/2023] [Revised: 06/26/2023] [Accepted: 07/04/2023] [Indexed: 07/18/2023]
Abstract
Ferric carboxymaltose (FCM) administration helps reduce transfusion requirements in the perioperative situation, which improves patient outcomes and reduces healthcare costs. However, there is increasing evidence of hypophosphataemia after FCM use. We aim to determine the incidence of hypophosphataemia after FCM administration and elucidate potential biochemical factors associated with the development of subsequent hypophosphataemia. A retrospective review of anonymised data of all FCM administrations in a single institution was conducted from August 2018 to August 2021. Each unique FCM dose administered was examined to assess its effect on Hb and serum phosphate levels within the subsequent 28 days from each FCM administration. Phosphate levels were repeatedly measured within the 28-day interval and the lowest phosphate level within that period was determined. Patients' serum phosphate levels within 28 days of FCM administration were compared against normal serum phosphate levels within 2 weeks before FCM administration. The odds ratios of various pre-FCM serum markers were calculated to elucidate potential biochemical predictors of post-FCM hypophosphataemia. In 3 years, a total of 1296 doses of FCM were administered to 1069 patients. The mean improvement in Hb was 2.45 g/dL (SD = 1.94) within 28 days of FCM administration, with the mean time taken to peak Hb levels being 6.3 days (SD = 8.63), which is earlier than expected, but was observed in this study and hence reported. The incidence of hypophosphataemia <0.8 mmol/L was 22.7% (n = 186), and <0.4 mmol/L was 1.6% (n = 9). This figure is lower than the numbers reported in previously published meta-analyses given that routine checks of serum phosphate levels were not conducted initially and hence could possibly be higher. The odds of developing hypophosphataemia (<0.8 mmol/L) were 27.7 (CI: 17.3-44.2, p < 0.0001) if baseline serum phosphate was less than 1 mmol/L. The odds of developing hypophosphataemia (<0.8 mmol/L) were 1.3 (CI: 1.08-1.59, p < 0.01) if the change in Hb levels observed after FCM administration were more than 4 g/dL. Hypophosphataemia after FCM administration is significant and FCM should be used by clinicians with caution.
Collapse
Affiliation(s)
- Zachary Chu
- Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore
| | - Tim Cushway
- The Integrative Medical Centre by The Iron Suites, Singapore, Singapore
| | - Marc Wong
- Department of Internal Medicine, Sengkang General Hospital, Singapore, Singapore
| | - Kai-Xiong Lim
- Department of Internal Medicine, Sengkang General Hospital, Singapore, Singapore
| | - Wee-Ming Peh
- Department of Internal Medicine, Sengkang General Hospital, Singapore, Singapore
| | - Choong-Tatt Ng
- Department of Internal Medicine, Sengkang General Hospital, Singapore, Singapore
| | - Wan-Yen Lim
- Department of Anaesthesiology, Sengkang General Hospital, Singapore, Singapore
| | - Sharon G K Ong
- Department of Anaesthesiology, Sengkang General Hospital, Singapore, Singapore
| | - Tze-Tong Tey
- Department of Gastroenterology, Sengkang General Hospital, Singapore, Singapore
| | - Fung-Joon Foo
- Department of General Surgery, Sengkang General Hospital, Singapore, Singapore
| | - Frederick H Koh
- Department of General Surgery, Sengkang General Hospital, Singapore, Singapore
| |
Collapse
|
5
|
Ma SY, An Y, Sun JX, Xu MY, Liu CQ, Xu JZ, Zhong XY, Zeng N, He HD, Xia QD, Wang SG. The effect of different timing of blood transfusion on oncological outcomes of patients undergoing radical cystectomy for bladder cancer: a systematic review and meta-analysis. Front Oncol 2023; 13:1223592. [PMID: 37719020 PMCID: PMC10499617 DOI: 10.3389/fonc.2023.1223592] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/16/2023] [Accepted: 08/16/2023] [Indexed: 09/19/2023] Open
Abstract
Highlights This meta-analysis and systematic review aim to analyze the association between BT and oncological outcomes of patients undergoing RC for bladder cancer, and tries to find out whether the timing of blood transfusion could also have an effect on this relationship. A total of 20 retrospective studies from online databases and other sources are identified and enrolled in this study. The results show that BT administration during RC operation or perioperative period is significantly associated with worse oncological outcomes including ACM, CSM and DR. Background Bladder cancer is one of the most common urological malignancies. Radical cystectomy (RC) remains the main treatment for localized muscle-invasive bladder cancer (MIBC) or high-grade non-muscle-invasive bladder cancer (NMIBC). In the process of RC, the administration of blood transfusion (BT) is sometimes needed, however, it may cause transfusion-related complications or lead to worse oncological outcomes. This meta-analysis and systematic review aims to give a comprehensive insight into the association between BT and oncological outcomes of patients undergoing RC, and tries to find out whether the timing of blood transfusion could also have an impact on this association. Methods This systematic review and meta-analysis were carried out according to the PRISMA 2020 reporting guideline. We have searched four bibliographic databases including PubMed (Medline), EMBASE, Cochrane Library, and Web of Science with no language limitation. Studies investigating the association between BT and oncological outcomes of patients undergoing RC are identified and included in this research from inception through March 20, 2023. This research calculates the pooled hazard ratios (pHR) and 95% confidence intervals (95% CI) of all-cause mortality (ACM), cancer-specific mortality (CSM) and disease recurrence (DR) using Random Effects models or Fixed Effects models. Subgroup analyses stratified by parameters such as timing of transfusion are also conducted. This meta-analysis was registered with PROSPERO, CRD42022381656. Results A total of 20 retrospective studies from online databases and other sources are identified and enrolled in this study. Results show that blood transfusion significantly increased the risks for ACM (HR = 1.33, 95% CI: 1.23-1.44), CSM (HR = 1.25, 95% CI: 1.15 - 1.35) and DR (HR = 1.26, 95% CI: 1.15 - 1.38). However, when stratified by the timing of BT, we find that only intraoperative and perioperative transfusion significantly increased in risks for worse prognosis, while postoperative transfusion raised none of the risks of ACM (HR = 1.26, 95% CI: 0.92-1.73), CSM (HR = 1.08, 95% CI: 0.93-1.26) nor DR (HR = 1.08, 95% CI: 0.90-1.29) significantly. Conclusion BT administration during RC operation or perioperative period is significantly associated with worse oncological outcomes including ACM, CSM and DR. Clinicians should consider carefully when deciding to administrate BT to patients undergoing RC and carry out according to current guidelines.
Collapse
Affiliation(s)
| | | | | | | | | | | | | | | | | | - Qi-Dong Xia
- *Correspondence: Qi-Dong Xia, ; Shao-Gang Wang,
| | - Shao-Gang Wang
- Department and Institute of Urology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
| |
Collapse
|
6
|
Wessels F, Bußoff I, Adam S, Kowalewski KF, Neuberger M, Nuhn P, Michel MS, Kriegmair MC. Comorbidity Scores and Machine Learning Methods Can Improve Risk Assessment in Radical Cystectomy for Bladder Cancer. Bladder Cancer 2022; 8:155-163. [PMID: 38993365 PMCID: PMC11181714 DOI: 10.3233/blc-211640] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/24/2021] [Accepted: 04/05/2022] [Indexed: 11/15/2022]
Abstract
BACKGROUND Pre-operative risk assessment in radical cystectomy (RC) is an ongoing challenge especially in elderly patients. OBJECTIVE To evaluate the ability of comorbidity indices and their combination with clinical parameters in machine learning models to predict mortality and morbidity after RC. METHODS In 392 patients who underwent open RC, complication and mortality rates were reported. The predictive values of the age-adjusted Charlson Comorbidity index (aCCI), the Elixhauser Index (EI), the Physical Status Classification System (ASA) and Gagne's combined comorbidity Index (GCI) were evaluated using regression analyses. Various machine learning models (Gaussian naïve bayes, logistic regression, neural net, decision tree, random forest) were additionally investigated. RESULTS The aCCI, ASA and GCI showed significant results for the prediction of complications (χ2 = 8.8, p < 0.01, χ2 = 15.7, p < 0.01 and χ2 = 4.6, p = 0.03) and mortality (χ2 = 21.1, p < 0.01, χ2 = 25.8, p < 0.01 and χ2 = 2.4, p = 0.04) after RC while the EI showed no significant prediction. However, areas under receiver characteristic curves (AUROCs) revealed good performance only for the prediction of mortality by the aCCI and ASA (0.81 and 0.78, CGI 0.63) while the prediction of complications was poor (aCCI 0.6, ASA 0.63, CGI 0.58). The combination of ASA, age, body mass index and sex in machine learning models showed a better prediction. Gaussian naïve bayes (0.79) and logistic regression (0.76) showed the best performance using a hold-out test set. CONCLUSIONS The ASA and aCCI show good prediction of mortality after RC but fail predicting complications accurately. Here, the combination of comorbidity indices and clinical parameters in machine learning models seems promising.
Collapse
Affiliation(s)
- Frederik Wessels
- Department of Urology and Urological Surgery, University Medical Center Mannheim, Medical Faculty of Heidelberg University, Mannheim, Germany
| | - Isabelle Bußoff
- Medical Faculty Mannheim of Heidelberg University, Mannheim, Germany
| | - Sophia Adam
- Medical Faculty Mannheim of Heidelberg University, Mannheim, Germany
| | - Karl-Friedrich Kowalewski
- Department of Urology and Urological Surgery, University Medical Center Mannheim, Medical Faculty of Heidelberg University, Mannheim, Germany
| | - Manuel Neuberger
- Department of Urology and Urological Surgery, University Medical Center Mannheim, Medical Faculty of Heidelberg University, Mannheim, Germany
| | - Philipp Nuhn
- Department of Urology and Urological Surgery, University Medical Center Mannheim, Medical Faculty of Heidelberg University, Mannheim, Germany
| | - Maurice S Michel
- Department of Urology and Urological Surgery, University Medical Center Mannheim, Medical Faculty of Heidelberg University, Mannheim, Germany
| | - Maximilian C Kriegmair
- Department of Urology and Urological Surgery, University Medical Center Mannheim, Medical Faculty of Heidelberg University, Mannheim, Germany
| |
Collapse
|
7
|
Lee G, Patel HV, Srivastava A, Ghodoussipour S. Updates on enhanced recovery after surgery for radical cystectomy. Ther Adv Urol 2022; 14:17562872221109022. [PMID: 35844831 PMCID: PMC9280843 DOI: 10.1177/17562872221109022] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/03/2022] [Accepted: 06/07/2022] [Indexed: 12/24/2022] Open
Abstract
Enhanced Recovery after Surgery (ERAS) is a multimodal pathway that provides evidence-based guidance for improving perioperative care and outcomes in patients undergoing surgery. In 2013, the ERAS society released its original guidelines for radical cystectomy (RC) for bladder cancer (BC), adopting much of its supporting data from colorectal literature. In the last decade, growing interest in ERAS has increased RC-specific ERAS research, including prospective randomized controlled trials (RCTs). Collective data suggest ERAS contributes to improved complication rates, decreased hospital length-of-stay, and/or time to bowel recovery. Various institutions have adopted modified versions of the ERAS pathway, yet there remains a lack of consensus on the efficacy of specific ERAS items and standardization of the protocol. In this review, we summarize updated evidence and practice patterns of ERAS pathways for RC since the introduction of the original 2013 guidelines. Novel target interventions, including use of immunonutrition, prehabilitation, alvimopan, and methods of local analgesia are reviewed. Finally, we discuss barriers to implementing and future steps in advancing the ERAS movement.
Collapse
Affiliation(s)
- Grace Lee
- Section of Urologic Oncology, Rutgers Cancer Institute of New Jersey and Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ, USA
| | - Hiren V Patel
- Section of Urologic Oncology, Rutgers Cancer Institute of New Jersey and Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ, USA
| | - Arnav Srivastava
- Section of Urologic Oncology, Rutgers Cancer Institute of New Jersey and Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ, USA
| | - Saum Ghodoussipour
- Section of Urologic Oncology, Rutgers Cancer Institute of New Jersey and Rutgers Robert Wood Johnson Medical School, 195 Little Albany Street, Room 4561, New Brunswick, NJ 08903, USA
| |
Collapse
|