1
|
Yu M, Yuan Z, Li R, Shi B, Wan D, Dong X. Interpretable machine learning model to predict surgical difficulty in laparoscopic resection for rectal cancer. Front Oncol 2024; 14:1337219. [PMID: 38380369 PMCID: PMC10878416 DOI: 10.3389/fonc.2024.1337219] [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: 11/22/2023] [Accepted: 01/15/2024] [Indexed: 02/22/2024] Open
Abstract
Background Laparoscopic total mesorectal excision (LaTME) is standard surgical methods for rectal cancer, and LaTME operation is a challenging procedure. This study is intended to use machine learning to develop and validate prediction models for surgical difficulty of LaTME in patients with rectal cancer and compare these models' performance. Methods We retrospectively collected the preoperative clinical and MRI pelvimetry parameter of rectal cancer patients who underwent laparoscopic total mesorectal resection from 2017 to 2022. The difficulty of LaTME was defined according to the scoring criteria reported by Escal. Patients were randomly divided into training group (80%) and test group (20%). We selected independent influencing features using the least absolute shrinkage and selection operator (LASSO) and multivariate logistic regression method. Adopt synthetic minority oversampling technique (SMOTE) to alleviate the class imbalance problem. Six machine learning model were developed: light gradient boosting machine (LGBM); categorical boosting (CatBoost); extreme gradient boost (XGBoost), logistic regression (LR); random forests (RF); multilayer perceptron (MLP). The area under receiver operating characteristic curve (AUROC), accuracy, sensitivity, specificity and F1 score were used to evaluate the performance of the model. The Shapley Additive Explanations (SHAP) analysis provided interpretation for the best machine learning model. Further decision curve analysis (DCA) was used to evaluate the clinical manifestations of the model. Results A total of 626 patients were included. LASSO regression analysis shows that tumor height, prognostic nutrition index (PNI), pelvic inlet, pelvic outlet, sacrococcygeal distance, mesorectal fat area and angle 5 (the angle between the apex of the sacral angle and the lower edge of the pubic bone) are the predictor variables of the machine learning model. In addition, the correlation heatmap shows that there is no significant correlation between these seven variables. When predicting the difficulty of LaTME surgery, the XGBoost model performed best among the six machine learning models (AUROC=0.855). Based on the decision curve analysis (DCA) results, the XGBoost model is also superior, and feature importance analysis shows that tumor height is the most important variable among the seven factors. Conclusions This study developed an XGBoost model to predict the difficulty of LaTME surgery. This model can help clinicians quickly and accurately predict the difficulty of surgery and adopt individualized surgical methods.
Collapse
Affiliation(s)
| | | | | | | | - Daiwei Wan
- Department of General Surgery, The First Affiliated Hospital of Soochow University, Suzhou, China
| | - Xiaoqiang Dong
- Department of General Surgery, The First Affiliated Hospital of Soochow University, Suzhou, China
| |
Collapse
|
2
|
Gachabayov M, Felsenreich DM, Bhatti S, Bergamaschi R. Does the visceral fat area impact the histopathology specimen metrics after total mesorectal excision for distal rectal cancer? Langenbecks Arch Surg 2023; 408:257. [PMID: 37389686 DOI: 10.1007/s00423-023-02981-7] [Citation(s) in RCA: 2] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/30/2022] [Accepted: 06/12/2023] [Indexed: 07/01/2023]
Abstract
BACKGROUND The aim of this study was to evaluate whether the visceral fat area (VFA) has an impact on the histopathology specimen metrics of male patients undergoing robotic total mesorectal resection (rTME) for distal rectal cancer. METHODS Prospectively collected data of patients undergoing rTME for resectable rectal cancer by five surgeons during a period of three years were extracted from the REgistry of Robotic SURgery for RECTal cancer (RESURRECT). VFA was measured in all patients at preoperative computed tomography. Distal rectal cancer was defined as <6cm from the anal verge. The histopathology metrics included circumferential resection margin (CRM) (in mm) and its involvement rate (if <1mm), distal resection margin (DRM), and quality of TME (complete, nearly-complete, incomplete). RESULTS Of 839 patients who underwent rTME, 500 with distal rectal cancer were included. One hundred and six (21.2%) males with VFA>100cm2 were compared to 394 (78.8%) males or females with VFA≤100cm2. The mean CRM of males with VFA>100cm2 was not significantly different from its counterpart (6.6 ± 4.8 mm versus 7.1 ± 9.5mm; p=0.752). CRM involvement rates were 7.6% in both groups (p=1.000). The DRM was not significantly different: 1.8±1.9cm versus 1.8±2.6cm; p=0.996. The quality of TME did not significantly differ: complete TME 87.3% vs. 83.7%; nearly complete TME 8.9% vs. 12.8%; incomplete TME 3.8% vs. 3.6%. Complications and clinical outcomes did not significantly differ. CONCLUSION This study did not find evidence to support that increased VFA would result in suboptimal histopathology specimen metrics when performing rTME in males with distal rectal cancer.
Collapse
Affiliation(s)
- M Gachabayov
- Section of Colorectal Surgery, Department of Surgery, New York Medical College, Taylor Pavilion, Suite D-365, 100 Woods Road, Valhalla, NY, 10595, USA
| | - D M Felsenreich
- Section of Colorectal Surgery, Department of Surgery, New York Medical College, Taylor Pavilion, Suite D-365, 100 Woods Road, Valhalla, NY, 10595, USA
- Division of Visceral Surgery, Department of General Surgery, Medical University of Vienna, Vienna, Austria
| | - S Bhatti
- Section of Colorectal Surgery, Department of Surgery, New York Medical College, Taylor Pavilion, Suite D-365, 100 Woods Road, Valhalla, NY, 10595, USA
| | - R Bergamaschi
- Section of Colorectal Surgery, Department of Surgery, New York Medical College, Taylor Pavilion, Suite D-365, 100 Woods Road, Valhalla, NY, 10595, USA.
| |
Collapse
|
3
|
Zhang Q, Wei J, Chen H. Advances in pelvic imaging parameters predicting surgical difficulty in rectal cancer. World J Surg Oncol 2023; 21:64. [PMID: 36843078 PMCID: PMC9969644 DOI: 10.1186/s12957-023-02933-x] [Citation(s) in RCA: 2] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/08/2022] [Accepted: 02/11/2023] [Indexed: 02/28/2023] Open
Abstract
Due to the fixed bony structure of the pelvis, the pelvic operation space is limited, complicating the surgical operation of rectal cancer, especially middle and low rectal cancer. The closer the tumor is to the anal verge, the smaller the operative field and operating space, the longer the operative time, and the greater the incidence of intraoperative side injuries and postoperative complications. To date, there is still no clear definition of a difficult pelvis that affects the surgical operation of rectal cancer. Few related research reports exist in the literature, and views on this aspect are not the same between countries. Therefore, it is particularly important to predict the difficulty of rectal cancer surgery in a certain way before surgery and to select the surgical method most suitable for each case during the treatment of rectal cancer.
Collapse
Affiliation(s)
- Qingbai Zhang
- grid.411491.8Department of General Surgery, The Fourth Affiliated Hospital of Harbin Medical University, Harbin, China
| | - Jiufeng Wei
- grid.411491.8Department of General Surgery, The Fourth Affiliated Hospital of Harbin Medical University, Harbin, China
| | - Hongsheng Chen
- Department of General Surgery, The Fourth Affiliated Hospital of Harbin Medical University, Harbin, China.
| |
Collapse
|
4
|
Abstract
BACKGROUND Central obesity is associated with surgical difficulties, but few studies explore the relationship with long-term results after colon cancer surgery. OBJECTIVE The purpose of this study was to investigate the association between perirenal fat surface area, a proxy for total visceral fat, and oncologic outcome after intestinal resection for colon cancer. DESIGN We investigated the association between perirenal fat surface area (exposure) on recurrence and death (outcome) in patients undergoing surgery with curative intent for colon cancer. SETTINGS The study was conducted at Stockholm South General Hospital, serving a population of 600,000. PATIENTS Patients (N = 733) without metastases at diagnosis who had a preoperative CT and had undergone elective colon resection between 2006 and 2016 were included. MAIN OUTCOME MEASURES We compared overall survival, recurrence-free survival, and cause-specific survival by perirenal fat surface area. RESULTS Patients with high perirenal fat surface area (fourth quartile) had more often left-sided tumors (45% vs 32% in the first quartile) and experienced more postoperative complications (29% vs 13%), but there were no differences in pathologic T and N stage, radicality of surgery, or adjuvant chemotherapy treatment. Overall survival decreased by increasing cancer stage but was not different between perirenal fat surface area categories. The HR for recurrence-free survival per centimeter squared increase in perirenal fat surface area was 1.00 (95% CI, 0.99-1.01) adjusted for age, sex, ASA category, tumor location, and postoperative complication Clavien-Dindo ≥2. The cumulative incidence of recurrence with death as a competing risk was not statistically different between perirenal fat surface area categories (p = 0.06). Subgroup analyses showed a nonsignificant tendency for men with low perirenal fat surface area to have a lower risk of recurrence and women a higher risk. LIMITATIONS In all register-based studies there can be randomly distributed errors. The results can only be generalized to colon resections. Our cohort ranged over a large year span. CONCLUSIONS We found no association between perirenal fat surface area and overall survival, recurrence-free survival, or cause-specific cumulative incidence of recurrence in patients undergoing colon resection for cancer. See Video Abstract at http://links.lww.com/DCR/B326. LA SUPERFICIE DE GRASA PERIRRENAL Y EL RESULTADO ONCOLGICO EN CIRUGA ELECTIVA DE CNCER DE COLON ANTECEDENTES:La obesidad central está asociada con dificultades quirúrgicas, pero pocos estudios exploran la relación de los resultados a largo plazo después de cirugía de cáncer de colon.OBJETIVO:Investigar la asociación entre la superficie de la grasa perirrenal, como un indicador de la grasa visceral total y el resultado oncológico después de una resección intestinal por cáncer de colon.DISEÑO:Se estudió la asociación entre el área de la superficie de la grasa perirrenal (expuesta) con la recurrencia y la muerte (resultado) de pacientes sometidos a cirugía con intención curativa por cáncer de colon.AJUSTES:Atención brindada por el Hospital General del Sur de Estocolmo a una población de 600,000 habitantes.PACIENTES:Aquellos pacientes sin metástasis (n = 733) en el momento del diagnóstico que tuvieron una tomografía computada preoperatoria y que se sometieron a una resección electiva de colon entre 2006-2016.PRINCIPALES MEDIDAS DE RESULTADO:Comparamos la sobrevida general, la sobrevida libre de recurrencia y la sobrevida específica de la causa, por área de superficie de grasa perirrenal.RESULTADOS:Los pacientes con una mayor área de superficie de grasa perirrenal (cuarto cuartil) tuvieron más frecuentemente tumores del lado izquierdo (45% frente a 32% en el primer cuartil) y sufrieron más complicaciones postoperatorias (29% frente a 13%), pero no hubieron diferencias en el Estadío patológico T y N, ni en lo radical de la cirugía o del tratamiento de quimioterapia adyuvante. La supervivencia general disminuyó al aumentar el estadio del cáncer, pero no fue diferente entre las categorías de área de superficie grasa perirrenal. La razón de riesgo para la sobrevida libre de recurrencia por aumento de cm2 en el área de la superficie grasa perirrenal fue de 1.00 (intervalo de confianza del 95%: 0.99-1.01) ajustada por edad, sexo, categoría de la Sociedad Americana de Anestesiólogos, ubicación del tumor y complicación postoperatoria según Clavien-Dindo ≥ 2) La incidencia acumulada de recurrencia con muerte como un riesgo competitivo no fue estadísticamente diferente entre las categorías de área de superficie grasa perirrenal (p = 0.06). Los análisis de subgrupos mostraron una tendencia no significativa para que los hombres con un área de superficie menor en grasa perirrenal tengan un menor riesgo de recurrencia y las mujeres un mayor riesgo.LIMITACIONES:En todos los estudios basados en registros puede haber errores distribuidos aleatoriamente. Los resultados solo pueden generalizarse a resecciones de colon. Nuestra cohorte osciló durante un gran lapso de años.CONCLUSIONES:No se encontró asociación entre el área de superficie de la grasa perirrenal y la sobrevida general, ni con la sobrevida libre de recurrencia o la incidencia acumulada de recurrencia específica de la causa en pacientes sometidos a resección de colon por cáncer. Consulte Video Resumen en http://links.lww.com/DCR/B326. (Traducción-Dr Xavier Delgadillo).
Collapse
|
5
|
Malamutmann E, Frenzen A, Karadag HI, Emre A, Klein CG, Paul A, Oezcelik A. Inner Abdominal Fat and Psoas Muscle as Predictive Factors for the Outcome After Liver Transplant. EXP CLIN TRANSPLANT 2021; 19:131-136. [PMID: 33455566 DOI: 10.6002/ect.2020.0265] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/11/2022]
Abstract
OBJECTIVES We analyzed the nutritional condition of liver transplant recipients and the body mass index, the inner abdominal fat tissue, the outer abdominal fat tissue, the psoas muscle size, and the psoas muscle index of the recipients and evaluated the effects of these factors on patient outcomes after liver transplant. MATERIALS AND METHODS We included recipients of liver transplants from January 2009 to December 2018 who had computed tomography at our center < 3 months before transplant. Preoperative, intraoperative, and postoperative data were evaluated. Outer abdominal fat tissue, inner abdominal fat tissue, and psoas muscle area were measured on the computed tomography abdominal images. We used univariate and multi-variate regression analyses to evaluate the data. RESULTS There were 265 patients; mean age was 54 years (SD, 13 years). The mean value for body mass index, calculated as weight in kilograms divided by height in meters squared, was 25 (SD, 5). The mean score for Model for End-Stage Liver Disease was 17 (SD, 6). All patients underwent orthotopic liver transplant by standard technique. After adjustment for multivariable analysis, the values for psoas muscle size and the psoas muscle index of the recipient were associated as independent factors for postoperative complications and duration of hospital stay. The survival rate at 1 year was 78.5%, and the rate of perioperative mortality was 16.6%. Independent factors associated with survival after liver transplant were inner abdominal fat tissue, etiology, and rate of major postoperative complications. CONCLUSIONS Inner abdominal fat tissue, psoas muscle size, and the psoas muscle index are significantly associated with postoperative complications and/or survival after liver transplant. Our results suggest that these prognostic factors may be useful to optimize the selection of appropriate candidates for liver transplant.
Collapse
Affiliation(s)
- Eugen Malamutmann
- From the Department of General, Visceral and Transplantation Surgery, University Hospital of Essen, University Duisburg-Essen, Essen, Germany
| | | | | | | | | | | | | |
Collapse
|
6
|
Clinical application of laparoscopic total mesorectal excision using the intersphincteric approach through the sacrococcygeal incision for treating patients with rectal cancer. Wideochir Inne Tech Maloinwazyjne 2019; 14:210-215. [PMID: 31118985 PMCID: PMC6528129 DOI: 10.5114/wiitm.2019.81316] [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: 07/15/2018] [Accepted: 10/30/2018] [Indexed: 11/17/2022] Open
Abstract
Introduction Colorectal cancer is the third most common cancer causing death in Western countries; laparoscopic surgery for colorectal cancer has many advantages and thus has been used widely. Laparoscopic total mesorectal excision through the sacrococcygeal incision under direct visualization to excise distal rectal cancer is an important procedure for super-low rectal carcinomas. Aim To investigate the feasibility of mesorectal excision and super-low rectal carcinoma excision using the intersphincteric approach through the sacrococcygeal incision. Material and methods From December 2009 to June 2017, intersphincteric resection was performed through the sacrococcygeal incision; the mesentery was excised in 27 patients with rectal cancer and a contracted pelvis (the lower edge of the tumor was 4 to 7 cm to the anal verge) through laparoscopy in the Gastrointestinal Surgery Department of our hospital. Results No death was recorded during surgery. The surgical time ranged from 190 to 310 min, the bleeding volume was 50 to 150 ml, and the post-surgical length of stay was 6 to 19 days. There were three cases of anastomotic fistulas, one case of anastomotic stenosis, and one case of fecal incontinence. Follow-up visits were scheduled for 19 patients, with a mean time of 37 months, ranging from 3 to 92 months; one case of local recurrence, one case of peritoneal metastasis, and two cases of hepatic metastasis were observed. Conclusions Laparoscopic total mesorectal excision using the intersphincteric approach through the sacrococcygeal incision is feasible for treating patients with a contracted pelvis and super-low rectal carcinoma.
Collapse
|
7
|
der Hagopian O, Dahlberg M, Heinius G, Nordberg J, Gustafsson J, Nordenvall C, Sandblom G, Farahnak P, Everhov ÅH. Perirenal fat surface area as a risk factor for perioperative difficulties and 30-day postoperative complications in elective colon cancer surgery. Colorectal Dis 2018; 20:1078-1087. [PMID: 29956867 DOI: 10.1111/codi.14322] [Citation(s) in RCA: 5] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/12/2018] [Accepted: 06/21/2018] [Indexed: 02/08/2023]
Abstract
AIM Visceral obesity is associated with perioperative and postoperative complications in colorectal surgery. We aimed to investigate the association between the perirenal fat surface area (PRF) and postoperative complications. METHOD Data on 610 patients undergoing curative, elective colon cancer resection between 2006 and 2016 at Stockholm South General Hospital were retrieved from a local quality register. We assessed perioperative and postoperative outcomes using a multinomial regression model adjusted for age, sex, American Society of Anesthesiologists classification and surgical approach (open/laparoscopy) in relation to PRF. RESULTS PRF could be measured in 605 patients; the median area was 24 cm2 . Patients with PRF ≥ 40 cm2 had longer operation time (median 223 vs 184 min), more intra-operative bleeding (250 vs 125 ml), reoperations (11% vs 6%), surgical complications (27% vs 13%) and nonsurgical infectious complications (16% vs 9%) than patients with PRF < 40 cm2 , but there were no differences in the need for intensive care or duration of hospital stay. The multivariate analyses revealed an increased risk of any complication [OR 1.68 (95% CI 1.1-2.6)], which was even more pronounced for moderate complications [Clavien-Dindo II, OR 2.14 (CI 1.2-2.4]; Clavien-Dindo III, OR 2.35 (CI 1.0-5.5)] in patients with PRF ≥ 40 vs < 40 cm2 . The absolute risk of complications was similar in men and women with PRF ≥ 40 cm2 . CONCLUSION PRF, an easily measured indirect marker of visceral obesity, was associated with overall and moderate complications in men and women and could serve as a useful tool in the assessment of preoperative risk.
Collapse
Affiliation(s)
- O der Hagopian
- Department of Clinical Science and Education Södersjukhuset (KI SÖS), Karolinska Institutet, Stockholm, Sweden
| | - M Dahlberg
- Department of Surgery, Södersjukhuset, Stockholm, Sweden
| | - G Heinius
- Department of Clinical Science and Education Södersjukhuset (KI SÖS), Karolinska Institutet, Stockholm, Sweden.,Department of Surgery, Södersjukhuset, Stockholm, Sweden
| | - J Nordberg
- Bilddiagnostiskt Centrum (BDC), Södersjukhuset, Stockholm, Sweden
| | - J Gustafsson
- Bilddiagnostiskt Centrum (BDC), Södersjukhuset, Stockholm, Sweden
| | - C Nordenvall
- Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden.,Center for Digestive Disease, Division of Coloproctology, Karolinska University Hospital, Stockholm, Sweden
| | - G Sandblom
- Department of Clinical Science and Education Södersjukhuset (KI SÖS), Karolinska Institutet, Stockholm, Sweden.,Department of Surgery, Södersjukhuset, Stockholm, Sweden
| | - P Farahnak
- Department of Clinical Science and Education Södersjukhuset (KI SÖS), Karolinska Institutet, Stockholm, Sweden.,Department of Surgery, Södersjukhuset, Stockholm, Sweden
| | - Å H Everhov
- Department of Clinical Science and Education Södersjukhuset (KI SÖS), Karolinska Institutet, Stockholm, Sweden.,Department of Surgery, Södersjukhuset, Stockholm, Sweden.,Clinical Epidemiology Unit, Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden
| |
Collapse
|
8
|
Body mass index and complications following major gastrointestinal surgery: a prospective, international cohort study and meta-analysis. Colorectal Dis 2018; 20:O215-O225. [PMID: 29897171 DOI: 10.1111/codi.14292] [Citation(s) in RCA: 44] [Impact Index Per Article: 7.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/23/2018] [Accepted: 05/29/2018] [Indexed: 02/08/2023]
Abstract
AIM Previous studies reported conflicting evidence on the effects of obesity on outcomes after gastrointestinal surgery. The aims of this study were to explore the relationship of obesity with major postoperative complications in an international cohort and to present a meta-analysis of all available prospective data. METHODS This prospective, multicentre study included adults undergoing both elective and emergency gastrointestinal resection, reversal of stoma or formation of stoma. The primary end-point was 30-day major complications (Clavien-Dindo Grades III-V). A systematic search was undertaken for studies assessing the relationship between obesity and major complications after gastrointestinal surgery. Individual patient meta-analysis was used to analyse pooled results. RESULTS This study included 2519 patients across 127 centres, of whom 560 (22.2%) were obese. Unadjusted major complication rates were lower in obese vs normal weight patients (13.0% vs 16.2%, respectively), but this did not reach statistical significance (P = 0.863) on multivariate analysis for patients having surgery for either malignant or benign conditions. Individual patient meta-analysis demonstrated that obese patients undergoing surgery for malignancy were at increased risk of major complications (OR 2.10, 95% CI 1.49-2.96, P < 0.001), whereas obese patients undergoing surgery for benign indications were at decreased risk (OR 0.59, 95% CI 0.46-0.75, P < 0.001) compared to normal weight patients. CONCLUSIONS In our international data, obesity was not found to be associated with major complications following gastrointestinal surgery. Meta-analysis of available prospective data made a novel finding of obesity being associated with different outcomes depending on whether patients were undergoing surgery for benign or malignant disease.
Collapse
|