Machine learning model for prediction of permanent stoma after anterior resection of rectal cancer: A multicenter study

BACKGROUND

The conversion from a temporary to a permanent stoma (PS) following rectal cancer surgery significantly impacts the quality of life of patients. However, there is currently a lack of practical preoperative tools to predict PS formation. The purpose of this study is to establish a preoperative predictive model for PS using machine learning algorithms to guide clinical practice.

METHODS

In this retrospective study, we analyzed clinical data from a total of 655 patients who underwent anterior resection for rectal cancer, with 552 patients from one medical center and 103 from another. Through machine learning algorithms, five predictive models were developed, and each was thoroughly evaluated for predictive performance. The model with superior predictive accuracy underwent additional validation using both an independent testing cohort and the external validation cohort. The Shapley Additive exPlanations (SHAP) approach was employed to elucidate the predictive factors influencing the model, providing an in-depth visual analysis of its decision-making process.

RESULTS

Eight variables were selected for the construction of the model. The support vector machine (SVM) model exhibited superior predictive performance in the training set, evidenced by an AUC of 0.854 (95 % CI:0.803-0.904). This performance was corroborated in both the testing set and external validation set, where the model demonstrated an AUC of 0.851 (95%CI:0.748-0.954) and 0.815 (95%CI:0.710-0.919), respectively, indicating its efficacy in identifying the PS.

CONCLUSIONS

The model(https://yangsu2023.shinyapps.io/psrisk/) indicated robust predictive performance in identifying PS after anterior resection for rectal cancer, potentially guiding surgeons in the preoperative stratification of patients, thus informing individualized treatment plans and improving patient outcomes.

Permanent stoma rate and long-term stoma complications in laparoscopic, robot-assisted, and transanal total mesorectal excisions: a retrospective cohort study

BACKGROUND

The surgical resection of rectal carcinoma is associated with a high risk of permanent stoma rate. Primary anastomosis rate is suggested to be higher in robot-assisted and transanal total mesorectal excision, but permanent stoma rate is unknown.

METHODS

Patients undergoing total mesorectal excision for MRI-defined rectal cancer between 2015 and 2017 in 11 centers highly experienced in laparoscopic, robot-assisted or transanal total mesorectal excision were included in this retrospective study. Permanent stoma rate, stoma-related complications, readmissions, and reoperations were registered. A multivariable regression analysis was performed for permanent stoma rate, stoma-related complications, and stoma-related reoperations.

RESULTS

In total, 1198 patients were included. Permanent stoma rate after low anterior resection (with anastomosis or with an end colostomy) was 40.1% in patients undergoing laparoscopic surgery, 21.3% in patients undergoing robot-assisted surgery, and 25.6% in patients undergoing transanal surgery (P < 0.001). Permanent stoma rate after low anterior resection with an anastomosis was 17.3%, 11.8%, and 15.1%, respectively. The robot-assisted and transanal techniques were independently associated with a reduction in permanent stoma rate in patients who underwent a low anterior resection (with anastomosis or with an end colostomy) (OR 0.39 [95% CI 0.25, 0.59] and OR 0.35 [95% CI 0.22, 0.55]), while this was not seen in patients who underwent a restorative low anterior resection. 45.4% of the patients who had a stoma experienced stoma-related complications, 4.0% were at least once readmitted, and 8.9% underwent at least one reoperation.

CONCLUSIONS

The robot-assisted and transanal techniques are associated with a lower permanent stoma rate in patients who underwent a low anterior resection.

Nomogram for predicting the probability of permanent stoma in patients with acute obstructive colorectal cancer

BACKGROUND

Acute obstructive colorectal cancer is a high-risk emergency among colorectal cancer (CRC). Approximately 20% of CRC patients are associated with a permanent stoma, which greatly affects the lifestyle of patients. This study aimed to investigate risk factors for predicting permanent stoma (PS) in patients with acute obstructive colorectal cancer.

METHODS

We retrospectively analyzed the clinical-pathological features of patients with acute obstructive colorectal cancer who underwent treatments from our hospital between January 2015 and December 2020. Univariate and multivariate logistic regression analyses were used to evaluate the risk factors for predicting PS chances of CRC patients using a nomogram method. Furthermore, the operating characteristic (ROC) curve and area under the ROC curve (AUC) were used to assess the discrimination power of the nomogram. Calibration plot was used to evaluate nomogram's calibration.

RESULTS

A total of 98 patients with acute obstructive colorectal cancer were enrolled in this study, including 24 PS patients with permanent stoma and 74 non-PS patients. Multivariate analysis showed that age [odds ratio (OR): 1.068, 95% confidence interval (CI): 1.006 ~ 1.135, P = 0.032], carcinoembryonic antigen (CEA) [OR: 1.015, 95% CI: 1.003 ~ 1.028, P = 0.013], and surgical method [emergency group vs. stent group, OR: 14.066, 95% CI: 3.625 ~ 54.572, p < 0.001] were independent risk factors for PS. These risk factors were incorporated into a nomogram and showed that the AUC of the nomogram was 0.867 (95% CI: 0.782-0.951). The calibration plot got consistent with prediction for PS in the nomogram.

CONCLUSION

Age, CEA, and surgical method were independent risk factors for PS in patients with acute obstructive colorectal cancer. Our nomogram has favorable predictive power for PS in CRC patients.

Risk factors for a permanent stoma after resection of left-sided obstructive colon cancer - A prediction model

INTRODUCTION

In patients with left-sided obstructive colon cancer (LSOCC), a stoma is often constructed as part of primary treatment, but with a considerable risk of becoming a permanent stoma (PS). The aim of this retrospective multicentre cohort is to identify risk factors for a PS in LSOCC and to develop a pre- and postoperative prediction model for PS.

MATERIALS AND METHODS

Data was retrospectively obtained from 75 hospitals in the Netherlands. Patients who had curative resection of LSOCC between January 1, 2009 to December 31, 2016 were included with a minimum follow-up of 6 months after resection. The interventions analysed were emergency resection, decompressing stoma or stent as bridge-to-elective resection. Main outcome measure was presence of PS at the end of follow-up. Multivariable logistic regression analysis was performed to identify risk factors for PS at primary presentation (T₀) and after resection, in patients having a stoma in situ (T₁). These risk factors were used to construct a web-based prediction tool.

RESULTS

Of 2099 patients included in the study (T₀), 779 had a PS (37%). A total of 1275 patients had a stoma in situ directly after resection (T₁), of whom 674 had a PS (53%). Median follow-up was 34 months. Multivariable analysis showed that older patients, female sex, high ASA-score and open approach were independent predictors for PS in both the T₀ and T₁ population. Other predictors at T₀ were sigmoid location, low Hb, high CRP, cM1 stage, and emergency resection. At T₁, subtotal colectomy, no primary anastomosis, not receiving adjuvant chemotherapy and high pTNM stage were additional predictors. Two predictive models were built, with an AUC of 0.74 for T₀ and an AUC of 0.81 for T₁.

CONCLUSIONS

PS is seen in 37% of the patients who have resection of LSOCC. In patients with a stoma in situ directly after resection, 53% PS are seen due to non-reversal. Not only baseline characteristics, but also treatment strategies determine the risk of a PS in patients with LSOCC. The developed predictive models will give physicians insight in the role of the individual variables on the risk of a PS and help in informing the patient about the probability of a PS.

Nomogram to predict permanent stoma in rectal cancer patients after sphincter-saving surgery

BACKGROUND

Approximately 20 percent of patients with a tumour localized in the low rectum still encounter the possibility of requiring permanent stoma (PS), which can cause drastic changes in lifestyle and physical perceptions.

AIM

To determine the risk factors for PS and to develop a prediction model to predict the probability of PS in rectal cancer patients after sphincter-saving surgery.

METHODS

A retrospective cohort of 421 rectal cancer patients who underwent radical surgery at Taipei Medical University Hospital between January 2012 and December 2020 was included in this study. Univariate and multivariate analyses were performed to identify the independent risk factors for PS. A nomogram was developed according to the independent risk factors obtained in the multivariate analysis. The performance of the nomogram was assessed using a receiver operating characteristic curve and a calibration curve.

RESULTS

The PS rate after sphincter-saving surgery was 15.1% (59/391) in our study after a median follow-up of 47.3 mo (range 7–114 mo). Multivariate logistic regression analysis demonstrated that local recurrence, perirectal abscess, anastomosis site stenosis, perineural invasion, tumor size and operative time were independent risk factors for PS. These identified risk factors were incorporated into the nomogram, and the concordance index of this model was 0.903 (95%CI: 0.851-0.955). According to the calibration curves, the nomogram represents a perfect prediction model.

CONCLUSION

Several risk factors for PS after sphincter-saving surgery were identified. Our nomogram exhibited perfect predictive ability and will improve a physician’s ability to communicate the benefits and risks of various treatment options in shared decision making.

Anastomotic leakage following restorative rectal cancer resection: treatment and impact on stoma presence 1 year after surgery-a population-based study

PURPOSE

Anastomotic leakage (AL) continues to be a challenge after restorative rectal resection (RRR). Various treatment options of AL are available; however, their long-term outcomes are uncertain. We explored the impact of AL on the risk of stoma presence 1 year after RRR for rectal cancer and described treatment of AL after RRR including impact on the probability of receiving adjuvant chemotherapy and stoma presence following different treatment options of AL.

METHODS

We included 859 patients undergoing RRR in Central Denmark Region between 2013 and 2019. Stoma presence was calculated as the proportion of patients with stoma 1 year after RRR. Multivariable logistic regression was conducted to estimate the impact of AL on stoma presence adjusting for potential predictors. Descriptive data of outcomes were stratified for various treatment options of AL.

RESULTS

The risk of stoma presence 1 year after surgery was 9.8% (95% CI 7.98-12.0). Predictors for having stoma 1 year after RRR were AL (OR 8.43 (95% CI 4.87-14.59)) and low tumour height (OR 3.85 (95% CI 1.22-13.21)). For patients eligible for adjuvant chemotherapy, the probability of receiving it was 42.9% (95% CI 21.8-66.0) if treated with endo-SPONGE and 71.4% (95% CI 47.8-88.7) if treated with other anastomosis preserving treatment options. The risk of having stoma 1 year after RRR was 33.9% (95% CI 21.8-47.8) for patients treated with endo-SPONGE and 13.5% (95% CI 5.6-25.8) for patients treated with other anastomosis preserving treatment options (p = 0.013).

CONCLUSION

AL is a strong predictor for stoma presence 1 year after RRR. Patients treated with endo-SPONGE seem to have worse outcomes compared to other anastomosis preserving treatment options.

Nomogram for Predicting the Probability of Permanent Stoma after Laparoscopic Intersphincteric Resection

PURPOSE

The purpose of this study was to determine the risk factors for the development of a permanent stoma in laparoscopic intersphincteric resection (LS-ISR) for ultralow rectal adenocarcinoma and to develop and validate a prediction model to predict the probability of permanent stoma after surgery.

METHODS

A primary cohort consisting of 301 consecutive patients who underwent LS-ISR was enrolled in this study. Multivariable logistic regression analysis was used to identify risk factors and develop the nomogram. The performance of the nomogram was assessed with respect to its calibration, discrimination, and clinical usefulness. An independent validation cohort contained 91 consecutive patients from January 2012 to January 2019.

RESULTS

The permanent stoma rate was 11.3% (34/301) in the primary cohort and 18.7% (17/91) in the validation cohort. Multivariable analysis revealed that nCRT (OR, 3.195; 95% CI, 1.169-8.733; P=0.024), ASA score of 3 (OR, 5.062; 95% CI, 1.877-13.646; P=0.001), distant metastasis (OR, 14.645; 95% CI, 3.186-67.315; P=0.001), and anastomotic leakage (OR, 11.308; 95% CI, 3.650-35.035; P<0.001) were independent risk factors for permanent stoma, and a nomogram was established. The AUCs of the nomogram were 0.842 and 0.858 in the primary and validation cohorts, respectively. The calibration curves showed good calibration in both cohorts. Decision curve analysis demonstrated that the nomogram was clinically useful.

CONCLUSION

We developed and validated a nomogram for ultralow rectal adenocarcinoma patients who underwent LS-ISR, and the nomogram could help surgeons identify which patients are at a higher risk of a permanent stoma after surgery.

Long-term stoma-related reinterventions after anterior resection for rectal cancer with or without anastomosis: population data from the Dutch snapshot study

BACKGROUND

The aim of this study was to analyze the stoma-related reinterventions, complications and readmissions after an anterior resection for rectal cancer, based on a cross-sectional nationwide cohort study with 3-year follow-up.

METHODS

Rectal cancer patients who underwent a resection with either a functional anastomosis, a defunctioned anastomosis, or Hartmann's procedure (HP) with an end colostomy in 2011 in 71 Dutch hospitals were included. The primary outcome was number of stoma-related reinterventions.

RESULTS

Of the 2095 patients with rectal cancer, 1400 patients received an anterior resection and were included in this study; 257 received an initially functional anastomosis, 741 a defunctioned anastomosis, and 402 patients a HP. Of the 1400 included patients, 62% were males, 38% were females and the mean age was 67 years (SD 11.1). Following a primary functional anastomosis, 48 (19%) patients received a secondary stoma. Stoma-related complications occurred in six (2%) patients, requiring reintervention in one (0.4%) case. In the defunctioned anastomosis group, stoma-related complications were present in 92 (12%) patients, and required reintervention in 23 (3%) patients, in 10 (1%) of these more than 1 year after initial resection. Stoma-related complications occurred in 92 (23%) patients after a HP, and required reintervention in 39 (10%) patients in 17 (4%) of cases more than 1 year after initial resection. The permanent stoma rate was 11% and 20%, in the functional anastomosis and the defuctioned anastomosis group, respectively. The end colostomy in the HP group was reversed in 4% of cases.

CONCLUSIONS

Construction of a stoma after resection for rectal cancer with preservation of the sphincter is accompanied with long-term stoma-related morbidity. Stoma complications are more frequent after a HP. Even after 1 year, a significant number of reinterventions are required.

Clinical characteristics and predictors of permanent stoma in rectal cancer patients underwent anterior resections: the value of preoperative prognostic nutritional index

OBJECTIVES

This study was aimed to investigate the correlation between preoperative prognostic nutritional index (PNI) and permanent stoma (PS) in patients with defunctioning stoma (DS) after anterior resections and, based on it, to reveal the clinical value of PNI on clinical strategies about the selection of stoma location.

METHODS

A total of 281 consecutive rectal cancer patients who accepted anterior resection and DS in the National Hepatobiliary and Enteric Surgery Research Center, Xiangya Hospital, Central South University from June 2008 to June 2018 were enrolled in this study. Receiver operating characteristic (ROC) curve for PNI was performed to discriminate PS. Univariate and multivariate analysis were conducted to identify the clinical characteristics and risk factors for PS. Specific reasons for patients with DS turned into PS were reviewed.

RESULTS

ROC curve analysis defined PNI cutoff level of 45.85 corresponding to PS (area under the curve (AUC) = 0.71, 77% sensitivity, 56.9% specificity). Low PNI (OR = 3.23, P = 0.005), tumor crossing the peritoneal reflection (PR) (OR = 3.42, P = 0.003), postoperative distant metastasis (OR = 6.31, P < 0.001) were independently associated with PS. Besides, anastomotic complications (31.4%), poor oncological outcomes (35.3%), and personal preferences (33.3%) were the specific reasons for patients turning into PS.

CONCLUSIONS

Preoperative PNI is an independent prognostic factor to predict PS in patients who underwent anterior resection and DS. Therefore, combined with other clinical characteristics and predictors, preoperative measurements of PNI could provide a significant support for clinical decision on patients prepared to accept anterior resection and DS.

Assessment of the risk of permanent stoma after low anterior resection in rectal cancer patients

Background

One of the most severe complications of low anterior rectal resection is anastomotic leakage (AL). The creation of a loop ileostomy (LI) reduces the prevalence of AL requiring surgical intervention. However, up to one-third of temporary stomas may never be closed.

The first aim of the study was to perform a retrospective assessment of the impact of LI on the risk of permanent stoma (PS) and symptomatic AL. The second aim of the study was to assess preoperative PS risk factors in patients with LI.

Methods

A total of 286 consecutive patients who underwent low anterior rectal resection were subjected to retrospective analysis. In 101 (35.3%) patients, diverting LI was performed due to low anastomosis, while in the remaining 185 (64.7%) patients, no ileostomy was performed. LIs were reversed after adjuvant treatment. Analyses of the effect of LI on symptomatic AL and PS were performed. Among the potential risk factors for PS, clinical factors and the values of selected peripheral blood parameters were analysed.

Results

PS occurred in 37.6% and 21.1% of the patients with LI and without LI, respectively (p < 0.01). Symptomatic ALs were significantly more common in patients without LI. In this group, symptomatic ALs occurred in 23.8% of patients, while in the LI group, they occurred in 5% of patients (p < 0.001). In the LI group, the only significant risk factor for PS in the multivariate analysis was preoperative plasma fibrinogen concentration (OR = 1.007, 97.5% CI 1.002–1.013, p = 0.013).

Conclusions

Although protective LI may reduce the incidence of symptomatic AL, it can be related to a higher risk of PS in this group of patients. The preoperative plasma fibrinogen concentration can be a risk factor for PS in LI patients and may be a useful variable in decision-making models.

Long distance to hospital is not a risk factor for non-reversal of a defunctioning stoma

PURPOSE

To see if road distance to hospital influences stoma reversal rate, time from index operation to stoma reversal, and occurrence of permanent stoma.

METHODS

Data from all diagnosed cases of rectal cancer from three counties in northern Sweden were extracted from the Swedish Rectal Cancer Registry. The three counties are sparsely populated, with a population density roughly one fifth the average density in Sweden. Distances to nearest, operating, and largest hospital were obtained using Google Maps™. Matched data on socioeconomic variables were retrieved from Statistics Sweden.

RESULTS

In univariate logistic regression analysis, patients living closer to the operating hospital had a higher likelihood of non-reversal than those living farther away (OR 0.3; 95% CI 0.12-0.76). However, no difference was seen in the multivariate analysis. Of the 717 cases included, 54% received a permanent stoma and 38% a defunctioning stoma at index surgery. The reversal rate of a defunctioning stoma was 83%. At follow-up, 61% still had a stoma, 89% of these were permanent, and 11% non-reversed defunctioning stomas. Median time to stoma reversal was 287 days (82-1557 days). Of all 227 stoma reversals, 77% were done more than 6 months after index surgery.

CONCLUSIONS

Longer distance to hospital is not a risk factor for non-reversal of a defunctioning stoma. Only 23% had their defunctioning stoma reversed within 6 months after index surgery. Future studies aiming to determine reversal rate need to extend their follow-up time in order to receive accurate results.

Application of cluster-based continuity nursing intervention in patients with permanent stoma after surgery for colorectal cancer

AIM

To investigate the effect of cluster-based continuity nursing intervention in patients with permanent colostomy after surgery for colorectal cancer.

METHODS

One hundred and twenty-eight patients with colorectal cancer were randomly divided into two groups: an intervention group and a control group. The intervention group was given cluster-based continuity nursing intervention, and the control group was given routine nursing and discharge guidance. Self-efficacy and self-care ability scores and the incidence of complications were compared between the two groups.

RESULTS

In this study, 128 patients aged from 48 to 76 (52.1 ± 7.3) years were investigated. There was no significant difference in gender, age, marriage, educational level, place of residence, or family income between the two groups (P > 0.05). The self-efficacy score was significantly higher in the observation group than in the control group (115.7 ± 9.8 vs 70.5 ± 7.5, t = 29.213, P < 0.001). The difference in self-care ability was also statistically significant between the two groups (t = 2.484, P = 0.014). The complication rate was 6.3% in the observation group and 17.2% in the control group, and the difference was statistically significant (χ² = 4.384, P = 0.036).

CONCLUSION

Cluster-based continuity care for patients with permanent stoma after surgery for colorectal cancer can improve the self-efficacy and self-care ability of patients, enhance patient's confidence, and reduce the incidence of complications.

Current use of diverting stoma in anterior resection for cancer: population-based cohort study of total and partial mesorectal excision

PURPOSE

A diverting stoma is commonly used to reduce the risk of anastomotic leakage when performing total mesorectal excision (TME) in anterior resection for rectal cancer. The purpose of this study was to evaluate the impact of fecal diversion in relation to partial mesorectal excision (PME).

METHODS

A retrospective analysis was undertaken on a national cohort, originally created to study the impact of central arterial ligation on patients with increased cardiovascular risk. Some 741 patients operated with anterior resection for rectal cancer during the years 2007 through 2010 were followed up for 53 months. Multivariate logistic regression was used to evaluate the impact of diverting stoma on the risk of anastomotic leakage and permanent stoma, expressed as odds ratios (ORs) and 95% confidence intervals (CIs).

RESULTS

The risk of anastomotic leakage was increased in TME surgery when not using a diverting stoma (OR 5.1; 95% CI 2.2-11.6), while the corresponding risk increase in PME patients was modest (OR 1.8; 95% CI 0.8-4.0). At study completion or death, 26 and 13% of TME and PME patients, respectively, had a permanent stoma. A diverting stoma was a statistically significant risk factor for a permanent stoma in PME patients (OR 4.7; 95% CI 2.5-9.0), while less important in TME patients (OR 1.8; 95% CI 0.6-5.5).

CONCLUSION

The benefit of a diverting stoma concerning anastomotic leakage in this patient group seems doubtful. Moreover, the diverting stoma itself may contribute to the high rate of permanent stomas.

Risk factors causing structural sequelae after anastomotic leakage in mid to low rectal cancer

AIM: To investigate the risk factors causing structural sequelae after anastomotic leakage in patients with mid to low rectal cancer.

METHODS: Prospectively collected data of consecutive subjects who had anastomotic leakage after surgical resection for rectal cancer from March 2006 to May 2013 at Korea University Anam Hospital were retrospectively analyzed. Two subgroup analyses were performed. The patients were initially divided into the sequelae (stricture, fistula, or sinus) and no sequelae groups and then divided into the permanent stoma (PS) and no PS groups. Univariate and multivariate analyses were performed to identify the risk factors of structural sequelae after anastomotic leakage.

RESULTS: Structural sequelae after anastomotic leakage were identified in 29 patients (39.7%). Multivariate analysis revealed that diversion ileostomy at the first operation increases the risk of structural sequelae [odds ratio (OR) = 6.741; P = 0.017]. Fourteen patients (17.7%) had permanent stoma during the follow-up period (median, 37 mo). Multivariate analysis showed that the tumor level from the dentate line was associated with the risk of permanent stoma (OR = 0.751; P = 0.045).

CONCLUSION: Diversion ileostomy at the first operation increased the risk of structural sequelae of the anastomosis, while lower tumor location was associated with the risk of permanent stoma in the management of anastomotic leakage.