Early Postoperative Prediction of Complications and Readmission After Colorectal Cancer Surgery Using an Artificial Neural Network.

BACKGROUND: Early predictors of postoperative complications can risk-stratify patients undergoing colorectal cancer surgery. However, conventional regression models have limited power to identify complex nonlinear relationships among a large set of variables. We developed artificial neural network models to optimize the prediction of major postoperative complications and risk of readmission in patients undergoing colorectal cancer surgery. OBJECTIVE: The aim of this study was to develop an artificial neural network model to predict postoperative complications using postoperative laboratory values, and compare these models' accuracy to standard regression methods. DESIGN: This retrospective study included patients who underwent elective colorectal cancer resection between January 1, 2016, and July 31, 2021. Clinical data, cancer stage, and laboratory data from postoperative day 1 to 3 were collected. Models of complications and readmission risk were created using multivariable logistic regression and single-layer neural networks. SETTING: National Cancer Institute-Designated Comprehensive Cancer Center. PATIENTS: Adult colorectal cancer patients. MAIN OUTCOME MEASURES: Accuracy of predicting postoperative major complication, readmission and anastomotic leak using the area under the receiver-operating characteristic curve. RESULTS: Neural networks had larger areas under the curve for predicting major complications compared to regression models (neural network 0.811; regression model 0.724, p < 0.001). Neural networks also showed an advantage in predicting anastomotic leak (p = 0.036) and readmission using postoperative day 1-2 values (p = 0.014). LIMITATIONS: Single-center, retrospective design limited to cancer operations. CONCLUSIONS: In this study, we generated a set of models for early prediction of complications after colorectal surgery. The neural network models provided greater discrimination than the models based on traditional logistic regression. These models may allow for early detection of postoperative complications as soon as postoperative day 2. See Video Abstract.

Testing Machine Learning Models to Predict Postoperative Ileus after Colorectal Surgery.

Background: Postoperative ileus (POI) is a common complication after colorectal surgery, leading to increased hospital stay and costs. This study aimed to explore patient comorbidities that contribute to the development of POI in the colorectal surgical population and compare machine learning (ML) model accuracy to existing risk instruments. Study Design: In a retrospective study, data were collected on 316 adult patients who underwent colorectal surgery from January 2020 to December 2021. The study excluded patients undergoing multi-visceral resections, re-operations, or combined primary and metastatic resections. Patients lacking follow-up within 90 days after surgery were also excluded. Eight different ML models were trained and cross-validated using 29 patient comorbidities and four comorbidity risk indices (ASA Status, NSQIP, CCI, and ECI). Results: The study found that 6.33% of patients experienced POI. Age, BMI, gender, kidney disease, anemia, arrhythmia, rheumatoid arthritis, and NSQIP score were identified as significant predictors of POI. The ML models with the greatest accuracy were AdaBoost tuned with grid search (94.2%) and XG Boost tuned with grid search (85.2%). Conclusions: This study suggests that ML models can predict the risk of POI with high accuracy and may offer a new frontier in early detection and intervention for postoperative outcome optimization. ML models can greatly improve the prediction and prevention of POI in colorectal surgery patients, which can lead to improved patient outcomes and reduced healthcare costs. Further research is required to validate and assess the replicability of these results.

Total Neoadjuvant Therapy for Rectal Cancer: Which Regimens to Use?

Total neoadjuvant therapy (TNT) is a novel strategy for rectal cancer that administers both (chemo)radiotherapy and systemic chemotherapy before surgery. TNT is expected to improve treatment compliance, tumor regression, organ preservation, and oncologic outcomes. Multiple TNT regimens are currently available with various combinations of the treatments including induction or consolidation chemotherapy, triplet or doublet chemotherapy, and long-course chemoradiotherapy or short-course radiotherapy. Evidence on TNT is rapidly evolving with new data on clinical trials, and no definitive consensus has been established on which regimens to use for improving outcomes. Clinicians need to understand the advantages and limitations of the available regimens for multidisciplinary decision making. This article reviews currently available evidence on TNT for rectal cancer. A decision making flow chart is provided for tailor-made use of TNT regimens based on tumor location and local and systemic risk.

Para-aortic Lymph Node Dissection for Colorectal Cancer: Predicting Pathologic Lymph Node Positivity and Optimizing Outcomes.

INTRODUCTION: In colorectal cancer, the presence of para-aortic lymph nodes (PALN) indicates extraregional disease. Appropriately selecting patients for whom PALN dissection will provide oncologic benefit remains challenging. This study identified factors to predict survival among patients undergoing PALN dissection for colorectal cancer. METHODS: An institutional database was queried for patients who underwent curative-intent resection of clinically positive PALN for colorectal cancer between 2007 and 2020. Preoperative radiologic images were reviewed, and patients who did and did not have positive PALN on final pathology were compared. Survival analysis was performed to evaluate the impact of pathologically positive PALN on recurrence-free (RFS) and overall survival (OS). RESULTS: Of 74 patients who underwent PALN dissection, 51 had PALN metastasis at the time of primary tumor diagnosis, whereas 23 had metachronous PALN disease. Preoperative chemotherapy ± radiotherapy was given in 60 cases (81.1%), and 28 (37.8%) had pathologically positive PALN. Independent factors associated with positive PALN pathology included metachronous PALN disease and pretreatment and posttreatment radiographically abnormal PALN. On multivariable analysis, pathologically positive PALN was significantly associated with decreased RFS (hazard ratio 3.90) and OS (HR 4.49). Among patients with pathologically positive PALN, well/moderately differentiated histology was associated with better OS, and metachronous disease trended toward an association with better OS. CONCLUSIONS: Pathologically positive PALN are associated with poorer RFS and OS after PALN dissection for colorectal cancer. Clinicopathologic factors may predict pathologic PALN positivity. Curative-intent surgery may provide benefit, especially in patients with well-to-moderately differentiated primary tumors and possibly metachronous PALN disease.

Application of Machine Learning in Predicting Perioperative Outcomes in Patients with Cancer: A Narrative Review for Clinicians.

This narrative review explores the utilization of machine learning (ML) and artificial intelligence (AI) models to enhance perioperative cancer care. ML and AI models offer significant potential to improve perioperative cancer care by predicting outcomes and supporting clinical decision-making. Tailored for perioperative professionals including anesthesiologists, surgeons, critical care physicians, nurse anesthetists, and perioperative nurses, this review provides a comprehensive framework for the integration of ML and AI models to enhance patient care delivery throughout the perioperative continuum.

The current multidisciplinary management of rectal cancer.

Multidisciplinary management of rectal cancer has rapidly evolved over the last several years. This review describes recent data surrounding total neoadjuvant therapy, organ preservation, and management of lateral pelvic lymph nodes. It then presents our treatment algorithm for management of rectal cancer at The University of Texas MD Anderson Cancer Center in the context of this and other existing literature. As part of this discussion, the review describes how we tailor management based upon both patient and tumor-related factors in an effort to optimize patient outcomes.

Automated, High-Throughput Platform to Generate a High-Reliability, Comprehensive Rectal Cancer Database.

PURPOSE: Dynamic operations platforms allow for cross-platform data extraction, integration, and analysis, although application of these platforms to large-scale oncology enterprises has not been described. This study presents a pipeline for automated, high-fidelity extraction, integration, and validation of cross-platform oncology data in patients undergoing treatment for rectal cancer at a single, high-volume institution. METHODS: A dynamic operations platform was used to identify patients with rectal cancer treated at MD Anderson Cancer Center between 2016 and 2022 who had magnetic resonance imaging (MRI) imaging and preoperative treatment details available in the electronic health record (EHR). Demographic, clinicopathologic, tumor mutation, radiographic, and treatment data were extracted from the EHR using a methodology adaptable to any disease site. Data accuracy was assessed by manual review. Accuracy before and after implementation of synoptic reporting was determined for MRI data. RESULTS: A total of 516 patients with localized rectal cancer were included. In the era after institutional adoption of synoptic reports, the dynamic operations platform extracted T (tumor) category data from the EHR with 95% accuracy compared with 87% before the use of synoptic reports, and N (lymph node) category with 88% compared with 58%. Correct extraction of pelvic sidewall adenopathy was 94% compared with 78%, and extramural vascular invasion accuracy was 99% compared with 89%. Neoadjuvant chemotherapy and radiation data were 99% accurate for patients who had synoptic data sources. CONCLUSION: Using dynamic operations platforms enables automated cross-platform integration of multiparameter oncology data with high fidelity in patients undergoing multimodality treatment for rectal cancer. These pipelines can be adapted to other solid tumors and, together with standardized reporting, can increase efficiency in clinical research and the translation of actionable findings toward optimizing patient outcomes.

Perioperative and oncological outcomes following robotic en bloc multivisceral resection for colorectal cancer.

AIM: As multidisciplinary treatment strategies for colorectal cancer have improved, aggressive surgical resection has become commonplace. Multivisceral and extended resections offer curative-intent resection with significant survival benefit. However, limited data exist regarding the feasibility and oncological efficacy of performing extended resection via a minimally invasive approach. The aim of this study was to determine the perioperative and long-term outcomes following robotic extended resection for colorectal cancer. METHOD: We describe the population of patients undergoing robotic multivisceral resection for colorectal cancer at our single institution. We evaluated perioperative details and investigated short- and long-term outcomes, using the Kaplan-Meier method to analyse overall and recurrence-free survival. RESULTS: Among the 86 patients most tumours were T3 (47%) or T4 (47%) lesions in the rectum (78%). Most resections involved the anterior compartment (72%): bladder (n = 13), seminal vesicle/vas deferens (n = 27), ureter (n = 6), prostate (n = 15) and uterus/vagina/adnexa (n = 27). Three cases required conversion to open surgery; 10 patients had grade 3 complications. The median hospital stay was 4 days. Resections were R0 (>1 mm) in 78 and R1 (0 to ≤1 mm) in 8, with none being R2. The average nodal yield was 26 and 48 (55.8%) were pN0. Three-year overall survival was 88% and median progression-free survival was 19.4 months. Local recurrence was 6.1% and distant recurrence was 26.1% at 3 years. CONCLUSION: Performance of multivisceral and extended resection on the robotic platform allows patients the benefit of minimally invasive surgery while achieving oncologically sound resection of colorectal cancer.

Serum Tumor Markers and Outcomes in Patients With Appendiceal Adenocarcinoma.

Importance: Serum tumor markers carcinoembryonic antigen (CEA), carbohydrate antigen 19-9 (CA19-9), and cancer antigen 125 (CA125) have been useful in the management of gastrointestinal and gynecological cancers; however, there is limited information regarding their utility in patients with appendiceal adenocarcinoma. Objective: To assess the association of serum tumor markers (CEA, CA19-9, and CA125) with clinical outcomes and pathologic and molecular features in patients with appendiceal adenocarcinoma. Design, Setting, and Participants: This is a retrospective cohort study at a single tertiary care comprehensive cancer center. The median (IQR) follow-up time was 52 (21-101) months. Software was used to query the MD Anderson internal patient database to identify patients with a diagnosis of appendiceal adenocarcinoma and at least 1 tumor marker measured at MD Anderson between March 2016 and May 2023. Data were analyzed from January to December 2023. Main Outcomes and Measures: Association of serum tumor markers with survival in patients with appendiceal adenocarcinoma. Cox proportional hazards regression analyses were also performed to assess associations between clinical factors (serum tumor marker levels, demographics, and patient and disease characteristics) and patient outcomes (overall survival). Results: A total of 1338 patients with appendiceal adenocarcinoma were included, with a median (range) age at diagnosis of 56.5 (22.3-89.6) years. The majority of the patients had metastatic disease (1080 patients [80.7%]). CEA was elevated in 742 of the patients tested (56%), while CA19-9 and CA125 were elevated in 381 patients (34%) and 312 patients (27%), respectively. Individually, elevation of CEA, CA19-9, or CA125 were associated with worse 5-year survival; elevated vs normal was 81% vs 95% for CEA (hazard ratio [HR], 4.0; 95% CI, 2.9-5.6), 84% vs 92% for CA19-9 (HR, 2.2; 95% CI, 1.4-3.4), and 69% vs 93% for CA125 (HR, 4.6; 95% CI, 2.7-7.8) (P < .001 for all). Quantitative evaluation of tumor markers was associated with outcomes. Patients with highly elevated (top 10th percentile) CEA, CA19-9, or CA125 had markedly worse survival, with 5-year survival rates of 59% for CEA (HR, 9.8; 95% CI, 5.3-18.0), 64% for CA19-9 (HR, 6.0; 95% CI, 3.0-11.7), and 57% for CA125 (HR, 7.6; 95% CI, 3.5-16.5) (P < .001 for all). Although metastatic tumors had higher levels of all tumor markers, when restricting survival analysis to 1080 patients with metastatic disease, elevated CEA, CA19-9, or CA125 were all still associated worse survival (HR for CEA, 3.4; 95% CI, 2.5-4.8; P < .001; HR for CA19-9, 1.8; 95% CI, 1.2-2.7; P = .002; and HR for CA125, 3.9; 95% CI, 2.4-6.4; P < .001). Interestingly, tumor grade was not associated with CEA or CA19-9 level, while CA-125 was slightly higher in high-grade tumors relative to low-grade tumors (mean value, 18.3 vs 15.0; difference, 3.3; 95% CI, 0.9-3.7; P < .001). Multivariable analysis identified an incremental increase in the risk of death with an increase in the number of elevated tumor markers, with an 11-fold increased risk of death in patients with all 3 tumor markers elevated relative to those with none elevated. Somatic mutations in KRAS and GNAS were associated with significantly higher levels of CEA and CA19-9. Conclusions and Relevance: In this retrospective study of serum tumor markers in patients with appendiceal adenocarcinoma, CEA, CA19-9, and CA125 were associated with overall survival in appendiceal adenocarcinoma. Given their value, all 3 biomarkers should be included in the initial workup of patients with a diagnosis of appendiceal adenocarcinoma.

The Clinical Significance of CEA, CA19-9, and CA125 in Management of Appendiceal Adenocarcinoma.

Importance: Serum tumor markers CEA, CA19-9, & CA125 have been useful in the management of gastrointestinal and gynecological cancers, however there is limited information regarding their utility in patients with appendiceal adenocarcinoma. Objective: Assessing the association of serum tumor markers (CEA, CA19-9, and CA125) with clinical outcomes, pathologic, and molecular features in patients with appendiceal adenocarcinoma. Design: This is a retrospective study with results reported in 2023. The median follow-up time was 43 months. Setting: Single tertiary care comprehensive cancer center. Participants: Under an approved Institutional Review Board protocol, the Palantir Foundry software system was used to query the MD Anderson internal patient database to identify patients with a diagnosis of appendiceal adenocarcinoma and at least one tumor marker measured at MD Anderson between 2016 and 2023. Results: A total of 1,338 patients with appendiceal adenocarcinoma were included, with a median age of 56.5 years. The majority of the patients had metastatic disease (80.7%). CEA was elevated in more than half of the patients tested (56%), while CA19-9 and CA125 were elevated in 34% and 27%, respectively. Individually, elevation of CEA, CA19-9, or CA125 were associated with worse 5-year survival; 82% vs 95%, 84% vs 92%, and 69% vs 93% elevated vs normal for CEA, CA19-9, and CA125 respectively (all p<0.0001). Quantitative evaluation of tumor markers increased prognostic ability. Patients with highly elevated (top 10th percentile) CEA, CA19-9 or CA125 had markedly worse survival with 5-year survival rates of 59%, 64%, and 57%, respectively (HR vs. normal : 9.8, 6.0, 7.6, all p<0.0001). Although metastatic tumors had higher levels of all tumor markers, when restricting survival analysis to 1080 patients with metastatic disease elevated CEA, CA19-9 or CA125 were all still associated worse survival (HR vs. normal : 3.4, 1.8, 3.9, p<0.0001 for CEA and CA125, p=0.0019 for CA19-9). Interestingly tumor grade was not associated with CEA or CA19-9 level, while CA-125 was slightly higher in high relative to low-grade tumors (18.3 vs. 15.0, p=0.0009). Multivariable analysis identified an incremental increase in the risk of death with an increase in the number of elevated tumor markers, with a 11-fold increased risk of death in patients with all three tumor markers elevated relative to those with none elevated. Mutation in KRAS and GNAS were associated with significantly higher levels of CEA and CA19-9. Conclusions: These findings demonstrate the utility of measuring CEA, CA19-9, and CA125 in the management of appendiceal adenocarcinoma. Given their prognostic value, all three biomarkers should be included in the initial workup of patients diagnosed with appendiceal adenocarcinoma.

Efficacy of Systemic Chemotherapy in Patients With Low-grade Mucinous Appendiceal Adenocarcinoma: A Randomized Crossover Trial.

Importance: Appendiceal adenocarcinoma is a rare tumor, and given the inherent difficulties in performing prospective trials in such a rare disease, there are currently minimal high-quality data to guide treatment decisions, highlighting the need for more preclinical and clinical investigation for this disease. Objective: To prospectively evaluate the effectiveness of fluoropyrimidine-based systemic chemotherapy in patients with inoperable low-grade mucinous appendiceal adenocarcinoma. Design, Setting, and Participants: This open-label randomized crossover trial recruited patients at a single tertiary care comprehensive cancer center from September 2013 to January 2021. The data collection cutoff was May 2022. Enrollment of up to 30 patients was planned. Eligible patients had histological evidence of a metastatic low-grade mucinous appendiceal adenocarcinoma, with radiographic imaging demonstrating the presence of mucinous peritoneal carcinomatosis and were not considered candidates for complete cytoreductive surgery. Key exclusion criteria were concurrent or recent investigational therapy, evidence of bowel obstruction, and use of total parenteral nutrition. Data were analyzed from November 2021 to May 2022. Interventions: Patients were randomized to either 6 months observation followed by 6 months of chemotherapy, or initial chemotherapy followed by observation. Main Outcomes and Measures: The primary end point was the percentage difference in tumor growth in treatment and observation groups. Key secondary end points included patient-reported outcomes in the chemotherapy and observation periods, objective response rate, rate of bowel complications, and differences in overall survival (OS). Results: A total of 24 patients were enrolled, with median (range) age of 63 (38 to 82) years, and equal proportion of men and women (eg, 12 men [50%]); all patients had ECOG performance status of 0 or 1. A total of 11 patients were randomized to receive chemotherapy first, and 13 patients were randomized to receive observation first. Most patients (15 patients [63%]) were treated with either fluorouracil or capecitabine as single agent; 3 patients (13%) received doublet chemotherapy (leucovorin calcium [folinic acid], fluorouracil, and oxaliplatin or folinic acid, fluorouracil, and irinotecan hydrochloride), and bevacizumab was added to cytotoxic chemotherapy for 5 patients (21%). Fifteen patients were available to evaluate the primary end point of difference in tumor growth during treatment and observation periods. Tumor growth while receiving chemotherapy increased 8.4% (95% CI, 1.5% to 15.3%) from baseline but was not significantly different than tumor growth during observation (4.0%; 95% CI, -0.1% to 8.0%; P = .26). Of 18 patients who received any chemotherapy, none had an objective response (14 patients [77.8%] had stable disease; 4 patients [22.2%] had progressive disease). Median (range) OS was 53.2 (8.1 to 95.5) months, and there was no significant difference in OS between the observation-first group (76.0 [8.6 to 95.5] months) and the treatment-first group (53.2 [8.1 to 64.1] months; hazard ratio, 0.64; 95% CI, 0.16-2.55; P = .48). Patient-reported quality-of-life metrics identified that during treatment, patients experienced significantly worse fatigue (mean [SD] score, 18.5 [18.6] vs 28.9 [21.3]; P = .02), peripheral neuropathy (mean [SD] score, 6.67 [12.28] vs 38.89 [34.88]; P = .01), and financial difficulty (mean [SD] score, 8.9 [15.2] vs 28.9 [33.0]; P = .001) compared with during observation. Conclusions and Relevance: In this prospective randomized crossover trial of systemic chemotherapy in patients with low-grade mucinous appendiceal adenocarcinoma, patients did not derive clinical benefit from fluorouracil-based chemotherapy, given there were no objective responses, no difference in OS when treatment was delayed 6 months, and no difference in the rate of tumor growth while receiving chemotherapy. Trial Registration: ClinicalTrials.gov Identifier: NCT01946854.

Insurance Disparities in Access to Robotic Surgery for Colorectal Cancer.

BACKGROUND: The use of the robotic approach is increasing for colorectal cancer operations, but the added cost of the platform has the potential to introduce challenges in its dissemination. We hypothesized that adoption of the robot is introducing new disparities in access to minimally invasive surgery (MIS) for colorectal cancer, especially across patient insurance groups. METHODS: This cross-sectional study analyzed surgical cases of stage I-III colorectal cancer from the National Cancer Database (NCDB) between 2010 and 2019. The primary outcome was surgical approach (robotic, laparoscopic, or the composite "MIS"). The predictor was a patient's primary payor. Potential confounders included sociodemographics, tumor characteristics, and the facility. Hierarchical multivariable models were generated, and sensitivity analyses were performed. RESULTS: For colorectal cancer operations, the MIS approach increased from 39% in 2010 to 73% in 2019, driven predominantly by an increase in the robotic approach from 2 to 24%. For laparoscopy, the size of the disparity between patients with Private insurance and Medicaid shrank from 11% (2010) to 4% (2019), whereas this disparity increased for the robotic approach from 1% (2010) to 5% (2019). On adjusted analysis, patients with Medicaid (odds ratio [OR] 0.86 [CI 0.79-0.95]) and the Uninsured (OR 0.67 [CI 0.56-0.79]) had lower odds of receiving a robotic operation than those with Private insurance in 2019. This disparity remained consistent across five sensitivity analyses. CONCLUSIONS: As the field of colorectal cancer surgery shifts away from laparoscopy and toward robotics, new inequities across patient insurance are emerging. Proactive efforts are needed to ensure all patients benefit from a minimally invasive approach.

Impact of cumulative operative time on postoperative complication risk in simultaneous resections of colorectal liver metastases and primary tumors.

BACKGROUND: Simultaneous resection of colorectal liver metastases (CLM) and primary colorectal cancers (CRC) is nuanced without firm rules for selection. This study aimed to identify factors associated with morbidity after simultaneous resection. METHODS: Using a prospective database, patients undergoing simultaneous CLM-CRC resection from 1/1/2017-7/1/2020 were analyzed. Regression modeling estimated impact of colorectal resection type, Kawaguchi-Gayet (KG) hepatectomy complexity, and perioperative factors on 90-day complications. RESULTS: Overall, 120 patients underwent simultaneous CLM-CRC resection. Grade≥2 complications occurred in 38.3% (n = 46); these patients experienced longer length of stay (median LOS 7.5 vs. 4, p < 0.001) and increased readmission (39% vs. 1.4%, p < 0.001) compared to patients with zero or Grade 1 complications. Median OR time was 298 min. Patients within highest operative time quartile (>506 min) had higher grade≥2 complications (57%vs. 23%, p = 0.04) and greater than 4-fold increased odds of grade≥2 morbidity (OR 4.3, 95% CI (Confidence Interval) 1.41-13.1, p = 0.01). After adjusting for Pringle time, KG complexity and colorectal resection type, increasing operative time was associated with grade≥2 complications, especially for resections in highest quartile of operative time (OR 7.28, 95% CI 1.73-30.6, p = 0.007). CONCLUSION: In patients undergoing simultaneous CLM-CRC resection, prolonged operative time is independently associated with grade≥2 complications. Awareness of cumulative operative time may inform intraoperative decision-making by surgical teams.

Prospective Study of Perioperative Circulating Tumor DNA Dynamics in Patients Undergoing Hepatectomy for Colorectal Liver Metastases.

OBJECTIVE: To evaluate the association of perioperative ctDNA dynamics on outcomes after hepatectomy for CLM. SUMMARY BACKGROUND DATA: Prognostication is imprecise for patients undergoing hepatectomy for CLM, and ctDNA is a promising biomarker. However, clinical implications of perioperative ctDNA dynamics are not well established. METHODS: Patients underwent curative-intent hepatectomy after preoperative chemotherapy for CLM (2013-2017) with paired prehepatectomy/postoperative ctDNA analyses via plasma-only assay. Positivity was determined using a proprietary variant classifier. Primary endpoint was recurrence-free survival (RFS). Median follow-up was 55 months. RESULTS: Forty-eight patients were included. ctDNA was detected before and after surgery (ctDNA+/+) in 14 (29%), before but not after surgery (ctDNA+/-) in 19 (40%), and not at all (ctDNA-/-) in 11 (23%). Adverse tissue somatic mutations were detected in TP53 (n = 26; 54%), RAS (n = 23; 48%), SMAD4 (n = 5; 10%), FBXW7 (n = 3; 6%), and BRAF (n = 2; 4%). ctDNA+/+ was associated with worse RFS (median: ctDNA+/+, 6.0 months; ctDNA+/-, not reached; ctDNA-/-, 33.0 months; P = 0.001). Compared to ctDNA+/+, ctDNA+/- was associated with improved RFS [hazard ratio (HR) 0.24 (95% confidence interval (CI) 0.1-0.58)] and overall survival [HR 0.24 (95% CI 0.08-0.74)]. Adverse somatic mutations were not associated with survival. After adjustment for prehepatectomy chemotherapy, synchronous disease, and ≥2 CLM, ctDNA+/- and ctDNA-/- were independently associated with improved RFS compared to ctDNA+/+ (ctDNA+/-: HR 0.21, 95% CI 0.08-0.53; ctDNA-/-: HR 0.21, 95% CI 0.08-0.56). CONCLUSIONS: Perioperative ctDNA dynamics are associated with survival, identify patients with high recurrence risk, and may be used to guide treatment decisions and surveillance after hepatectomy for patients with CLM.

Peritoneal recurrence after resection for Stage I-III colorectal cancer: A population analysis.

BACKGROUND: Colorectal cancer (CRC) often recurs in the peritoneum, although the pattern of peritoneal recurrence (PR) has received less attention. We sought to describe the presentation and risk factors for PR following CRC resection. METHODS: We performed a cohort study of patients undergoing resection of Stage I-III CRC from 2006 to 2007 using merged data from a Commission on Cancer Special Study and the National Cancer Database. We estimated the timing, method of detection, and risk factors for isolated PR. RESULTS: Here, 8991 patients were included and isolate PR occurred in 77 (0.9%) patients. The median time to PR was 16.2 months (intrquartile range = 9.3-28.0 months) and most patients were identified via new symptoms (36.4%). Pathologic factors associated with increased odds of PR included higher T stage (T3 vs. T2, odds ratio [OR] = 4.8, 95% confidence interval [CI] = 1.5-15.7), N stage (N1 vs. N0, OR = 2.00, CI = 1.1-3.7), and signet ring (OR = 8.2, CI = 3.0-22.3) or mucinous histology (OR = 2.6, CI = 1.5-4.7). CONCLUSIONS: The majority of PR was detected within 18 months and few were identified by surveillance. Advanced T/N stage and signet ring/mucinous histology were associated with increased odds of PR.

Oncologic Outcomes of Multi-Institutional Minimally Invasive Inguinal Lymph Node Dissection for Melanoma Compared with Open Inguinal Dissection in the Second Multicenter Selective Lymphadenectomy Trial (MSLT-II).

BACKGROUND: Minimally invasive inguinal lymphadenectomy (MILND) is safe and feasible, but limited data exist regarding oncologic outcomes. METHODS: This study performed a multi-institutional retrospective cohort analysis of consecutive MILND performed for melanoma between January 2009 and June 2016. The open ILND (OILND) comparative cohort comprised patients enrolled in the second Multicenter Selective Lymphadenectomy Trial (MSLT-II) between December 2004 and March 2014.The pre-defined primary end point was the same-basin regional nodal recurrence, calculated using properties of binomial distribution. Time to events was calculated using the Kaplan-Meier method. The secondary end points were overall survival, progression-free survival, melanoma-specific survival (MSS), and distant metastasis-free survival (DMFS). RESULTS: For all the patients undergoing MILND, the same-basin regional recurrence rate was 4.4 % (10/228; 95 % confidence interval [CI], 2.1-7.9 %): 8.2 % (4/49) for clinical nodal disease and 3.4 % (6/179) for patients with a positive sentinel lymph node (SLN) as the indication. For the 288 patients enrolled in MSLT-II who underwent OILND for a positive SLN, 17 (5.9 %) had regional node recurrence as their first event. After controlling for ulceration, positive LN count and positive non-SLNs at the time of lymphadenectomy, no difference in OS, PFS, MSS or DMFS was observed for patients with a positive SLN who underwent MILND versus OILND. CONCLUSION: This large multi-institutional experience supports the oncologic safety of MILND for melanoma. The outcomes in this large multi-institutional experience of MILND compared favorably with those for an OILND population during similar periods, supporting the oncologic safety of MILND for melanoma.