Faculty Bibliography

BACKGROUND:Robotic surgery is attractive for resection of low rectal cancer due to greater dexterity and visualization, but its benefit is poorly understood. We aimed to determine if operative approach impacts radial margin positivity (RMP) and postoperative outcomes among patients undergoing abdominoperineal resection (APR). METHODS:This was a retrospective cohort study of patients from the National Surgical Quality Improvement Program who underwent APR for low rectal cancer from 2016 to 2019. Patients were stratified by operative approach: robotic, laparoscopic, and open APR (R-APR, L-APR, and O-APR). Emergent cases were excluded. The primary outcome was RMP. 30-day postoperative outcomes were also evaluated, using logistic regression analysis. RESULTS:Among 1,807 patients, 452 (25.0%) underwent R-APR, 474 (26.2%) L-APR, and 881 (48.8%) O-APR. No differences regarding RMP (13.5% R-APR vs. 10.8% L-APR vs. 12.3% O-APR, p = 0.44), distal margin positivity, positive nodes, readmission, or operative time were observed between operative approaches. Adjusted analysis confirmed that operative approach did not predict RMP (p > 0.05 for all). Risk factors for RMP included American Society of Anesthesiologists (ASA) classification III (ASA I-II ref; OR 1.46, p = 0.039), pT3-4 stage (T0-2 ref, OR 4.02, p < 0.001), pN2 stage (OR 1.98, p = 0.004), disseminated cancer (OR 1.90, p = 0.002), and lack of preoperative radiation (OR 1.98, p < 0.01). CONCLUSIONS:No difference in RMP was observed among R-APR, L-APR, and O-APR. Postoperatively, R-APR yielded greater benefit when compared to O-APR, but was comparable to that of L-APR. Minimally invasive surgery may be an appropriate option and worthy consideration for patients with distal rectal cancer requiring APR.

BACKGROUND:The purpose of this study was to assess the impact of surgical delays on short- and long-term survival among colon cancer patients. METHODS:Adult patients undergoing surgery for stage I, II, or III colon cancer were identified from the National Cancer Database (2010-2016). After categorization by wait times from diagnosis to surgery (<1 week, 1-3 weeks, 3-6 weeks, 6-9 weeks, 9-12 weeks, and >12 weeks), 30-day mortality, 90-day mortality, and 5-year overall survival were compared between patients both overall and after stratification by pathological disease stage. RESULTS:< .001 for all). Subgroup analysis after stratification by disease stage demonstrated that patients with stage III colon cancer were able to wait up to 9 weeks before exhibiting worse 5-year overall survival, compared to 6 weeks for patients with stage I or II disease. CONCLUSIONS:Colon cancer patients should undergo surgery 3-6 weeks after diagnosis, as all surgical delays beyond 6 weeks were associated with worse 30-day mortality, 90-day mortality, and 5-year overall survival.

INTRODUCTION:The increased use of minimally invasive surgery in the management of colorectal cancer has led to a renewed focus on how certain factors, such as insurance status, impact the equitable distribution of both laparoscopic and robotic surgery. Our goal was to analyze surgical wait times between robotic, laparoscopic, and open approaches, and to determine whether insurance status impacts timely access to treatment. METHODS:After IRB approval, adult patients from the National Cancer Database with a diagnosis of colorectal cancer were identified (2010-2016). Patients who underwent radiation therapy, neoadjuvant chemotherapy, had wait times of 0 days from diagnosis to surgery, or had metastatic disease were excluded. Primary outcomes were days from cancer diagnosis to surgery and days from surgery to adjuvant chemotherapy. Multivariable Poisson regression analysis was performed. RESULTS:Among 324,784 patients, 5.9% underwent robotic, 47.5% laparoscopic, and 46.7% open surgery. Patients undergoing robotic surgery incurred the longest wait times from diagnosis to surgery (29.5 days [robotic] vs. 21.7 [laparoscopic] vs. 17.2 [open], p < 0.001), but the shortest wait times from surgery to adjuvant chemotherapy (48.9 days [robotic] vs. 49.9 [laparoscopic] vs. 54.8 [open], p < 0.001). On adjusted analysis, robotic surgery was associated with a 1.46 × longer wait time to surgery (IRR 1.462, 95% CI 1.458-1.467, p < 0.001), but decreased wait time to adjuvant chemotherapy (IRR 0.909, 95% CI 0.905-0.913, p < 0.001) compared to an open approach. Private insurance was associated with decreased wait times to surgery (IRR 0.966, 95% CI 0.962-0.969, p < 0.001) and adjuvant chemotherapy (IRR 0.862, 95% CI 0.858-0.865, p < 0.001) compared to Medicaid. CONCLUSION:Though patients undergoing robotic surgery experienced delays from diagnosis to surgery, they tended to initiate adjuvant chemotherapy sooner compared to those undergoing open or laparoscopic approaches. Private insurance was independently associated not only with access to robotic surgery, but also shorter wait times during all stages of treatment.

AIM:We aim to compare machine learning with neural network performance in predicting R0 resection (R0), length of stay >Â 14Â days (LOS), major complication rates at 30Â days postoperatively (COMP) and survival greater than 1 year (SURV) for patients having pelvic exenteration for locally advanced and recurrent rectal cancer. METHOD:A deep learning computer was built and the programming environment was established. The PelvEx Collaborative database was used which contains anonymized data on patients who underwent pelvic exenteration for locally advanced or locally recurrent colorectal cancer between 2004 and 2014. Logistic regression, a support vector machine and an artificial neural network (ANN) were trained. Twenty per cent of the data were used as a test set for calculating prediction accuracy for R0, LOS, COMP and SURV. Model performance was measured by plotting receiver operating characteristic (ROC) curves and calculating the area under the ROC curve (AUROC). RESULTS:Machine learning models and ANNs were trained on 1147 cases. The AUROC for all outcome predictions ranged from 0.608 to 0.793 indicating modest to moderate predictive ability. The models performed best at predicting LOS >Â 14Â days with an AUROC of 0.793 using preoperative and operative data. Visualized logistic regression model weights indicate a varying impact of variables on the outcome in question. CONCLUSION:This paper highlights the potential for predictive modelling of large international databases. Current data allow moderate predictive ability of both complex ANNs and more classic methods.

BACKGROUND:Anal squamous cell carcinoma (ASCC) is the most common histological subtype of anal cancer. Rates have been observed to increase in recent years. Combined chemoradiotherapy (CCRT) is currently the gold standard of treatment. The aim of this study is to assess ASCC prevalence, treatment trends, and overall survival (OS) in the United States. METHODS:Patients diagnosed with stage I-IV ASCC were identified from the National Cancer Database from 2004 to 2015. The primary outcome was 5-year OS, which was analyzed using Kaplan-Meier survival curves, log-rank test, and Cox proportional hazards models. RESULTS:34,613 cases were included (stage I: 21.45%; II: 41.00%; III: 31.62%; IV: 5.94%), with an increasing trend in prevalence. CCRT was the most used treatment. Multimodal treatment, combining surgery with CCRT, offered the best OS rates for stage I, II, and IV cancers (I: 84.87%; II: 75.12%; IV: 33.08%), comparable with survival of stage III patients treated with CCRT (III: 61.14%). Radiation alone had the worse OS rates, and on adjusted analysis, radiation treatment alone had the greatest risk of mortality (I: hazard ratio, 2.01; 95% confidence interval, 1.14-3.54; P = 0.016; II: 2.05, 1.44-2.93, P < 0.001; IV: 1.99, 0.99-4.02, P = 0.054). CONCLUSIONS:ASCC has increased in prevalence, notably in stage III and IV disease. Although CCRT is the most commonly used treatment type for all stages of ASCC, multimodal treatment offers better OS in stages I, II, and IV. Treatment with radiation alone offers the worst OS no matter the stage and should no longer be used as a solitary treatment modality.

BACKGROUND: This study aims to directly compare measurements of tissue oxygenation obtained using the Intra.Ox (Vioptix Inc., Fremont, CA) near infrared spectrometer with the perfusion assessment of the indocyanine green (ICG)-based SPY Elite imaging system (Stryker Co., Kalamazoo, MI) in a porcine bowel model. METHODS: Two live minipigs underwent laparotomy and isolation of a 30-cm segment of a large bowel. Standardized oximetry measurements were taken along the segment of bowel immediately before, after, and serially for 30 minutes following transection. A 0.5 mg/kg dose of ICG was then injected intravenously and the SPY Elite system was used to visualize and quantify tissue perfusion. Pearson's correlation coefficients were calculated using the outcomes. RESULTS: = 0.645). CONCLUSION/CONCLUSIONS: Both the Intra.Ox and the SPY detected clinically relevant changes in bowel oxygenation following transection and ligation. The use of intravenous ICG dye did not appear to affect measurements of tissue oxygenation obtained using the Intra.Ox.

BACKGROUND:Current guidelines recommend extended venothromboembolism (VTE) prophylaxis for most patients following colorectal cancer surgery, but provider uptake has been limited. The purpose of this study was to identify thresholds for when such extended prophylaxis (ePpx) may be value-appropriate. METHODS:All colorectal cancer postoperative discharges were identified within a private payer administrative database (MarketScan® 2010-2014, IBM Truven Health Analytics). Outcomes of interest were VTE event rate, mortality, and overall costs of care. The data along with published literature were used as parameter estimations for a decision analysis model with probabilistic sensitivity analysis. RESULTS:We identified 22,463 colorectal cancer surgical patients (4.0% with ePpx) that served as the parameter estimates for the decision model with a VTE event rate of 0.2%. Decision analysis demonstrated that prescribing ePpx was dominated by usual practice with the former having higher probability-adjusted incremental costs ($1078.68 per person) and lower probability-adjusted benefits (- 0.000098 quality adjusted life years). Broad sensitivity analysis found that probability of a VTE event, bleeding case fatality rate, and probability of an ePpx-associated bleeding event were the primary effectors of the model. VTE event rates of greater than 3.0% benefited from prescribing ePpx to all patients. CONCLUSIONS:Very few patients are discharged on ePpx following colorectal cancer surgery despite its endorsement by national guidelines. A decision analysis model does not support the use of ePpx except in cases of markedly high VTE rates. Clinical guidance could be improved by further recognizing the role of risk stratification in the determination of high-risk patients requiring ePpx.