The Minimally Invasive Approach Confers Improved Outcomes in Frail Cancer Patients Undergoing Hepatectomy: An ACS NSQIP Analysis

BACKGROUND

Morbidity for liver resection has decreased and frail patients are undergoing surgery. The impact of the minimally invasive approach (MILR) is unknown. This study aims to evaluate the impact of approach on outcomes in frail patients.

METHODS

Elective hepatectomies from the 2014-2020 ACS-NSQIP Hepatectomy-specific PUF were reviewed. The five-factor modified frailty index (mFI-5) was used. It includes diabetes, hypertension, functional status, heart failure and dyspnea. Patients were considered frail if their mFI-5 was >2.

RESULTS

3,116 patients were included: 2117 (67.9%) minor and 999 (32%) major. There were 2, 254 open cases and 862 MILR. Postoperatively (minor) MILR conferred lower risk of prolonged LOS, discharge not-to-home, transfusion, major complications, and minor complications (P<0.05). Postoperatively (major) MILR had lower incidence of prolonged LOS and any complication (P<0.05). In the minor group, MILR remained independently predictive of lower rates of prolonged LOS (OR: 0.34, 95% CI: 0.28-0.42), discharge not-to-home (0.58, 0.41-0.84), transfusion (0.72, 0.54-0.96), major complication (0.78, 0.62-1) and any complication (0.73, 0.58-0.92). In the major group, MILR remained independently predictive of prolonged LOS (0.60, 0.40-0.89).

CONCLUSION

MILR conferred lower rates of complications in minor as well as shorter LOS in major hepatectomy. The minimally invasive approach to hepatectomy may benefit frail cancer patients.

The development of a deep learning model for automated segmentation of the robotic pancreaticojejunostomy

BACKGROUND

Minimally invasive surgery provides an unprecedented opportunity to review video for assessing surgical performance. Surgical video analysis is time-consuming and expensive. Deep learning provides an alternative for analysis. Robotic pancreaticoduodenectomy (RPD) is a complex and morbid operation. Surgeon technical performance of pancreaticojejunostomy (PJ) has been associated with postoperative pancreatic fistula. In this work, we aimed to utilize deep learning to automatically segment PJ RPD videos.

METHODS

This was a retrospective review of prospectively collected videos from 2011 to 2022 that were in libraries at tertiary referral centers, including 111 PJ videos. Each frame of a robotic PJ video was categorized based on 6 tasks. A 3D convolutional neural network was trained for frame-level visual feature extraction and classification. All the videos were manually annotated for the start and end of each task.

RESULTS

Of the 100 videos assessed, 60 videos were used for the training the model, 10 for hyperparameter optimization, and 30 for the testing of performance. All the frames were extracted (6 frames/second) and annotated. The accuracy and mean per-class F1 scores were 88.01% and 85.34% for tasks.

CONCLUSION

The deep learning model performed well for automated segmentation of PJ videos. Future work will focus on skills assessment and outcome prediction.

Impact of Neoadjuvant Treatment and Minimally Invasive Surgery on Perioperative Outcomes of Pancreatoduodenectomy: an ACS NSQIP Analysis

BACKGROUND

There is an increasing use of neoadjuvant treatment (NAT) for pancreatic cancer (PC) followed by minimally invasive pancreatoduodenectomy (MIPD). We evaluate the impact of the surgical approach on 30-day outcomes in PC patients who underwent NAT.

METHODS

Patients with PC who had NAT followed by MIPD or open pancreatoduodenectomy (OPD) were identified from a pancreatectomy-targeted dataset (2014-2020) of the National Surgical Quality Improvement Program. Comparisons were made between MIPD and OPD within NAT groups.

RESULTS

A total of 5588 patients were analyzed. Of those, 4907 underwent OPD and 476 underwent MIPD. In addition, 3559 patients received neoadjuvant chemotherapy alone and 1830 received neoadjuvant chemoradiation. In the chemotherapy-alone group, the MIPD subgroup had lower rates of any complication (38.2% vs. 45.8%, P = 0.005), but there were no differences in mortality (2.1% for MIPD vs 1.9% for OPD, P=0.8) or serious complication (11.8% for MIPD vs 15% for OPD, P=0.1). On multivariable analysis, MIPD was independently predictive of lower rates of any complication (OR: 0.74, 95% CI 0.6-0.93, P = 0.0009), CR-POPF (OR: 0.58, 95% CI 0.35-0.96, P = 0.04), and shorter LOS (estimate: -1.03, 95% CI -1.73 to -0.32, P = 0.004). In the chemoradiation group, patients undergoing MIPD had higher rates of preoperative diabetes (P < 0.05), but there were no significant differences in any outcomes between the two approaches in this group.

CONCLUSION

MIPD is safe and feasible after NAT. Patients having neoadjuvant chemotherapy alone followed by MIPD had lower rates of complications, shorter LOS, and fewer CR-POPFs compared to OPD.

Risk factors and outcomes in patients undergoing minimally invasive hepatectomy with unplanned conversion: a contemporary NSQIP analysis

BACKGROUND

Minimally invasive techniques are growing for hepatectomies. Laparoscopic and robotic liver resections have been shown to differ in conversions. We hypothesize that robotic approach will have decreased conversion to open and complications despite being a newer technique than laparoscopy.

METHODS

ACS NSQIP study using the targeted Liver PUF from 2014 to 2020. Patients grouped based on hepatectomy type and approach. Multivariable and propensity scored matching (PSM) was used to analyze the groups.

RESULTS

Of 7767 patients who underwent hepatectomy, 6834 were laparoscopic and 933 were robotic. The rate of conversions was significantly lower in robotic vs laparoscopic (7.8% vs 14.7%; p < 0.001). Robotic hepatectomy was associated with decreased conversion for minor (6.2% vs 13.1%; p < 0.001), but not major, right, or left hepatectomy. Operative factors associated with conversion included Pringle (OR = 2.09 [95% CI 1.05-4.19]; p = 0.0369), and a laparoscopic approach (OR = 1.96 [95% CI 1.53-2.52]; p < 0.001). Undergoing conversion was associated with increases in bile leak (13.7% vs 4.9%; p < 0.001), readmission (11.5% vs 6.1%; p < 0.001), mortality (2.1% vs 0.6%; p < 0.001), length of stay (5 days vs 3 days; p < 0.001), and surgical (30.5% vs 10.1%; p < 0.001), wound (4.9% vs 1.5%; p < 0.001) and medical (17.5% vs 6.7%; p < 0.001) complications.

CONCLUSION

Minimally invasive hepatectomy with conversion is associated with increased complications, and conversion is increased in the laparoscopic compared to a robotic approach.

Does CT scanning after pancreatoduodenectomy reduce readmission rates: an analysis of 900 resections at a high-volume center

BACKGROUND

Pancreatoduodenectomy (PD) remains associated with significant complication and readmission rates. Infection constitutes a significant proportion of morbidity. We aim to evaluate whether CT scans performed prior to discharge for suspected infection prevents readmission.

METHODS

A retrospective review of patients undergoing PD at a tertiary referral center from 2010 to 2018.

RESULTS

A total of 982 patients underwent PD: 74% had no clinical infection at the index admission. Of the non-infected patients, 59% exhibited leukocytosis, 27% underwent a CT scan, and 33.6% were readmitted. Of the non-infected patients, 148 (20.3%) experienced major complications, and this was the strongest predictor of readmission (OR: 10.5, [95% CI: 6.5-17], p = 0.0001). In the non-infected patients who had major complications, CT scanning was predictive of lower risk of readmission (OR: 0.38, [95% CI: 0.17-0.83], p = 0.015). Leukocytosis was also found to be predictive of lower risk of readmission (OR: 0.42, [95% CI: 0.18-0.98], p = 0.044). These findings did not hold true for those who had yet to experience major complications on their index admission.

CONCLUSION

CT scanning without evidence of infection was associated with reduction of readmission in the cohort with major complications and showed a trend towards preventing readmission in the overall cohort. Development of clinical algorithms to maximize the utility of this test is warranted.

Bile leak incidence, risk factors and associated outcomes in patients undergoing hepatectomy: a contemporary NSQIP propensity matched analysis

BACKGROUND

Despite advances in surgical technique, bile leak remains a common complication following hepatectomy. We sought to identify incidence of, risk factors for, and outcomes associated with biliary leak.

STUDY DESIGN

This is an ACS-NSQIP study. Distribution of bile leak stratified by surgical approach and hepatectomy type were identified. Univariate and multivariate factors associated with bile leak and outcomes were evaluated.

RESULTS

Robotic hepatectomy was associated with less bile leak (5.4% vs. 11.4%; p < 0.001) compared to open. There were no significant differences in bile leak between robotic and laparoscopic hepatectomy (5.4% vs. 5.3%; p = 0.905, respectively). Operative factors risk factors for bile leak in patients undergoing robotic hepatectomy included right hepatectomy [OR 4.42 (95% CI 1.74-11.20); p = 0.002], conversion [OR 4.40 (95% CI 1.39-11.72); p = 0.010], pringle maneuver [OR 3.19 (95% CI 1.03-9.88); p = 0.044], and drain placement [OR 28.25 (95% CI 8.34-95.72); p < 0.001]. Bile leak was associated with increased reoperation (8.7% vs 1.7%, p < 0.001), 30-day readmission (26.6% vs 6.8%, p < 0.001), 30-day mortality (2% vs 0.9%, p < 0.001), and complications (67.2% vs 23.4%, p < 0.001) for patients undergoing MIS hepatectomy.

CONCLUSION

While MIS confers less risk for bile leak than open hepatectomy, risk factors for bile leak in patients undergoing MIS hepatectomy were identified. Bile leaks were associated with multiple additional complications, and the robotic approach had an equal risk for bile leak than laparoscopic in this time period.

Mentorship and formal robotic proficiency skills curriculum improve subsequent generations' learning curve for the robotic distal pancreatectomy

BACKGROUND

Minimally invasive distal pancreatectomy is the accepted standard of care. The robotic distal (RDP) learning curve is 20-40 surgeries with operating time (ORT) as the most significant factor. This study evaluates how formal mentorship and a robotic skills curriculum impact the learning curve for subsequent generation surgeons.

METHODS

Consecutive RDP from 2008 to 2017 were evaluated. First Generation was two surgeons who started program without training or mentorship. Second Generation was the two surgeons who joined the program with mentorship. Third Generation was fellows who benefited from both formal training and mentorship. Multivariable models (MVA) were performed for ORT, clinically relevant pancreatic fistula (CR-POPF), and major complications (Clavien≥3).

RESULTS

A total of 296 RDP were performed of which 187 did not include other procedures: First Generation (n = 71), Second Generation (n = 50), and Third Generation (n = 66). ORT decreased by generation (p < 0.001) without any differences in CR-POPF or Clavien≥3. On MVA, earlier generation (p = 0.019), pre-operative albumin (p = 0.001) and pancreatic adenocarcinoma (p = 0.019) were predictive of ORT. Increased BMI (p = 0.049) and neoadjuvant therapy (p = 0.046) were predictive of CR-POPF. Fellow participation at the console increased over time.

CONCLUSION

Formal mentorship and a skills curriculum decreased the learning curve and complications were largely dependent on patient factors.

Development of a Novel Pancreatoduodenectomy-Specific Risk Calculator: an Analysis of 10,000 Patients

BACKGROUND

Pancreatoduodenectomy (PD) is often performed in frail patients and is associated with significant morbidity. The five-factor modified frailty index (mFI-5) has been utilized to predict adverse postoperative outcomes, but has not been tested in PD. We aimed to develop risk tools to generate and predict 30-day outcomes after PD and compare their performance with the mFI-5. Risk tools were then used to generate a PD-specific calculator.

METHODS

Elective PDs from the 2014-2016 ACS NSQIP® Procedure Targeted Pancreatectomy PUFs were identified. Multivariable logistic regression models were constructed to predict postoperative mortality, any complication, serious complication, clinically relevant postoperative pancreatic fistula (CR-POPF), and discharge not-to-home. Predictive accuracy was evaluated through repeated stratified tenfold cross-validation and compared to the mFI-5.

RESULTS

Nine thousand eight hundred sixty-seven PDs were captured. Nine risk factors were retained: sex, age, BMI, DM, HTN, ASA classification, pancreatic duct size, gland texture, and adenocarcinoma. Cross-validated C-indices ranged from 0.49 to 0.61 for the mFI-5 and 0.63 to 0.75 for our risk models. The best-performing model was for discharge not-to-home (C = 0.75), and the model delivering the largest increase in predictive accuracy was for CR-POPF (C_mFI-5/Model = 0.49/0.70). A user-friendly risk calculator was created predicting the five outcomes of interest.

CONCLUSION

We have created a PD-specific risk calculator that outperforms the mFI-5. This calculator may serve as a useful adjunct in shared decision-making for patients and surgeons.

500 Minimally Invasive Robotic Pancreatoduodenectomies: One Decade of Optimizing Performance

OBJECTIVES

This study aims to present the outcomes of our decade-long experience of robotic pancreatoduodenectomy and provide insights into successful program implementation.

BACKGROUND

Despite significant improvement in mortality over the past 30 years, morbidity following open pancreatoduodenectomy remains high. We implemented a minimally invasive pancreatic surgery program based on the robotic platform as one potential method of improving outcomes for this operation.

METHODS

A retrospective review of a prospectively maintained institutional database was performed to identify patients who underwent robotic pancreatoduodenectomy (RPD) between 2008 and 2017 at the University of Pittsburgh.

RESULTS

In total, 500 consecutive RPDs were included. Operative time, conversion to open, blood loss, and clinically relevant postoperative pancreatic fistula improved early in the experience and have remained low despite increasing complexity of case selection as reflected by increasing number of patients with pancreatic cancer, vascular resections, and higher Charlson Comorbidity scores (all P<0.05). Operating room time plateaued after 240 cases at a median time of 391 minutes (interquartile rang 340-477). Major complications (Clavien >2) occurred in less than 24%, clinically relevant postoperative pancreatic fistula in 7.8%, 30- and 90-day mortality were 1.4% and 3.1% respectively, and median length of stay was 8 days. Outcomes were not impacted by integration of trainees or expansion of selection criteria.

CONCLUSIONS

Structured implementation of robotic pancreatoduodenectomy can be associated with excellent outcomes. In the largest series of RPD, we establish benchmarks for the surgical community to consider when adopting this approach.

Association of Mentorship and a Formal Robotic Proficiency Skills Curriculum With Subsequent Generations' Learning Curve and Safety for Robotic Pancreaticoduodenectomy

Importance

Learning curves are unavoidable for practicing surgeons when adopting new technologies. However, patient outcomes are worse in the early stages of a learning curve vs after mastery. Therefore, it is critical to find a way to decrease these learning curves without compromising patient safety.

Objective

To evaluate the association of mentorship and a formal proficiency-based skills curriculum with the learning curves of 3 generations of surgeons and to determine the association with increased patient safety.

Design, Setting, and Participants

All consecutive robotic pancreaticoduodenectomies (RPDs) performed at the University of Pittsburgh Medical Center between 2008 and 2017 were included in this study. Surgeons were split into generations based on their access to mentorship and a proficiency-based skills curriculum. The generations are (1) no mentorship or curriculum, (2) mentorship but no curriculum, and (3) mentorship and curriculum. Univariable and multivariable analyses were used to create risk-adjusted learning curves by surgical generation and to analyze factors associated with operating room time, complications, and fellows completing the full resection. The participants include surgical oncology attending surgeons and fellows who participated in an RPD at University of Pittsburgh Medical Center between 2008 and 2017.

Main Outcomes and Measures

The primary outcome was operating room time (ORT). Secondary outcomes were postoperative pancreatic fistula and Clavien-Dindo classification higher than grade 2.

Results

We identified 514 RPDs completed between 2008 and 2017, of which 258 (50.2%) were completed by first-generation surgeons, 151 (29.3%) were completed by the second generation, and 82 (15.9%) were completed by the third generation. There was no statistically significant difference between groups with respect to age (66.3-67.3 years; P = .52) or female sex (n = 34 [41.5%] vs n = 121 [46.9%]; P = .60). There was a significant decrease in ORT (P < .001), from 450.8 minutes for the first-generation surgeons to 348.6 minutes for the third generation. Additionally, across generations, Clavien-Dindo classification higher than grade 2 (n = 74 [28.7%] vs n = 30 [9.9%] vs n = 12 [14.6%]; P = .01), conversion rates (n = 18 [7.0%] vs n = 7 [4.6%] vs n = 0; P = .006), and estimated blood loss (426 mL vs 288.6 mL vs 254.7 mL; P < .001) decreased significantly with subsequent generations. There were no significant differences in postoperative pancreatic fistula.

Conclusions and Relevance

In this study, ORT, conversion rates, and estimated blood loss decreased across generations without a concomitant rise in adverse patient outcomes. These findings suggest that a proficiency-based curriculum coupled with mentorship allows for the safe introduction of less experienced surgeons to RPD without compromising patient safety.

Uretero-Arterial Fistula: A Case Report and Review of the Literature

Uretero-arterial fistula (UAF) is a rare complication of either aneurysmal disease primarily or pelvic inflammation secondary to urologic, oncologic, or vascular interventions. Diagnosis can be difficult to confirm and treatment may need to proceed on high index of suspicion alone. We present the case of a 56-year-old woman suffering from intermittent hematuria after laser lithotripsy leading to UAF between her left ureter and left Dacron aortobifemoral bypass limb. The fistula was successfully treated with endovascular intervention.

Serum CA19-9 Response to Neoadjuvant Therapy Predicts Tumor Size Reduction and Survival in Pancreatic Adenocarcinoma

BACKGROUND

The optimal cutoffs for carbohydrate antigen 19-9 (CA19-9) response after neoadjuvant therapy (NT) for pancreatic adenocarcinoma (PDAC) are not well characterized. This study aimed to analyze the relationship of serum CA19-9 to other markers of response and to identify thresholds correlating to outcomes.

METHODS

A retrospective review of resected PDAC patients from 2010 to 2017 at an academic tertiary referral center was conducted.

RESULTS

The analysis enrolled 250 subjects. Normalization and multiple cutoff points for CA19-9 response were assessed. Normalization was not associated with improved survival (35.17 vs. 29.43 months; p = 0.173). Although a response 45% or higher was associated with longer survival (35 vs. 20 months; p = 0.018), a response of 85% or higher was optimal (55.7 vs. 25.97 months; p < 0.0001). A response of 85% or higher remained a strong independent predictor of survival [hazard ratio (HR), 0.47; p = 0.007]. Subjects with a response of 85% or higher had received more NT cycles [3 (range 2-6) vs. 3 (range 2-4) cycles; p = 0.006] and fewer adjuvant cycles [4 (range 3-6) vs. 5 (range 3-6) cycles; p = 0.027]. Reduction in T-size correlated with a drop in CA19-9 and a size reduction of 25% or higher (56.97 vs. 28.17 months; p = 0.016) improved survival. A serum CA19-9 response of 85% or higher was a strong independent predictor of a reduction in T-size of 25% or higher (HR 2.40; p = 0.007).

CONCLUSION

A CA19-9 response of 85% or higher is the optimal threshold for predicting survival. It is predictive of T-size reduction. Future NT trials should incorporate CA19-9 response as an end point.

Prognostic Value of the Systemic Immune-Inflammation Index (SII) After Neoadjuvant Therapy for Patients with Resected Pancreatic Cancer

BACKGROUND

The systemic immune-inflammation index (SII), calculated using absolute platelet, neutrophil, and lymphocyte counts, has recently emerged as a predictor of survival for patients with pancreatic ductal adenocarcinoma (PDAC) when assessed at diagnosis. Neoadjuvant therapy (NAT) is increasingly used in the treatment of PDAC. However, biomarkers of response are lacking. This study aimed to determine the prognostic significance of SII before and after NAT and its association with the pancreatic tumor biomarker carbohydrate-antigen 19-9 (CA 19-9).

METHODS

This study retrospectively analyzed all PDAC patients treated with NAT before pancreatic resection at a single institution between 2007 and 2017. Pre- and post-NAT lab values were collected to calculate SII. Absolute pre-NAT, post-NAT, and change in SII after NAT were evaluated for their association with clinical outcomes.

RESULTS

The study analyzed 419 patients and found no significant correlation between pre-NAT SII and clinical outcomes. Elevated post-NAT SII was an independent, negative predictor of overall survival (OS) when assessed as a continuous variable (hazard ratio [HR], 1.0001; 95% confidence interval [CI] 1.00003-1.00014; p = 0.006). Patients with a post-NAT SII greater than 900 had a shorter median OS (31.9 vs 26.1 months; p = 0.050), and a post-NAT SII greater than 900 also was an independent negative predictor of OS (HR, 1.369; 95% CI 1.019-1.838; p = 0.037). An 80% reduction in SII independently predicted a CA 19-9 response after NAT (HR, 4.22; 95% CI 1.209-14.750; p = 0.024).

CONCLUSION

Post-treatment SII may be a useful prognostic marker in PDAC patients receiving NAT.

Correction to: Serum CA19-9 Response to Neoadjuvant Therapy Predicts Tumor Size Reduction and Survival in Pancreatic Adenocarcinoma

In the original article, Caroline J. Rieser's last name is spelled wrong. It is correct as reflected here.

Impact of adjuvant chemotherapy regimen on survival outcomes in immunohistochemical subtypes of ampullary carcinoma

BACKGROUND AND OBJECTIVES

Ampullary adenocarcinoma (AA) is classified by immunohistochemical (IHC) subtypes into intestinal (IN), pancreatobiliary (PB), and ambiguous (AM). The impact of adjuvant therapy on IHC subtype and disease stage is unclear. We examined the effect of adjuvant chemotherapy regimen on survival of ampullary cancers by IHC subtype and disease stage.

METHODS

Review of pancreatoduodenectomy (PD) performed for AA between 2005 and 2013 at a single center. The impact of regimen on IHC subtype and stage was analyzed.

RESULTS

One hundred and twenty-one patients were subtyped: IN = 32%, PB = 48%, and AM = 20% with overall survival of 45.6, 31.3, and 46.9 months, respectively. PB had higher pathologic T-stage, positive lymph node disease, and perineural and lymphovascular invasion (P < .05). 5-Fluorouracil (FU)-based adjuvant therapy improved survival compared to no treatment (87.4 vs 32.1 months; P = .046), and receipt of 5-FU emerged as an independent predictor of improved survival (hazard ratio [HR] 0.244; P = .031) regardless of subtype. 5-FU was superior to Gemcitabine in advanced-stage disease (stage IIB and III vs I+IIA, HR: 0.35; P < .05).

CONCLUSIONS

Adjuvant therapy with 5-FU confers a survival benefit in patients with advanced-stage AA regardless of subtype. The impact of various chemotherapy regimens on subtypes of ampullary cancer warrants further investigation.

Age-related risks of complications after distal pancreatectomy for neuroendocrine and cystic neoplasms

BACKGROUND

Distal pancreatic neuroendocrine tumors (PNET) and pancreatic cystic neoplasms (PCN) are often incidentally found in older adults, requiring careful consideration between operative management and watchful waiting. This study analyzes the short-term complications associated with distal pancreatectomy (DP) for PNET and PCN in older adults to inform clinical decision-making.

METHODS

Patients undergoing DP for PNET and PCN were analyzed using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database and the pancreatectomy procedure-targeted dataset. Associations between decade of age and 30-day outcomes were evaluated.

RESULTS

1626 patients were analyzed from 2014 to 2015. 692 (42.6%) were younger than 60 years, 507 (31.2%) were sexagenarians, 342 (21.0%) were septuagenarians, and 85 (5.2%) were octogenarians. Minimally invasive approaches were used in 62.7%. While septuagenarians and octogenarians constituted 26.3% of the cohort, they were affected by 55.6% of reintubations, 66.7% of failures to wean, 82.4% of myocardial infarctions, and 57.1% of septic shock. Septuagenarians and octogenarians had longer hospital stays, as compared to those younger than 60 years.

CONCLUSION

Septuagenarians and octogenarians are disproportionately affected by perioperative complications after DP for PNET and PCN. Careful patient selection and thorough counseling should be provided when surgery is considered.

Predictors of neoadjuvant chemotherapy efficacy in localized pancreatic head cancer: A 15-year experience at a high volume pancreatic cancer center

e15724

Background: Surgery followed by adjuvant chemotherapy (SF) is the standard approach for localized pancreatic adenocarcinoma (PC). Although neoadjuvant chemotherapy (NC) is being increasingly used, its efficacy remains debatable. This study aims to identify baseline factors associated with benefit from NC as compared to SF for PC. Methods: Review of a prospectively maintained database of resected localized head PC undergoing NC or SF at a tertiary referral hospital from 2002 to 2018. Baseline factors investigated included clinical stage by EUS (7th edition AJCC), CT vessel involvement, and serum CA19-9. Baseline CA19-9 levels were grouped into mild (37-200), moderate (200-900), and marked elevations (≥ 900). Patients with baseline serum CA19-9 < 37 IU/dL were excluded from the analysis. Results: In total, 269 patients were evaluated: 165 (61%) received NC and 104 (39%) had SF. Mean patient age was 68.3 years and 50% were females. Of these, 42% had stage IIB, 44% had N1 disease, and 38% had vascular involvement. Compared to SF approach, NC was associated with improved median survival for CA19-9 of < 200 IU/dL (30 vs 21 months, p = 0.08), 200-900 IU/dL (32 vs 17 months, p < 0.01), and the ≥900 IU/dL (33 vs 12 months, p < 0.01) groups. On Cox regression, NC was the only baseline factor predictive of improved survival in the groups with moderate (HR:0.35,[95% CI:0.20-0.62];p < 0.01) and marked (HR:0.36,[95% CI:0.18-0.71];p < 0.01) pretreatment CA19-9 elevation. Notably, NC was not predictive of improved survival in the cohort with mild CA19-9 elevation. Conclusions: This study identifies a cohort of patients that benefits from neoadjuvant chemotherapy over a surgery first approach for localized pancreatic head cancer. Elevation in baseline CA19-9 may serve as the strongest indicator for chemotherapy before resection.

Methodology for Developing an Educational and Research Video Library in Minimally Invasive Surgery

OBJECTIVE

Explore the methods used and costs necessary for the creation and maintenance of a surgical video library with an emphasis on its applications in surgical education and scholarship.

DESIGN

A methodology paper highlighting how to develop and utilize a surgical video library for trainee operative preparation, development of research projects, and surgeon credentialing.

SETTING

The study was conducted at the University of Pittsburgh Medical Center, a tertiary care medical center.

PARTICIPANTS

Not applicable.

RESULTS

The video library includes all recorded robotic operations performed by the Division of Surgical Oncology at the University of Pittsburgh from 2010 to 2018. It includes 929 videos of which 110 selected videos are uploaded for trainee review online to prepare for upcoming operations. These procedures are broken into steps to create intraoperative time metrics for trainee integration. Fellows operated from console in 85% of robotic cases and all 30 fellows could obtain robotic privileges based on case logs. To date 102 short scholarly videos have been created: 7 for manuscripts, 13 as video submissions, 27 in book chapters, and 55 for presentations. Three papers have been published using video review to determine clinical outcomes with four more under evaluation. The cost of the program is <$10,000 annually.

CONCLUSIONS

Video libraries can be efficiently created utilizing intraoperative recorders for minimally invasive surgery. Breaking surgeries into distinct steps can aid in deliberate integration and answer clinical questions. Overall video libraries are cost effective tools for trainee education, research, and ultimately surgeon credentialing.

Crowdsourced Assessment of Inanimate Biotissue Drills: A Valid and Cost-Effective Way to Evaluate Surgical Trainees

OBJECTIVE

Providing feedback to surgical trainees is a critical component for assessment of technical skills, yet remains costly and time consuming. We hypothesize that statistical selection can identify a homogenous group of nonexpert crowdworkers capable of accurately grading inanimate surgical video.

DESIGN

Applicants auditioned by grading 9 training videos using the Objective Structured Assessment of Technical Skills (OSATS) tool and an error-based checklist. The summed OSATS, summed errors, and OSATS summary score were tested for outliers using Cronbach's Alpha and single measure intraclass correlation. Accepted crowdworkers then submitted grades for videos in 3 different compositions: full video 1× speed, full video 2× speed, and critical section segmented video. Graders were blinded to this study and a similar statistical analysis was performed.

SETTING

The study was conducted at the University of Pittsburgh Medical Center (Pittsburgh, PA), a tertiary care academic teaching hospital.

PARTICIPANTS

Thirty-six premedical students participated as crowdworker applicants and 2 surgery experts were compared as the gold-standard.

RESULTS

The selected hire intraclass correlation was 0.717 for Total Errors and 0.794 for Total OSATS for the first hire group and 0.800 for Total OSATS and 0.654 for Total Errors for the second hire group. There was very good correlation between full videos at 1× and 2× speed with an interitem statistic of 0.817 for errors and 0.86 for OSATS. Only moderate correlation was found with critical section segments. In 1 year 275hours of inanimate video was graded costing $22.27/video or $1.03/minute.

CONCLUSIONS

Statistical selection can be used to identify a homogenous cohort of crowdworkers used for grading trainees' inanimate drills. Crowdworkers can distinguish OSATS metrics and errors in full videos at 2× speed but were less consistent with segmented videos. The program is a comparatively cost-effective way to provide feedback to surgical trainees.

CA19-9 on Postoperative Surveillance in Pancreatic Ductal Adenocarcinoma: Predicting Recurrence and Changing Prognosis over Time

BACKGROUND

Serum carbohydrate antigen 19-9 (CA19-9) correlates with response to therapy and overall survival (OS) for patients with pancreatic ductal adenocarcinoma (PDAC). This study aimed to define the chronologic relationship between CA19-9 elevation and radiographic recurrence to develop a model that can predict the risk of recurrence (RFS) and prognosis during interval surveillance for patients with resected PDAC.

METHODS

A retrospective review examined patients undergoing surgery for pancreatic adenocarcinoma from January 2010 to May 2016. Their CA19-9 levels were classified at diagnosis, after surgery, and at 6-month surveillance intervals. Recurrence was defined by radiographic evidence. The CA19-9 levels were correlated with RFS and OS at every time point using multivariate analysis.

RESULTS

The study examined 525 patients. Five patterns of CA19-9 were identified: normal ("nonsecretors," 18.5%), always elevated, and high at diagnosis but normal after resection involving three patterns with varied behavior during surveillance. These five patterns had implications for RFS and OS. When elevation of CA19-9, as assessed at 6-month intervals, was analyzed relative to detection of radiographic disease, CA19-9 had poor positive predictive value (average, 35%) but high negative predictive value (average, 92%) for radiographic recurrence. Conditional RFS showed that CA19-9 elevation did not equal radiographic recurrence but predicted subsequent RFS. Additionally, conditional OS showed that CA19-9 elevation alone was predictive at each time point.

CONCLUSION

This study showed that CA19-9 patterns beyond the post-resection period predict RFS and OS. High CA19-9 frequently is discordant with recurrence on imaging and may precede it by more than 6 months. At each surveillance interval, CA19-9 is predictive of prognosis, which may help in counseling patients and could be used to direct protocols of salvage chemotherapy.

Roux-en-Y gastric bypass, sleeve gastrectomy, or one anastomosis gastric bypass as rescue therapy after failed adjustable gastric banding: a multicenter comparative study

BACKGROUND

To date, laparoscopic adjustable gastric banding remains the third most commonly performed surgical procedure for weight loss. Some patients fail to get acceptable outcomes and undergo revisional surgery at rates ranging from 7% to 60%. Roux-en-Y gastric bypass (RYGB), one anastomosis gastric bypass (OAGB), and sleeve gastrectomy (SG) are among the most common salvage options for failed laparoscopic adjustable gastric banding.

OBJECTIVE

To compare the outcomes of converting failed laparoscopic adjustable gastric banding to RYGB, OAGB, or SG.

METHODS

Data collected from 7 experienced bariatric centers around the world were retrospectively collected, reviewed, and analyzed. Final body mass index (BMI), change in BMI, percentage excess BMI loss, and major complications with particular attention to leaks, hemorrhage, and mortality were reported.

RESULTS

Of 1219 patients analyzed, 74% underwent RYGB, 16% underwent OAGB, and 10% underwent SG after banding failure. The mean age was 38 years (±10 yr), and 82% of patients were women. The mean follow-up was 33 months. The follow-up rate was 100%, 87%, and 52% at 1, 3, and 5 years, respectively. At the latest follow-up, percentage excess BMI loss >50% was achieved by 75% of RYGB, 85% of OAGB, and 67% of SG patients. Postoperative complications occurred in 13% of patients after RYGB, 5% after OAGB, and 15% after SG.

CONCLUSION

Our data show that it is possible to achieve or maintain significant weight loss with an acceptable complication rate with all 3 surgical options.