THE USE OF AXIAL IMAGING IN THE EARLY POSTOPERATIVE PERIOD FOLLOWING PANCREATECTOMY: IS IT EVER “TOO EARLY”?

Background
Management of patients with margin negative, T1-T3, N0 (stage IB–IIIA), resected gallbladder cancer (GBC) remains poorly defined. Current guidelines consider observation, chemotherapy (CT), and chemoradiation (CRT) as options. The current study investigates the impact of CT/CRT on overall survival (OS) in these patients.

Methods
Patients with R0 resected stage IB–IIIA GBC were identified within the National Cancer Database. Relevant patient, tumor, and treatment data were analyzed. Multiple logistic regressions were performed for factors associated with receipt of hepatectomy and CT/CRT. Kaplan-Meier analysis for OS was performed. Logrank tests compared OS between treatment groups. Stage-by-stage multivariable Cox regressions assessed the impact of CT and CRT on mortality while adjusting for other risk factors.

Results
Of 2,070 patients identified, resection included cholecystectomy in 950 (45.9%) and cholecystectomy+hepatectomy in 1,120 (54.1%). Overall, 1,419 had resection alone (68.6%), 313 had resection+CT (15.1%), and 338 had resection+CRT (16.3%). Hepatectomy was more commonly performed at academic facilities (OR 2.538; CI 1.969-3.281; P<.01) and for stage IIIA disease (OR 2.670; CI 1.913-3.742; P<.01). Hepatectomy was associated with improved OS in stage IB-IIIA disease (logrank P<.01). Receipt of CT or CRT was associated with stage IIA-IIIA disease, hepatectomy, and private insurance (all P<.05). In univariable analysis, both CT (logrank P<.05) and CRT (logrank P<.01) were associated with improved OS only in patients with stage IIA-IIB disease who did not undergo hepatectomy. Within a multivariable Cox regression model adjusting for age, gender, comorbidities, insurance status, facility type, and tumor grade, only CRT was associated with decreased mortality for patients with stage IIA-IIB disease who did not undergo hepatectomy (HR 0.609; CI 0.400-0.888; P<.05).

Conclusions
The current national study demonstrates the importance of adequate surgical therapy for patients with potentially curable GBC. Among these patients, CT and CRT were not associated with improved OS. However, CRT did result in an OS benefit in patients who did not undergo optimal resection. Prospective trials focused on CT/CRT for adequately resected, node negative GBC are needed.

Introduction: The robotic platform is gaining a wider adoption in minimally invasive liver surgery, especially for complex resections. Difficulty scoring systems are a useful tool to predict the technical difficulty of each hepatectomy and to guide surgeons during operative planning. To date, difficulty scoring systems are available for laparoscopic but not for robotic hepatectomy. Because the robotic platform has unique technical characteristics when compared to the laparoscopic method in liver surgery, there is a need to design a difficulty scoring system specifically for robotic hepatectomy which we propose herein.
Methods: 328 consecutive patients undergoing robotic hepatectomy from a single institution were utilized to develop a difficulty scoring system. Patients requiring concomitant colorectal resection or undergoing planned conversion to ‘open’ hepatectomy were excluded. Operative duration and estimated blood loss (EBL) were utilized, among many markers, reflecting operative difficulty. Multivariate analysis was applied to determine the relationships between these markers of difficulty and relevant clinical factors.
Clinical factors utilized in the proposed robotic DSS were compiled after a comprehensive review of factors used by previously published IWATE, Institut Mutualiste Montsouris, and Southampton laparoscopic difficulty scoring systems, in addition to other relevant variables.
Every patient was given a corresponding robotic difficulty score and subsequently categorized into four groups based on their cumulative score (Group 1: 1-6, Group 2: 7-10, Group 3: 11-15, Group 4: 16-21). Patients’ perioperative outcomes between each group was compared. Data are presented as median(mean±SD). Significance was accepted at p≤0.05.
Results: Past surgical history, use of neoadjuvant chemotherapy, tumor location, tumor size, tumor type, proximity to major vessels, extent of parenchymal resection, need for portal lymphadenectomy, and need for biliary resection/reconstruction were significantly correlated with either operative time and/or EBL. Upon further analysis, 22 (7%) patients received a score of 1-6, 143 (44%) patients received a score of 7-10, 154 (47%) received a score of 11-15, and 9 (3%) patients received a score greater than 16. When stratified by difficulty score, there was a significant difference in patients’ Childs-Pugh score, rate of major resection, 30-day readmission, total cost, and fixed direct cost(p≤0.05).
Conclusions: Herein, we propose a novel difficulty scoring system (Table 1) for robotic hepatectomy utilizing clinical factors mentioned, which reflect technical difficulty. A further validation of this robotic difficulty scoring system is needed using larger multi-institutional collaborative datasets.

Introduction: Access to specialized oncologic care is critical for improved prognosis in pancreatic cancer. Studies have shown that patients living in high Area Deprivation Index (ADI) regions are more likely to have low median income, lack means of transportation and access to basic health care facilities. We aim to understand the impact of residing in areas of high ADI on access to care in patients with pancreatic cancer.

Methods: We performed a retrospective review of patients identified from an institutional database, from October 2017 to August 2022. Individual patient ADI ranks were assigned using a 9-digit zip code and divided into terciles, with high tercile indicating the most disadvantaged group. Chi-square test statistics are reported for factors impacted by ADI rank.

Results: 52.6% (n=105) of 223 patients resided in areas of medium to high deprivation. 50% of Black patients (n=24) lived in areas of high ADI compared with 10% of White patients (n=17). Residence in areas of higher ADI was significantly associated with decreased receipt of guideline concordant care (GCT) with 10% (n=14) of patients residing in high ADI areas receiving GCT, compared with 32.1% (n=45) of patients in medium ADI, and 57.9% (n=81) of patients in low ADI (χ2=18.69 p<0.001). Residence in areas of medium to high ADI were also associated with increased Emergency Department utilization (χ2=14.70 p=0.001), and with decreased prior cancer screening uptake (13.1% in high ADI compared with 29% in medium ADI and 57.9% in low ADI, χ2=8.08 p=0.018). There was no significant difference in access to primary care based on residence (χ2=4.21 p=0.122).

Conclusion: Patients with pancreatic cancer residing in areas of high ADI are less likely to receive evidence-based guideline-concordant care and are more likely to have higher ED utilization. Patients residing in high ADI areas are more likely to be Black and have lower cancer screening uptake, potentially indicating decreased access to healthcare. Further studies are needed to assess the impact of residence on delivery of GCT, as well as to assess specific barriers that may be addressed through targeted interventions.

Introduction
Treatment of pancreatic ductal adenocarcinoma (PDAC) can be challenging in the setting of liver cirrhosis. We sought to evaluate the effect of cirrhosis on surgical complications, time to recurrence (TTR), and overall survival (OS) following the resection of PDAC in cirrhotic patients.

Methods
This was a single-center retrospective analysis of patients who underwent curative-intent resection for PDAC. Patients with a preoperative diagnosis of cirrhosis were compared to non-cirrhotic patients. Propensity score matching (PSM) was performed using the AJCC 8^th edition TNM stage, preoperative serum level of CA 19-9, and surgical margin status. Postoperative complications, TTR, and OS were analyzed before and after PSM.

Results
We identified 1243 patients who fulfilled the inclusion criteria. Thirty-six patients (2%) had liver cirrhosis (80% Child-Pugh A, 17% Child-Pugh B, 3% Child-Pugh C) with an overall median MELD-Na of 11 (IQR 7.5-12.5) and were matched to 36 non-cirrhotic controls. After PSM, patients with cirrhosis had higher estimated blood loss (1000 ml, IQR 675-1600 vs. 650 ml, IQR 390-900, p=0.003), higher rates of major complications (42% vs. 12%, p=0.005), and a higher 30-day readmission rate (28% vs. 9%, p=0.042) than non-cirrhotic patients. No differences were found in TTR between both groups. However, median OS was substantially lower in patients with cirrhosis (14 months, 95% CI 11-20 vs. 33 months, 95% CI 20-not reached, p<0.001).

Conclusion
Patients with liver cirrhosis are at an increased risk of postoperative complications and associated with substantially worse survival following surgical resection of PDAC. Therefore, individualized risk-benefit assessment and multidisciplinary counseling are needed prior to consideration of pancreatectomy.

Background
The step-up approaches – percutaneous or endoscopic catheter drainage followed, if necessary, by minimally invasive surgical or endoscopic necrosectomy – are leading approaches for infected necrotizing pancreatitis after trials showed reduced morbidity compared to traditional open surgical necrosectomy. However, both the surgical and endoscopic step-up approaches are associated with the need for repeat interventions and prolonged hospitalization. The surgical transgastric approach to pancreatic necrosectomy (direct STGN) has been introduced for retrogastric collections to overcome these shortcomings. In this study, we aimed to describe outcomes for patients who have undergone direct STGN for infected necrosis.

Methods
This observational cohort study included patients who underwent direct STGN for infected pancreatic necrosis between 2011 and 2022 at two centers. Patients with sterile necrosis, a prior pancreatic intervention (i.e., interventional radiology drain, endoscopic transgastric drain) or a laparotomy after diagnosis of pancreatitis prior to STGN were excluded from analysis. Primary outcomes included mortality, length of hospital and intensive care unit (ICU) stay, new-onset organ failure, repeat interventions, readmissions and time to disease resolution (defined as the date of removal of the last percutaneous drain or date of discharge of initial admission or readmission).

Results
Forty-five patients underwent direct STGN for infected necrosis (21 open, 24 laparoscopic). The median age was 57 years (IQR 46–62) and 14 patients (31.1%) were female. On pre-operative imaging, 29 (64.4%) patients had more than 30% pancreatic necrosis. The median time to intervention from diagnosis of acute pancreatitis was 48 days (IQR 32–70). Pancreatitis-related mortality rate was 6.7% (n = 3). Following direct STGN, the median length of hospital stay was 8 days (IQR 6–17). ICU admission was required in 23 patients (51.1%) for a median of 1 day (IQR 0–3). New-onset organ failure occurred in 8 patients (17.8%). In 10 patients (22.2%), a surgical drain was left intraoperatively. Two patients (4.4%) required re-intervention (1 interventional radiology drain, duration 101d; 1 endoscopic transgastric drain). A pancreatic fistula was seen in one patient (2.2%). Eleven patients (24.4%) were readmitted. The median time to disease resolution was 6 days (IQR 6–22).

Conclusion
When anatomically possible, direct STGN offers an opportunity to treat patients with infected necrosis in a single procedure resulting in a short time to disease resolution. Due the low need for repeat interventions and few pancreatic fistulas, the direct STGN challenges the step-up approach for infected necrosis. Additional research is required to compare the different approaches.

Lymph node metastasis is a well-established negative predictor of survival in PDAC, however the independent predictive value of specific lymph nodes remains equivocal. In particular, the importance of a positive common hepatic artery lymph node (CHALN+) (station 8a) has been long debated. Earlier studies showed significantly impaired survival associated with CHALN+, but others have not. We sought to contribute to this debate utilizing a more contemporary cohort where the CHALN was prospectively identified.

We analyzed a cohort of patients with head of the pancreas PDAC who underwent pancreatoduodenectomy from 2010 to 2017. A single surgeon prospectively identified the CHALN intraoperatively and submitted it separately for permanent pathological analysis. Actual 5-year overall survival (OS) was available and calculated. Overall and disease-free survival (DFS) beyond 5 years was estimated per Kaplan Meier analysis and compared via log-rank test.

217 patients had excision of the CHALN in addition to other peripancreatic lymph nodes (PPLN). 75 (35%) were PPLN-/CHALN-, 125 (58%) were PPLN+/CHALN-, and 17 (7%) were PPLN+/CHALN+. No patients had CHALN+ in absence of PPLN+. Actual 5-year OS was 50% in PPLN-/HALN-, 14% in PPLN+/HALN-, and 20% in PPLN+/HALN+. Kaplan-Meier estimates of median OS showed a significant difference between PPLN-/CHALN- and PPLN+/CHALN- (47 vs 22 months; p<0.01) and between PPLN-/CHALN- and PPLN+/CHALN+ (47 vs 25 months; p=0.02) (Figure 1). However, there was no significant difference in OS between PPLN+/CHALN- and PPLN+/CHALN+ (p=0.8). Comparative values for DFS were 38, 13, and 9 months respectively, again with no significant difference between PPLN+/CHALN- and PPLN+/CHALN+.

This contemporary PDAC cohort with prospectively collected CHALN shows that when this lymph node is positive there is no difference in survival when compared to patients who have metastatic spread to other lymph nodes. Additionally, it shows an actual 5-year survival of 50% in patients who underwent pancreatoduodenectomy for PDAC and have negative lymph nodes, and 14-20% in patients with positive lymph nodes. This improvement from previously reported rates is likely a reflection of current adjuvant and neoadjuvant therapies.

Introduction:
Over the last decade a neoadjuvant-first approach has garnered increasing popularity in the management of pancreatic ductal adenocarcionma (PDAC). Systemic therapy is associated with considerable chemotoxocity, and chemotherapy intolerance is associated with performance status and comordibidities. Over a third of patients diagnosed with PDAC are aged ≥ 75 years. The aim of this study was to assess the utlization of neoadjuvant therapy (NAT) and its impact on survival in this cohort.
Methods:
The National Cancer Database (NCDB) was used to identify patients diagnosed with PDAC between 2010 and 2017 who underwent pancreatectomy. Patients were staged using the American Joint Committee on Cancer staging system. Patients with stage IV disease or those with missing data on stage or NAT were excluded. Demographic and clinicopathological characteristics were assessed using Chi Squared Test. Factors associated with receipt of NAT were identified using logistic regression. The association between NAT and hazard of mortality was assessed using Cox proportional hazards model. Median overall survival (OS) was examined.
Results:
A total of 26,346 patients were included of whom 21% were ≥75 years of age. Younger patients were more likely to have a Charlson-Deyo comorbidity score of 0 (64% vs. 60%, p<0.001). No significant differences were observed between the two cohorts in terms of tumor size, margin status, lymphovascular invasion and grade (all p-values>0.05).
NAT was administered in 12% of the elderly patients as compared to 24% in the younger cohort (p<0.001). After controlling for sex, race, tumor size, grade, clinical stage, and Charlson-Deyo comorbiditiy score, elderly patients were less likely to receive NAT (OR:0.46, 95%CI: 0.40-0.53, p<0.001). Charlson-Deyo score was not associated with receipt of NAT in the elderly cohort (p=0.986). On multivariate analysis, NAT was associated with improved survival in both the elderly (HR:0.80, 95%CI: 0.70-0.92, p=0.002) and younger cohort (HR:0.78, 95%CI: 0.73-0.83, p<0.001). In the younger cohort, median OS was 30.1 months for patients that received NAT and 23.2 months for those that did not (p<0.001) . In the elderly, median OS was 24.9 months for patients that received NAT and 17.8 months for those that did not (p<0.001).


Conclusion:
NAT is associated with improved OS in patients with PDAC who are ≥75 years of age compared to those that did not receive NAT. However, its utilization in this cohort remains significantly lower than in their younger counterparts. Comorbidities are not associated with the receipt of NAT in the older cohort. Furture studies are required to identify factors driving these lower rates which when addressed, could potentially help improve management in these patients.

Introduction:
Laparoscopic pancreaticoduodenectomy (LPD) has been shown to be safe, feasible, and potentially advantageous over open pancreaticoduodenectomy (OPD). However, there are reports of higher complications, namely postoperative pancreatic fistula (POPF), with LPD. We hypothesized that OPD was associate with lower complications and the aim of this study was to compare LPD to OPD regarding postoperative outcomes (Quality) as well as the financial burden (Cost) in order to obtain an overview of the Value (Quality/Cost) of LPD compared to OPD.
Methods:
From January 2010 to December 2020, OPD (n=347) and LPD (n=242) were performed by a surgical team using the same selection criteria, operative technique, and recovery protocols at a single institution. Total pancreatectomy, additional complex multivisceral resections, and major vascular reconstructions were excluded to diminish selection bias. Operative data and 90 day outcomes were compared and a statistical analysis was performed. Additionally, Cost required for surgery and recovery (additional OR time, readmissions and total hospital days, ICU days, postoperative imaging studies, radiologic and GI interventions, and need for TPN and blood transfusions) gathered and tabulated.
Results:
In all, 261 patients undergoing OPD were compared to 183 patients undergoing LPD. LPD patients were significantly younger (64.3 vs. 67.4, p=0.009) with higher BMI (28.4 vs. 26.7, p=0.002) but had similar rates of comorbidities, ASA, and ECOG status. LPD was associated with similar operative blood loss, transfusion rates, and need for vascular resection but had significantly longer operative times (472 vs 271 min, p=0.0001).
Regarding Quality of the operation, LPD was associated with similar mortality but significantly higher major complications, pancreatic fistula, hemorrhage, delayed gastric emptying, intra-abdominal abscess, need for postop imaging, need for intensive care, readmission, and total hospital stay (Table 1).
Regarding Cost, the mean Cost for LPD was significantly higher than for OPD (Table 1).
Conclusions:
According to this review at our institution, performing LPD appears to significantly worsen the Quality as well as increase the Cost of PD, therefore detracting significant Value from patients requiring PD. Additional unmeasured Costs including quality of life for the patient, unnecessary burden on the interventional and endoscopic support systems, and opportunity costs lost by the surgeon and OR team by the increase resources necessary for LPD should be examined and added to the overall Cost of LPD in future analyses.

Introduction:
Failure-to-rescue remains central to reducing mortality following pancreatic resection. Postoperative CT imaging has been shown to play an important role in intercepting failure-to-rescue precursors, though the timing of such imaging remains relatively understudied. We sought to examine the utility of immediate, early, and delayed pre-discharge abdominal CTs in pancreatectomy patients.

Methods:
Patients who underwent pancreatic resection at our institution from 2017-2022 were reviewed retrospectively. Clinical data and outcomes were recorded until 90 days postoperatively. Management before and after CT was analyzed for treatment changes and correlated with imaging findings. Patients were subdivided by the postoperative day that the first CT scan was obtained: immediate (POD<3), early (POD 3-7), and delayed (POD>7). Fisher exact/Chi-square, Student’s t, and rank sum tests were used with pairwise comparisons for the early and delayed groups.

Results:
Of 370 patients, 110 (30%) had a CT during the initial surgical stay. Indications included suspicion of infection in 60 (55%), bleeding in 10 (9%), and other concerns, such as severe pain or GI symptoms, in 40 (36%). A change in treatment was observed in 59% following CT, with 15% undergoing invasive interventions and 27% treated medically. Of those who had CT imaging, 12% had scans within 3 days (immediate), 54% days 3-7 (early), and 34% after day 7 (delayed). The three groups were similar in baseline characteristics such as age, comorbidities, pathology, operative time, and number of scans (table 1). Regarding imaging indications, concerns for bleeding (15%) were more common in the immediate period while infectious suspicions predominated the early (55%) and delayed (65%) groups. The proportion of patients undergoing pancreaticoduodenectomy increased with POD group (p=0.026). Comparing the early to the delayed group (table 2), antibiotic use and initiation day (80% vs 70%, POD 5.05 vs 6.15), percutaneous drainage and placement day (12% vs 14%, POD 12.9 vs 14.6), and overall invasive interventions during surgical stay (20% vs 30%) were all similar (p=NS). Both 30 and 90-day readmission rates were equivalent as well at 18% vs 15% and 32% vs 30%, respectively (p=NS). Importantly, those scanned in the early period had significantly shorter length of stay (17.05 vs 22.82, p=0.0008) and fewer composite days hospitalized (20.10 vs 24.93, p=0.0109) relative to the delayed group. Further, surgical stay mortality rates were significantly lower in the early versus delayed group (0% vs 11%, p=0.0191).

Conclusion:
In our cohort, patients imaged early after pancreatectomy experienced shorter hospital stays and lower inpatient mortality relative to those scanned after the first postoperative week. When intra-abdominal complications are suspected following pancreatectomy, axial imaging should not be delayed.