Contemporary report of surgical outcomes after single‐stage total pancreatectomy: A 10‐year experience

Surgeons rarely perform elective total pancreatectomy (TP). Our study seeks to report surgical outcomes in a contemporary series of single‐stage (SS) TP patients.

6][7][8] Significant morbidity and mortality, lifelong debilitating pancreatic insufficiency, impaired patient quality of life (QoL), 7,9 and questionable oncologic benefit contribute to the general disfavor and reluctance to perform this complex operation electively.
1][12] Furthermore, advances in perioperative care and surgical techniques, as well as the evolution of high-volume specialized pancreatic centers have led to improved outcomes in pancreatic surgery. 7,13Concurrently, the advent of modestly effective oncologic therapies for pancreatic adenocarcinoma and the increasing frequency of detection of premalignant lesions and nonadenocarcinoma-variants provide sound rationale for future patients now considered potential candidates for SSTP with curative intent. 14e indications and decision-making process, whether made preoperatively or intraoperatively, to perform an elective SSTP is complex and often multifactorial, predicated in part upon disease biology, surgical-anatomic considerations to achieve R0 resection, and safety.In this contemporary report, we review our perioperative and oncologic outcomes in a series of 60 consecutively treated patients that underwent SSTP over a 10-year period at a high-volume NCIdesignated quaternary care center.Our overall objective is to contribute to a small but growing body of evidence supporting the feasibility of SSTP in appropriately selected patients that will facilitate more informed discussion about the benefits and risks of this operation.

| PATIENTS AND METHODS
This is a single-center, retrospective study approved by the Thomas Jefferson University Hospital (TJUH) Institutional Review Board (IRB).
Consecutive patients undergoing SSTP with en-bloc splenectomy between June 2013 and January 2023 were included and analyzed.
Clinical data were abstracted from our prospectively maintained pancreatic cancer (PC) database and institutional digital records obtained on EPIC.The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were followed during conductance of this study. 15A STROBE checklist was used in conjuction for data reporting.
Performance of elective single-stage total pancreatectomy was defined as either preoperatively planned or intraoperative decision to convert to a TP.Extent of resection was conducted according to the International Study Group of Pancreatic Surgery (ISGPS), 4 and our surgical technique is described below.Standard/classic and pylorussparing TP as defined by ISGPS includes either resection and/or preservation of the pylorus and antrum, respectively.Preoperatively, the Eastern Cooperative Oncology Group (ECOG) 16 and the American Society of Anesthesiologists-Physical Status (ASA-PS) 17 provide parameters that contribute to selection criteria to evaluate each patients' performance and physiologic status, thus assessing candidacy and suitability for pancreatectomy.Major postoperative morbidity reported herein details exclusively Clavien-Dindo grade ≥III complications. 18Specifically, delayed gastric emptying (DGE) and post pancreatectomy hemorrhage (PPH) are classified as per the ISGPS definitions 19,20 ; grade B and C complications were both considered clinically relevant.The duration of postoperative hospital stay was determined from the day of surgery defined as Day 0 to the day of discharge.Data pertaining to demographics, final pathology and patient survival were abstracted and reported from our PC database.
Patient, tumor, and treatment characteristics are described using numbers and percentages for categorical variables analyzed by patient frequency.Continuous variables are summarized using median and interquartile ranges (IQRs) while survival was analyzed using Kaplan-Meier actuarial methodology.All analyses were performed using Statistical Package for the Social Sciences (SPSS) Statistics Version 28.0 (IBM SPSS).A p < 0.05 was considered statistically significant.

| Surgical technique
Experienced pancreatic surgeons at our institution approach en bloc single-stage total pancreatectomy using several main steps. 21,22atomic considerations are referenced as illustrated in Figure 1A.The pancreas, spleen, proximal jejunum, duodenum, distal bile duct with or without pylorus-preservation are, in most cases, resected enbloc.(7) Lastly, Reconstruction: Restoration of gastrointestinal continuity entails end-to-side biliary-enteric and gastro (or duodeno)-enteric anastomoses.In the scenario of nonpyloruspreservation, a GJ reconstruction is performed (Figure 1B).Typically, both the hepaticojejunostomy and the GJ/DJ are performed in retrocolic fashion, through separate rents in the transverse mesocolon.

| Indications for surgery, decision making, and pathologic details
During this 10-year study period, we are reporting on 60 elective open single-stage total pancreatectomy operations, which occurred as the index procedure, inclusive.Thus, no patients had a prior partial pancreatectomy.SSTP accounted for 3% (60/1859) of all formal pancreas resections performed at our institution with an increasing trend apparent during the latter years of this study.Surgical decisionmaking varied as shown in Figure 2; twenty-nine patients (48%) were preoperatively scheduled for TP, and underwent SSTP.In thirty-one patients (52%), the decision to convert from a partial pancreatectomy to SSTP occurred after intended pancreaticoduodenectomy (PD) or distal pancreatectomy (DP) in 23 and eight patients, respectively.To summarize, in 11 patients (35%), the decision for conversion was due to tumor size, while the decision to extend the resection was due to positive intraoperative microscopic margin status in seven patients (23%), and to multifocal disease in five patients (16%).In a single patient (3%), the pancreatic duct appeared grossly abnormal on intraoperative pancreatic ductoscopy.In two patients (6%), the decision was made to convert to TP due to extensive vascular involvement.In three patients (10%) the presence of pancreatic necrosis and in two patients (6%) inflamed, soft, friable pancreata contributed to this decision.In those cases, the primary attending surgeon decided that TP would be safer and more beneficial when partial pancreatectomy was not possible for anatomic or technical reasons.
All operations were performed through an upper midline incision and each resection included splenectomy.The median size of the resected pancreatic tumors was 3.6 cm (IQR: 2.5).In 36 cases (60%) a classic TP was performed, with pylorus preservation carried out in the remaining 24 patients.In certain cases (20 patients, 33%), the extent of resection included adjacent vascular structures necessitating an array of vascular resections and reconstructive procedures to achieve a complete resection, thus facilitating an R0 resection with curative intent (Table 2).Patients with large centrally positioned tumors involving the neck or body of the pancreas with tumor-vessel interface >180°represented the most technically challenging of these procedures.In two patients (3%), concomitant resection of visceral structures included subtotal gastrectomy and small bowel beyond the first segment of the jejunum.Among this cohort of 60 patients that underwent a total pancreatectomy, 59 patients (98%) had an R0 resection with negative microscopic margins.A single PDAC patient had a microscopically positive posterior margin and, thus, by definition underwent an R1 resection.Overall, the pathologic characteristics for PDAC (n = 54) patients were consistent with more advanced T3 and/or T4 staged tumors in 65% (n = 35/54), with N1/N2 lymph nodes positive for metastasis in 41% (n = 22).Specimen lymph node yields revealed a median of 24 regional lymph nodes (IQR: 24) harvested from the PDAC group of patients.In patients that underwent unplanned conversion to SSTP, PD patients had notably larger tumors compared to DP patients (T3 60.9% vs. 37.5%, p = 0.024).However, no difference was noted in nodal staging (P = NS).For the entire cohort, the median duration of surgery was 516 min (IQR: 135) and the estimated blood loss (EBL) was 700 mL (IQR: 662).Operative and pathologic details are further summarized in Tables 2 and 3. Compared to planned TP, PD/DP cases that underwent unplanned conversion to SSTP, were characterized with shorter operating times (517.5 vs. 593.2min, p = 0.024) and decreased EBL (775.8 vs. 1634.5mL, p < 0.001).

| Outcomes
The distribution of postoperative complications are shown in Table 4.   Overall, among the 60 patients reported in this study, 1-, 3-, and 5-year composite survival estimates were 68%, 43%, and 16%, respectively, with a median overall survival of 24.4 months (±3.05).
Specifically for the patients with a diagnosis of PDAC (n = 54 patients), 1-, 3-, and 5-year survival was 68%, 42%, and 16%, respectively, with a median overall survival of 22.7 months (±3.3) (Figure 3).Is total pancreatectomy justified?The decision to perform elective TP in the setting of an invasive cancer continues to be a difficult one for surgeons.Herein, 60 consecutive patients underwent SSTP.Notably, elective SSTP represented only 3% of pancreatectomy procedures performed at our institution, of which the vast majority 90% were conducted for PDAC.In this report, all patients underwent TP as the index procedure with intraoperative conversion from partial pancreatectomy to TP occurring in over half (52%).
Central (neck/body) tumors found intraoperatively to be more advanced than previously believed and positive microscopic margins account for the most common reasons for conversion to TP.
Interestingly, multicentric PDAC, which is generally regarded as a rare occurrence, was encountered in five such patients (16%) that necessitated a TP.[25][26] Importantly, significant advancements improving postprocedural TP metabolic consequences allow for the further consideration and performance of this operation.1][12]   mortality rates now reported after complex pancreatic surgery. 7,138][29] Importantly, during this study's 10-year contemporary timeline, the therapeutic benefits of neoadjuvant treatments garnered greater interest with evidencebased reports demonstrating increasing rates of R0 resection.
Remarkably, in a recent landmark phase III PREOPANC trial, 30,31 neoadjuvant therapy conferred a distinct survival advantage for resectable and borderline resectable pancreatic cancer patients, although likely dependent on a multitude of factors, contributes to a rationale and justification to consider aggressive surgical resection such as a total pancreatectomy. 32 whom should we do it?In this contemporary report, the overall indication to convert a partial pancreatectomy to SSTP was to achieve an R0 resection.Microscopic margin status remains one of the most significant predictors of survival after surgery for pancreatic cancer. 5,33,34Fifty-three patients (98%) among our PDAC cohort underwent an R0 resection.Arguably, the oncologic benefit of performing a total pancreatectomy is still in question. 35e utility of TP as inferred in the recent literature appears fourfold, resting firstly on the necessity for complete extirpation of the pancreas in those patients in whom this is the only way to obtain a R0 resection. 5,6Second, we concur with prior authors, that tumors requiring conversion to a total pancreatectomy do not necessarily represent an underlying more biologically aggressive disease. 6ther, as reported herein, clinical scenarios such as persistently positive neck margins and/or suspicious diseased-appearing pancreatic parenchyma generally regarded as harbingers of bad disease biology otherwise suggests presence of residual diffuse or macroscopic disease, 36 requiring a total pancreatectomy to achieve an R0 resection 3,5 ; alternatively avoiding an R2 resection with minimal to no survival benefit. 37The fact that survival is superior in our patient cohort with generally large (55%, T3/T4), node-positive tumors (41%, N1/N2) compared to previously published survival data in PD patients with microscopic negative surgical margins, perhaps imparts a reflection of adequacy on this operation. 35Third, from a technical and oncologic standpoint, the consensus remains that clearance of standardized TP pathologic margins of resection is paramount.Importantly, in our hands, during the challenging retroperitoneal resection-, all visible tissue from the right lateral aspect of the superior mesenteric artery during uncinate-takedown necessitates a meticulous dissection; hence, the SMA is skeletonized during separation of the total pancreatectomy specimen from the visceral vessels.Ideally, no gross or microscopic residual disease should be left behind. 38As recently reported by Zheng and colleagues, one rationale for not performing a SSTP includes patients with a concurrent positive R1 (or R2) uncinate margin that obviates an R0 resection. 35Thus, we agree with previous investigators that do not currently advocate proceeding with a TP in this clinical scenario. 6Lastly, the impact of clearing resection margins is becoming increasingly more debatable as pancreatic cancer patients who are left with microscopic R1 disease fare better than patients without surgery with the ever-expanding armamentarium of more effective neoadjuvant and/or adjuvant therapies mitigating microscopic residual disease. 39Consequently, burgeoning data about perioperative benefits of systemic therapy in combination with surgery as illustrated in previous reports support a narrative for further discussions expanding the indications and increasing the role of SSTP in the future. 30,31,40,41 TP safe?Although historically reported to be a high-risk procedure, in our experience and the current literature, 30-and 90-day mortality rates in contemporary TP patients now | 1241 approximate 3% and 6%, respectively.The association between the surgical complexity of TP and short-term outcomes is well described, [42][43][44] however as recently shown, appears proceduralrelated as defined by a four-tiered classification system.Based upon Loos and colleagues criteria, 24  (n = 501, 20.9 months) 35 or distal pancreatectomy 25 (n = 101, 16 months).Indeed, this encouraging result may be simply due to the favorable characteristics inherent to a highly select cohort of patients but represents data that now seemingly contests the previous unmet theoretical advantages for which this operation was historically proposed 46 : that total pancreatectomy for adenocarcinoma in carefully select patients may offer a survival advantage over less-than-total resection.To this end, singlestage total pancreatectomy is now increasingly recommended in patients with extensive main duct intraductal papillary mucinous neoplasm (IPMN) or multifocal IPMN, with the evidence of IPMN being a pre-malignant condition. 47In this study, all four patients with IPMNs and two patients with a rare diffuse multicentric cystic neuroendocrine tumor of the pancreas remain alive at a median follow-up of 20.5 months.
A clear shortcoming of this report is that we did not study the short and long-term endocrine, metabolic consequences and quality of life (QoL) in these patients, as these topics are important and should be addressed.Consistent with our regional and interstate referral patterns, long-term follow-up remains a challenge and ultimately is underreported.While we are witnessing a significant improvement in perioperative management with comparable complication rates after PD and TP, recent studies paint a varied picture when it comes to the inevitable presence of long-term pancreatic insufficiency. 10,11,42,43,48is study has several additional limitations-this is a retrospective study and potentially subject to selection biases and errors of omission and classification.However, the study is based on a prospectively maintained database that includes a wellcharacterized cohort of pancreatectomy patients and is therefore less likely to contain selection bias or errors of omission.We also strived to verify the accuracy of the data by multiple data crosschecking to minimize any errors.Other limitations include the single-institution nature of this analysis and the high-volume character of our pancreatic resection practice.The latter two elements may limit the generalizability of the results.Since this study was undertaken in a high-volume specialized center, the nature of the patient cohort may bias certain patient characteristics unique to pancreatic tertiary referral centers.Among other limitations, we acknowledge the fact that our data lack recurrencefree survival analysis and does not focus on adjuvant therapies and their impact on the oncologic outcomes, as many of our patients return long-distances to their oncologist (thus subsequent treatments and follow-up cannot be fully assessed).In addition, we did not directly design a study comparing SSTP to partial pancreatectomy (PD and/or DP) yet relied on published data, which indeed introduces significant bias.We also did not include data regarding long-term changes to quality of life, secondary to the expected occurrence of diabetes.This paper is focused on surgical decisionmaking and its implications including short-term complications and overall survival.Further studies and discussion are certainly required to further establish SSTP as an acceptable surgical approach.

| CONCLUSION
Heretofore, the role of elective total pancreatectomy has faced staunch criticism.This report contributes to a small but growing body of evidence suggesting the justification, the feasibility and the safety of single-staged total pancreatectomy that will facilitate more informed discussion about the benefits and risks of this operation.
We contend that SSTP should be in the armamentarium of surgeons performing pancreatic resection.

( 1 )
Exposure: First, partial mobilization of the right colon and hepatic flexure to gain exposure and access to the duodenum.Performance of an extensive Kocher maneuver, retroperitoneal dissection of the pancreas and duodenum off the vena cava and the aorta allowing access to the superior mesenteric vein (SMV).Division of the gastrocolic ligament (with preservation of the right gastro-epiploic arcade) permits access to the body of the pancreas.(2) Dissection and mobilization: Tumor location permitting, the anterior aspect of the retro-pancreatic SMV once dissected, creates an avascular tunnel beyond the neck of the pancreas superiorly and continues until a circumferential plane around the neck is accomplished and can be elevated.A cholecystectomy is performed and the hepatoduodenal ligament is incised allowing hepatic arterial anatomic assessment and transection of the bile duct.Splenic ligamentous attachments are divided and the spleen and tail of the pancreas are mobilized medially, opening the retro-pancreatic space towards the pancreatic neck and portal confluence.(3) Vessel ligation: Test-clamping of the gastroduodenal artery (GDA) confirms normal pulsation in the common hepatic artery (CHA) and proper hepatic artery (PHA) before GDA ligation.The proximal splenic artery just beyond its origin from the celiac is suture ligated and divided allowing splenic decompression.The splenic vein is subsequently controlled where it enters the SMV.(4) Resection of antrum, duodenum & jejunum: Antroduodenal mobilization precedes duodenal transection 3 cm below the pylorus with pylorus preservation (in the case of classic TP, includes distal gastrectomy).The jejunum is divided 20 cm distal to the ligament of Trietz with vascular control of the resected duodenojejunal (DJ) mesentery.The DJ junction is then mobilized behind the mesenteric vessels to the patient's right side thereby permitting subsequent careful dissection and separation of the specimen.(5) Completion of uncinate takedown: This is performed meticulously, staying right on the right lateral aspect of the SMV and portal vein (PV) and working down onto the right lateral aspect of the superior mesenteric artery.The final mesopancreatic tissue is taken over clamps and silk-ties allowing complete separation of the specimen from the visceral vessels (6) Total pancreatectomy specimen: Sixty consecutive patients underwent single-stage total pancreatectomy at the Thomas Jefferson University Hospital from 2013 to 2023 and were included in this study.Fifty-four patients (90%) were diagnosed with pancreatic ductal adenocarcinoma (PDAC), four (7%) with extensive intraductal papillary mucinous neoplasm (IPMN), and two (3%) with extensive pancreatic neuroendocrine tumor (PNET).The median age was 68 years (IQR: 14), while 34 were males (57%) and 26 were females (43%).Thirty patients (50%) received preoperative chemotherapy and nine (15%) received preoperative radiation.High-quality contrast-enhanced multidetector computed tomography (CT) scan, magnetic resonance imaging (MRI), and endoscopic ultrasound represent common modalities employed to assess neoadjuvant treatment (NAT) response (if conducted) and anatomic assessment for resectability.Corroborative serologic tumor marker CA 19-9 were recorded in the setting of a bilirubin level <2 mg/dL.Comprehensive preoperative evaluation included perioperative risk assessment as designated by ECOG and ASA-PS parameters (for the entire patient cohort: ECOG scores ≤2; ASA criteria ≤3 in 58/60 patients), physiotherapist, nutritional, and endocrine consultations.

F
I G U R E 1 Anatomy.(A) Normal anatomy with multicentric disease before total pancreatectomy.(B) Anatomy following total pancreatectomy.(Illustrations by Nitzan Zohar, MD.).CHD, common hepatic duct.
resistant to initial treatment.Eleven patients (18%) required an interventional radiology procedure and four patients underwent surgical re-exploration (7% reoperation rate).Thirty and 90-day perioperative mortality was 1.67% (one patient) and 5% (three patients), respectively.The single patient who succumbed within 30 days after a pylorus-preserving total pancreatectomy (which included a CHA resection and reconstruction) was readmitted after POD 6 discharge with sepsis.Subsequent CT imaging revealed multiple abdominal and liver segment 5 & 6 collections that required radiologic-guided drain placement.Thereafter, the patient continued to experience a clinical decline and became refractory to surgical intensive care treatment leading to a family decision to withdraw further life-sustaining measures.Among the two 90-day mortalities, one patient developed an upper gastrointestinal bleed.A subsequent arteriogram identified a ruptured splenic artery stump aneurysm that was coiled with initial source control.Later hemodynamic instability led to next of kin's decision for comfort care measures only.The second patient's cause of death remains undetermined.
Notably, 15 patients (28%) remain alive from the PDAC group at the time of preparation of this manuscript.Survivorship includes two 5-year survivors.No procedural or disease-related deaths occurred in the nonadenocarcinoma study group, which includes noninvasive multicentric or diffuse intraductal papillary mucinous neoplasms (IPMN, four patients) and multifocal neuroendocrine tumors (NETs, two patients) during a median follow up study period of 20.5 months (IQR: 39).

F
I G U R E 2 (A) Decision making process for elective single-stage total pancreatectomy (SSTP).Distribution of preoperatively planned versus intraoperative decision to convert to SSTP (Left), distribution of partial pancreatectomy procedures (pancreaticoduodenectomy and distal pancreatectomy) before SSTP (percentage out of total converted operations, right).(B) Distribution of indications for intraoperative conversion from partial pancreatectomy to SSTP (n = 31, number of converted surgeries during study period.Percentage out of 31 converted SSTP).*Includes persistent positive resection margins.Total pancreatectomy for pancreatic cancer and other neoplastic entities remains controversial.In this context, over the recent years there has been a renewed interest in the performance of TPs, which resonates as the central theme and rationale for undertaking this study.We set out to address three principal questions: (1) Is total pancreatectomy justified?(2) In whom should we do it?(3) Is it safe?This current study is thus timely and warrants reporting as it represents a contemporary experience of SSTP patients treated at a single high-volume NCI-designated center.

T A B L E 3
Abbreviations: IQR, interquartile range; R0, margin negative resection; R1, removal of all macroscopic disease, but microscopic margins are positive for tumor.a Report for PDAC patients only, percentage out of the PDAC patient cohort.b Percentage out of the non-PDAC patients cohort (IPMN, NET).

T A B L E 4
Abbreviation: IQR, interquartile range.
36 of our patients (60%) underwent a Type I SSTP (TP without vascular or adjacent organ resection); 17 patients (28%), Type 2 (TP with portal vein and/or superior mesenteric vein resection); two patients (3%), Type 3 (TP with multivisceral resection); and five patients (8%), Type 4 (TP with arterial resection).Similarly, as noted by Loos, and observed in our patient cohort, a gradual increase in morbidity was experienced with the increasing complexity of SSTP type performed.Notwithstanding, extent of resection should be individualized, oncologically and when technically feasible, undertaken with surgical expertise inherent to a high-volume center.Importantly, data from this contemporary report demonstrates that while overall patient morbidity remains high (60%), the majority (56%) of complications were minor complications (Clavien type II and I).More serious types-III, IV, and V in our experience occurred in 16 patients (27%) and were seemingly related to the increasing complexity of the operation performed.These results remain high enough to be precautionary when performing more complex elective SSTP-type procedures.Notably, our perioperative outcome metrics such as relatively short postoperative hospital length of stay (median, 6 days) and 30-day readmission rate (27%) remains consistent with our established enhanced pancreatectomy recovery clinical pathways45 and routine early outpatient follow-up, respectively.Overall, these data and the growing body of evidence suggest that the increasing numbers of TPs now being performed at high-volume centers can be undertaken safely.Not surprisingly, the prognosis in patients undergoing SSTP varied based upon the pathology.The patients studied herein undergoing SSTP for ductal adenocarcinoma had a median overall survival (22.7 months) that outperformed contemporaneous cohorts of PDAC patients undergoing pancreaticoduodenectomy

F I G U R E 3
Kaplan-Meier composite overall survival function for pancreatic ductal adenocarcinoma (PDAC) patients only (n = 54) undergoing total pancreatectomy during study period (median survival for PDAC 22.7 months, ±3.3).

Table 1
provides further details regarding preoperative patient demographics, comorbidities, NATs, tumor, and treatment-related variables.