A Review of the Indications, Outcomes, and Postoperative Management After Total and Completion Pancreatectomy for Pancreatic Cancer: More Is Not Necessarily Better

INTRODUCTION

Early reports in the literature suggest that Dr Theodor Billroth completed the first successful total pancreatectomy (TP) in 1884.¹ In 1943, Dr Eugene Rockey documented performing a TP for a patient with pancreatic adenocarcinoma; unfortunately, this patient died 15 days postoperatively.² Given insufficient documentation of Billroth’s case and apprehension regarding the reported favorable outcome in the pre-insulin era, many give Rockey credit for the first TP.^1,3 Only 1 year later in 1944, Dr James Priestley documented the first successful TP for the treatment of hyperinsulinism secondary to an islet cell adenoma.¹ These early recounts established the technical feasibility of performing a TP, but this procedure gained little traction as subsequent studies failed to demonstrate improved oncologic outcomes and raised concerns regarding poor quality of life (QoL), attributed to both brittle diabetes and malabsorptive symptoms.⁴ With improvements in surgical technique and advances in pancreatic endocrine and exocrine replacement therapy, the indications for and threshold to perform total or completion pancreatectomy in the modern surgical era are ever evolving. This review will evaluate the indications for total and completion pancreatectomy for pancreatic cancer, as well as evaluate postoperative management and longterm oncologic outcomes after TP.

INDICATIONS FOR TOTAL OR COMPLETION PANCREATECTOMY

Benign Disease, Intraductal Papillary Mucinous Neoplasms, and Familial Pancreatic Cancer

Briefly, for benign disease, TP with islet cell auto-transplantation is indicated for select patients with chronic pancreatitis who have preserved endocrine function but are experiencing pain refractory to optimal medical and endoscopic therapies.^3,5 For neoplastic lesions, the increasingly common indication for TP is for the management of intraductal papillary mucinous neoplasms (IPMNs). Knowing that the risk of invasive carcinoma (IC) and high-grade dysplasia (HGD) approaches 60% for main-duct IPMNs (MD-IPMNs), surgical resection is strongly recommended for all surgically fit patients with “high-risk stigmata” based on the updated Fukuoka guidelines. The goal of surgical resection for MD-IPMNs is to obtain negative margins (ie, no cancer or HGD, thus, occasionally, TP is indicated in this clinical scenario).⁶ For those with familial pancreatic cancer (defined as patients with at least 2 first-degree relatives with pancreatic ductal adenocarcinoma [PDAC]), familial cancer syndromes (such as Peutz–Jeghers syndrome, familial adenomatous polyposis, hereditary non-polyposis colorectal carcinoma, familial breast–ovarian cancer, and familial atypical multiple mole melanoma), or hereditary pancreatitis, there is evidence in the literature to support prophylactic TP for select patients with high-risk lesions.^3,7 Multiple studies have demonstrated that surveillance allows for an early detection of IPMNs, pancreatic intraepithelial neoplasia (PanIN) or PDAC in patients with high-risk family histories, thus presenting a scenario where TP may be indicated.^7–11

Pancreatic Ductal Adenocarcinoma, Neuroendocrine Tumors, and Pancreatic Metastases

For malignant disease, TP may be indicated in properly selected patients with PDAC, pancreatic neuroendocrine tumors, or metastatic disease from primary tumors such as renal cell carcinoma or melanoma.^4,5,12 The indication to perform TP over a pancreaticoduodenectomy (PD) or distal pancreatectomy (DP) may be multifactorial—such as high-risk pancreatic anatomy (small pancreatic duct or soft gland texture), multifocal disease, or persistent positive margins on frozen section (FS) analysis.^13–15 Other common indications for a completion pancreatectomy would be as a rescue procedure for hemorrhage, pancreaticojejunostomy leak causing sepsis, or for recurrent disease after index pancreatectomy.¹⁵

Over the last few decades, medical and technologic advances have improved postoperative endocrine management such that there has been a renewed interest in performing total or completion pancreatectomy for the management of PDAC. In 2016, Johnston and colleagues queried the National Cancer Database and identified 5726 patients who underwent TP for PDAC between 1998 and 2011. These authors reported a median overall survival (OS) of 15 months with 1, 3, and 5 year survival rates of 59%, 21%, and 12%, respectively, but concluded that “these mortality rates are lower than those reported (after TP) in previous decades, but high enough to discourage elective total pancreatectomy if a partial resection would be oncologically and technically feasible.”¹⁵ Multiple systematic reviews and pooled analyses have demonstrated a potential survival benefit for patients who undergo completion pancreatectomy for recurrent disease after index resection for PDAC.^16–19 Despite reporting a 5 year survival rate of 40.6% after second pancreatectomy in their pooled-analysis, Zhou and colleagues comment that the improved outcomes are likely related to patient selection and note that the proportion of patients with unfavorable prognostic factors (such as nodal disease, tumor size, poor differentiation) were lower for those undergoing completion versus index pancreatectomy.¹⁶

PANCREATIC DUCTAL ADENOCARCINOMA

Pancreatic Ductal Adenocarcinoma: Pancreaticoduodenectomy Versus Total Pancreatectomy

Given that there has never been a randomized controlled trial comparing PD versus TP, controversy remains in the literature regarding the oncologic outcomes of these 2 procedures. Many retrospective studies have indicated equivalent survival after TP and PD when performed for PDAC (Table 1).^20–25 Two of these studies initially demonstrated that OS was improved for patients who underwent PD, but after matched-pairs analysis (matching for variables such as age, gender, tumor size, adjuvant therapy, and resection margin status), this difference was no longer statistically significant.^22,23 Zhu and colleagues reported the largest and most recently published study comparing cancer-specific outcomes for TP versus PD.²⁰ These authors used the Surveillance, Epidemiology, and End Results (SEER) database to identify all patients with PDAC of the head of the pancreas who underwent PD or TP over a 28 year time period. After propensity score matching, they demonstrated that there was no difference in 30 or 90 day mortality or cancer-specific survival between the TP and PD groups. Interestingly, on univariate analysis, tumor size, tumor stage, nodal stage, and systemic therapy were identified as independent predictors of survival. After adjusting for these variables in multivariable analysis, TP and PD again had comparable survival (hazard ratio [HR] 1.1 [95% CI 1.0–1.2]; P = .06).²⁰

Table 1.

Overview of retrospective studies comparing outcomes of total pancreatectomy and pancreaticoduodenectomy for pancreatic ductal adenocarcinoma

Author, Publication year Institution Country	Study Period	Database	Number of Patients Undergoing TP	Number of Patients Undergoing PD	Matching	Outcome	P-value
Zhu et al,20 2023 The First Affiliated Hospital of Nanchang University China	1988–2016	SEER	740	1479	Propensity score matching	Cancer-specific survival	P = .51b
Passeri et al,21 2019 Carolinas Medical Center USA	1998–2006	NCDB	807	5058	None	Median survival	P = .32a
Xiong et al,22 2017 West China Hospital China	2009–2015	Institutional database	50	325	Matched pair analysis	OS	P = .49b
Satoi et al,23 2016 Kansai Medical University Japan	2001–2011	7 Institutional databases	45	885	Matched pair analysis	OS	P = .215b
Reddy et al,25 2009 Johns Hopkins University School of Medicine USA	1970–2007	Institutional database	100	1286	None	OS	P > .05
Nathan et al,24 2008 Johns Hopkins University School of Medicine USA	1998–2004	SEER	292	2988	None	OS	P = .79

Abbreviations: SEER, surveillance, epidemiology, and end results; NCDB, United States National Cancer Database.

^aReport P-value corresponds to the comparison of median survival of TP versus PD.

^bReported P-value corresponds to comparison OS of TP versus PD after matching.

From Ref.^20–25

Positive Intraoperative Neck Margin: Indication for Total Pancreatectomy?

There are multiple pancreatic margins assessed in the setting of PD: neck, radial, uncinate/superior mesenteric artery, and posterior/retroperitoneal. Given the pancreatic neck margin is the most amenable to surgical remediation, many advocate for reresection of microscopically positive intraoperative neck margins during PD for PDAC.^26,27 Similar to the previous discussion on oncologic outcomes of TP versus PD, data regarding the oncologic utility of re-resection to achieve a tumor-free neck margin during PD are conflicting. The Central Pancreatic Consortium (CPC) reported the largest retrospective, multicenter study on this topic, exploring 1399 patients undergoing PD for PDAC at 8 high-volume academic centers in the United States. On intraoperative frozen section(FS), 10.9% of patients had an R1 resection at the neck, while 3.6% had a false-negative FS-R0 neck margin, as the final margin analysis revealed microscopic disease. A permanent section (PS)-R0 neck margin was achieved in 85.5%, whereas 9.3% remained PS-R1, and 5.1% were converted from FS-R1 to PS-R0 by additional surgical resection. Median OS was significantly longer for PS-R0-neck patients when compared to both PS-R1-neck (P < .001) and FS-R1 to PS-R0 (P < .001) patients. On univariate Cox regression analysis, histologic grade, tumor size, SMA margin positivity, lymph node positivity, lymphovascular invasion, perineural invasion (PNI), T-stage, and neck margin status were all associated with worse OS. Importantly, when adjusting for these factors on multivariable analysis, conversion of a positive FS margin to a PS-R0-neck margin remained associated with significantly worse OS when compared to PS-R0 patients (HR 1.55; P = .009), suggesting that a more aggressive surgical resection cannot “make up” for worse tumor biology.²⁸

In contrast to the earlier CPC study,²⁸ other authors have reported that re-resection of a positive pancreatic neck margin on FS analysis may be associated with an improved survival.^29,30 For example, the Dresden group examined 483 patients who underwent surgery (PD, DP) for an exocrine pancreas malignancy between 1993 and 2014 using an institutional database. These authors reported a median OS for patients with an initial R0 resection (29 months) similar to those who underwent re-resection to convert an FS-R1 to PS-R0 (36 months; P = .849).²⁹ Although, cross-study comparison should be taken with caution; the CPC study only included patients who underwent PD for PDAC, while the Dresden group’s analysis included patients who underwent PD and DP for varying tumor histology, of which only 69.4% reresections were done for PDAC.^26,28,29 Another study of 986 patients who underwent pancreatectomy (PD, TP) for PDAC between 1998 and 2012 at Massachusetts General Hospital (MGH) and the University of Verona demonstrated improved survival with re-resection of a positive intraoperative FS neck margin. These authors reported significantly increased OS for patients with PS-R0 resection (28 months; P = .01) and FS-R1 to PS-R0 resection (24 months; P = .02) when compared to patients with PS-R1 (19 months). Of the patients with a positive intraoperative FS, 28.3% underwent TP to achieve PS-R0 margins while only 1.6% in the PS-R0 and no patients with PS-R1 underwent TP.³⁰ As noted in a recent review on this topic by Datta and colleagues,²⁷ when compared to the CPC patient cohort,²⁸ the MGH-Verona cohort included smaller tumors, a lower proportion of patients with T3-T4 disease, and a higher proportion of patients who received multimodal neoadjuvant/adjuvant therapies, suggesting that selection bias might account for the reported survival advantage.^27,30

Patterns of Recurrences after Pancreatectomy for Pancreatic Ductal Adenocarcinoma

Following surgical resection of PDAC, the majority of patients will recur within 2 years, suggesting that many patients have micro-metastatic disease at the time of resection.^31,32 The pattern of disease recurrence includes locoregional disease, generally defined as recurrent disease in the pancreas, surgical bed, or within locoregional lymph nodes, versus distant disease, such as metastases to the liver and/or lung.³¹ A large retrospective study by Groot and colleagues on 531 patients who underwent pancreatectomy (PD, TP) at Johns Hopkins between 2000 and 2010 reported recurrence data that demonstrated that at the time of first recurrence, 57.8% of patients had isolated distant metastatic disease, 23.7% had isolated local disease, and 18.5% had both local and distant disease. Interestingly, the location of recurrence differed by the time of diagnosis of recurrence. For example, when recurrence was diagnosed within 6 months of resection, liver-only, lung-only, and local-only recurrence accounted for 47.6%, 7.9%, and 10.3% of all recurrences, respectively. In contrast, when diagnosed after 2 years, liver-only, lung-only, and local-only recurrence accounted for 12.1%, 27.3%, and 39.4% of all recurrences, respectively.³³ These data suggest that early recurrences are more likely to occur in the liver, while late recurrences are more likely to occur locally or in the lung. Using multivariable regression analysis, these authors demonstrated that recurrence locations can be predicted by different clinicopathological factors. For example, the presence of a positive surgical margin was predictive of local-only recurrence, while patients who received adjuvant chemoradiotherapy were significantly less likely to have local-only recurrence. Tumor size, poor-to-moderate tumor differentiation, and a lymph node ratio of greater than 0.2 (defined as the number of positive lymph nodes divided by the total number of lymph nodes resected) were all significant predictors of liver-only recurrence. Perivascular invasion and a lymph node ratio of greater than 0.2 were significant predictors of lung-only recurrence.³³

The European Study Group for Pancreatic Cancer 4 trial (ESPAC-4) was an international, multicenter phase 3 randomized controlled trial that included 730 patients who underwent pancreatectomy (PD, DP, and TP) for PDAC followed by adjuvant gemcitabine versus combination gemcitabine plus capecitabine.³⁴ Dissimilar to the retrospective study by Groot and colleagues,³³ a secondary analysis of ESPAC-4 by Jones and colleagues showed more patients (49.7%) had local-only recurrence compared to distant-only (40.3%) and simultaneous local/recurrent disease (10%).³¹ Similar to the Hopkins study,³³ Jones and colleagues found that liver metastases occurred earlier than lung metastases (9.6 vs 15.3 months; P < .001). Interestingly, patients with lungonly recurrent disease also had a significantly longer survival from time of recurrence when compared to those with liver-only recurrent disease (HR 0.6; P = .01). Given that a combined total of only 78 patients included in the Hopkins³³ and ESPAC-4 secondary analysis³¹ underwent TP, we examined the recurrence data in the previously discussed matched-pairs study by Satoi and colleagues.²³ These authors reported a significantly increased rate of liver-only recurrence in the TP versus PD cohorts (52% vs 34%; P = .036).²³ These data suggest that patients who require TP to ensure an appropriate oncologic resection likely have differing and more aggressive tumor biology than those undergoing PD. Altogether, these studies suggest biologic heterogeneity among primary PDAC tumors that primes divergent patterns of disease recurrence and survival. One may infer that the inherent biologic characteristics of a PDAC that mandate a TP to achieve an oncologically appropriate resection also likely leads to early distant disease recurrence, thus raising into question the utility of the more aggressive resection.

INTRADUCTAL PAPILLARY MUCINOUS NEOPLASM/CARCINOMA

Classification of Intraductal Papillary Mucinous Neoplasms

In 2006, the International Association of Pancreatology (IAP) published the initial guidelines for the management of IPMNs and mucinous cystic neoplasms which have subsequently been updated in 2012, 2018, and 2023.^6,35–37 IPMNs are initially classified based on imaging characteristics and then pathology, if resection is pursued. Based on the consensus statements, an MD-IPMN is characterized by segmental or diffuse dilation of the main pancreatic duct (MPD) by more than 5 mm, a branch duct IPMN (BD-IPMN) is defined as a pancreatic cyst that is greater than 5 mm in diameter and communicates with the MPD, while mixed-type IPMN meet criteria for both MD-IPMN and BD-IPMN (Fig. 1A). For the diagnosis of an MD-IPMN, other causes of MPD obstruction need to be ruled out. For BD-IPMN diagnosis, pancreatic pseudocysts secondary to pancreatitis or trauma should be ruled out.⁶ Based on the updated World Health Organization classification, IPMNs are graded pathologically as low-grade dysplasia (LGD), HGD, or IC.^6,38 Given that the rates of HGD and IC differ between MD-IPMN and BD-IPMN, this classification schema has clinical significance when considering which patients will benefit from surgical resection.³⁷ IPMNs are further distinguished based on 3 morphologic subtypes: gastric, intestinal, and pancreaticobiliary type. These subtypes are associated with prognosis, pathologic types of IC (colloid vs tubular adenocarcinoma), and risk of the development of recurrence in the remnant pancreas after resection.^6,39–41

Fig. 1.

(A–C) IPMN-related carcinoma sequence. Created with BioRender.com. (From Ref.^6,35–38,42)

Indication for Resection of Intraductal Papillary Mucinous Neoplasms

The updated 2023 Kyoto guidelines present “operative principles” and a revised decision tree based on “high-risk stigmata” and “worrisome features” for both MD-IPMN and BD-IPMN (Fig. 1B). Ohtsuka and colleagues state that when considering resection for an IPMN, patients must receive thorough preoperative counseling given the operative plan may change significantly based on intraoperative findings. Organ-preserving (or non-anatomic) pancreatectomy can be considered if a noninvasive diagnosis is favored, while radical pancreatectomy with lymphadenectomy should be pursued if there is concern for or cytologic evidence of IC. The goal of resection is to achieve complete removal of the tumor with a negative margin, thus, evaluation of intraoperative FS is recommended to determine the extent of resection. The authors state additional resection should be pursued if the intraoperative margin is positive for HGD or IC but recommend against additional resection if the margin has evidence of LGD. On the topic of TP for IPMN, the authors specifically state “due to its endocrine and exocrine metabolic consequences, prophylactic total pancreatectomy is not recommended, however, possibility of total pancreatectomy should be informed to all patients (during preoperative counseling).”⁶

IPMNs have been described as a “dual” precursor of pancreatic carcinogenesis.⁴² First, an IPMN can progress from LGD to HGD and then become an IC, such as PDAC, colloid carcinoma, or tubular ductal carcinoma. Second, IPMNs are associated with an increased risk of PanIN-derived PDAC, which can occur as a concomitant (synchronous) tumor or a metachronous tumor, after resection of an IPMN (Fig. 1C).^42–45 Several retrospective studies have found that outcomes for patients with IPMN-associated carcinoma are superior than those with PDAC not associated with an IPMN, suggesting differing molecular pathways and favorable biology for IPMN-associated PDAC.^44,46 For example, in the recent study by Ziogas and colleagues,⁴⁶ the authors utilized the National Cancer Database to demonstrate that OS among patients with PDAC was significantly shorter than those with invasive IPMNs (8.3 vs 33.4 months, respectively; P < .001).⁴⁶ Given that patients with known IPMN remain under radiographic surveillance, it is possible that when they undergo resection for an invasive cancer, it may have been discovered at an earlier stage, thus contributing to improved survival.

Intraductal Papillary Mucinous Neoplasm Surveillance and Recurrence

Given the field defect and persistent risk of pancreatic malignancy associated with IPMNs, the IAP consensus guidelines recommend continued surveillance following resection of IPMNs. The updated Kyoto Guideline surveillance schedule is differentiated based on the pathology of the resected IPMN. If surgical pathology reveals a noninvasive lesion, surveillance should include a combination of biannual or annual physical examination, imaging, and blood tests based on institutional guidelines until the patient is no longer fit for surgical resection. However, if the resected lesion has evidence of invasion, Ohtsuka and colleagues recommend that the patient should undergo surveillance similar to PDAC protocols.⁶ Many authors have attempted to retrospectively identify risk factors associated with survival and recurrence. A retrospective study of 213 patients who underwent surgical resection for invasive IPMNs at Johns Hopkins between 1995 and 2018, reported that 92 patients (43.2%) experienced disease recurrence. Interestingly, the most common pattern of recurrence was systemic (55.4%), while 34.8% had local-only and 9.8% had both local and systemic recurrence. Additionally, patients with recurrent disease had significantly worse median OS when compared to those who did not have disease recurrence (34.3 vs 81.3 months, P < .001). Cox regression analysis identified lymph node positivity, elevated CA 19–9, poor tumor grade or differentiation, and PNI as predictors of a decreased RFS.⁴⁷ In another retrospective, multi-institutional study from the Japan Pancreas Society, of 1074 patients who underwent surgical resection for an IPMN between 1996 and 2014, 155 patients (14.4%) had postoperative recurrence, of which 70 (6.5%) were high-risk lesions (defined as PDAC, recurrent IPMN, or both) within the remnant pancreas and 88 (8.2%) were extrapancreatic recurrences. Of the 70 patients who had recurrence within the remnant pancreas, 36 (51.4%) underwent re-resection of the high-risk lesion. OS was significantly improved for patients who underwent re-resection of a metachronous tumor as compared to those who did not undergo re-resection (P = .04). Unfortunately, the proportion of patients who underwent completion TP for recurrent disease was not reported. Several factors were identified on multivariable analysis to be significantly associated with increased recurrence of a high-risk lesion in the remnant pancreas: preoperative symptoms, IPMN located in the body/tail, MPD greater than 10 mm, and HGD/IC.⁴⁸ A prior study from Johns Hopkins evaluated patients who underwent an index partial pancreatectomy for a noninvasive IPMN. Of the 130 included patients, 22 (17%) developed evidence of a new IPMN on postoperative surveillance imaging. Eleven of the 22 patients who had IPMN recurrence underwent completion pancreatectomy: 8 completion DP and 3 completion PD, and no mortalities were reported to be associated with completion pancreatectomy, highlighting the feasibility of this approach.⁴⁹

Considerations for Total or Completion Pancreatectomy for Intraductal Papillary Mucinous Neoplasm

Altogether, the authors of this review recommend that consideration of total or completion pancreatectomy for an IPMN be reserved for young and/or healthy patients with a longlife expectancy, evidence of persistent intraoperative positive margins (HGD or IC), or recurrent disease that warrants re-resection. In addition to the preoperative counseling recommended by the 2023 Kyoto guidelines, we strongly advise that preoperative counseling prior to total or completion pancreatectomy for invasive disease also include conversations regarding the real potential for systemic recurrence regardless of the extent of resection.

MANAGEMENT AND QUALITY OF LIFE FOLLOWING TOTAL OR COMPLETION PANCREATECTOMY

Endocrine Insufficiency

The American Diabetes Association classifies diabetes secondary to structural or functional pancreatic dysfunction as type 3c diabetes, a disease state that has colloquially been termed “brittle diabetes.”^50–52 This unique form of diabetes, characterized by complete insulin deficiency and the absence of glucagon, manifests as frequent, severe, and unpredictable fluctuations in blood glucose despite consistency in insulin regimens and diet.^51–53 The morbidity associated with these resections have prompted many studies on postoperative QoL following TP or completion pancreatectomy.^54–56 PANORAMA was a retrospective, cohort study completed by the Dutch Pancreatic Cancer Group that included all patients who underwent TP between 2006 and 2016 at 17 centers nationwide. These authors found that after completion pancreatectomy or TP, patients reported significantly lower global (P = .03) and daily (P < .01) QoL when compared with the general Dutch population, and mean QoL scores did not significantly change as time increased after pancreatectomy (<3, 3–5, or >5 years). Interestingly, patients reported overall satisfaction with their diabetes treatment regimen and had similar rates of diabetes-related distress as patients with type 1 diabetes.⁵⁴

Endocrine management of the a-pancreatic state requires an algorithmic approach and collaborative effort of a multidisciplinary care team that includes the patient, their caregivers, as well as the surgeon, endocrinologist, nurse educators, and dieticians. This treatment algorithm begins in the preoperative setting with education from diabetes nurse educators and dieticians, and, ideally, the patient would establish care with an endocrinologist who specializes in the management of type 3c diabetes.^52,57 Immediate postoperative insulin management generally begins with a continuous insulin infusion and dextrose replacement with frequent blood glucose monitoring, per institutional guidelines. Once stabilized, patients are transitioned to multiple daily insulin injections, insulin pumps, or continuous subcutaneous insulin infusion pumps and as needed glucagon.^52,58 A recent randomized controlled trial demonstrated improved glucose control with a bi-hormonal “artificial pancreas” system that simultaneously infuses both insulin and glucagon, highlighting the strides being made in the management of type 3c diabetes.⁵⁹ Despite some small studies demonstrating the feasibility of islet cell auto-transplantation in the setting of malignancy,^60–66 many institutions still consider cancer a relative or absolute contraindication to islet cell transplantation due to the potential seeding of premalignant or metastatic disease into the liver.⁶⁷

Exocrine Insufficiency and Other Complications

Due to exocrine pancreatic insufficiency, patients most commonly experience diarrhea, as well as steatorrhea.^51,53 This is generally managed with pancreatic enzyme replacement and nutritional supplementation, but despite efforts, many patients still experience malabsorption and weight loss, although some authors have demon-strated that weight and nutritional status stabilize with time.^51,53,68,69 Long-term metabolic sequelae of total and completion pancreatectomies include osteoporosis and hepatic steatosis.^{51–53,69,70} Another infrequent and severe complication after TP is gastric venous congestion (GVC) necessitating partial/total gastrectomy or venous reconstruction.^71–73 This complication can be identified in the operating room or present postoperatively as gastric perforation or hemorrhage.^71,73 GVC occurs due to compromised venous outflow from ligation of the right-sided gastric venous drainage in addition to ligation of the coronary vein. This results in dilation of the gastric veins, increased intravascular pressure, edema, discoloring of the stomach, and eventual hemorrhage and/or perforation.⁷¹ In a retrospective study by Loos and colleagues, among 585 patients who underwent TP or completion pancreatectomy at the University Hospital Heidelberg in Germany between 2015 and 2019, 163 patients (27.9%) had evidence of GVC with 154 (26.3%) identified intraoperatively requiring partial gastrectomy.⁷¹ The authors reported that 29 patients required re-exploration for GVC, of which 20 had already undergone partial gastrectomy during their index operation. Furthermore, they found that 90 day mortality was significantly higher for patients with GVC than those without (7.4% vs 2.8%, P = .014).⁷¹

SUMMARY

In the late nineteenth century, Dr William Halsted bolstered support for the theory that cancer spreads locally via lymphatics prior to becoming metastatic disease.^74–76 With the use of a radical mastectomy, Halsted achieved impressive locoregional control with a cited recurrence rate as low as 4% to 6%, but this came at the expense of a highly morbid operation and no difference in OS.^74,77 The paradigm shift from the radical mastectomy to the precision medicine model of breast cancer therapy practiced today took over a century to establish. With the advent of targeted systemic therapies and modern-day screening practices, surgical resection has become one part of a multidisciplinary care model for breast cancer managment.^74,77 Similarly, given the inherent aggressive biology of PDAC and advances in personalized targeted therapy for pancreatic cancer, one has to pursue TP with caution, as removal of more pancreas may not necessarily be better. Additionally, similar to the lymphedema experienced by many patients following radical mastectomy, GVC remains a frequent and dangerous complication after TP.^71,78

In conclusion, TP or completion pancreatectomy is feasible and can be considered in properly selected cases. These authors would reserve consideration for young and/or physically fit patients with an extended life expectancy who have the financial and social support necessary to endure the inevitable postoperative endocrine and exocrine pancreatic morbidities. Overall, the decision to perform a total or completion pancreatectomy should be a combined decision made between the patient, their support system, and a multidisciplinary care team. However, one must always be reminded that the biology of the tumor drives survival. As demonstrated many years ago by Dr William Halsted, increasingly radical tumor resections come at the risk of additional morbidity and mortality and do not always overcome the aggressive tumor biology associated with the index cancer.

KEY POINTS.

• Presently, there are no randomized controlled trials comparing outcomes for pancreaticoduodenectomy versus total pancreatectomy or completion for pancreatic cancer.

• Although studies propose that aggressive pancreatic resections may provide survival benefit in a subset of patients, current literature also proposes that biologic heterogeneity exists among primary pancreatic ductal adenocarcinoma tumors that prime divergent patterns of disease recurrence and survival.

• While studies suggest that a subset of patients with recurrent intraductal papillary mucinous neoplasm (IPMN) may experience survival benefit after completion pancreatectomy, it is also suggested that a majority of patients with resected IPMN-associated invasive carcinomas recur systemically rather than locally.

• More research is needed to identify the ideal population of patients with pancreatic cancers who may benefit from total or completion pancreatectomy.

CLINICS CARE POINTS.

• TP or completion pancreatectomy is feasible for pancreatic cancer and can be considered in properly selected cases.

• Presently, there are no randomized controlled trials comparing outcomes for PD versus TP or completion pancreatectomy for pancreatic cancer.

• Many retrospective studies have indicated equivalent survival after TP and PD when performed for PDAC.

• In patients with recurrent disease after index resection for PDAC, multiple systematic reviews and pooled analyses demonstrate improved survival in those who undergo completion pancreatectomy.

• Although studies propose that aggressive pancreatic resection may provide survival benefit in a subset of patients, current literature also proposes that biologic heterogeneity exists among primary PDAC tumors that prime divergent patterns of disease recurrence and survival.

• While retrospective studies suggest that a subset of patients with recurrent IPMN experience survival benefit after completion pancreatectomy, it is also suggested that a majority of patients with resected IPMN-associated IC recur systemically rather than locally.

• More research is needed to identify the ideal population of patients with pancreatic cancer who may benefit from TP or completion pancreatectomy, but given the significant postoperative morbidity, proceeding with such an aggressive resection should be a combined decision among the patient, their support system, and a multidisciplinary care team.