Comprehensive comparisons of total pancreatectomy versus pancreaticoduodenectomy for pancreatic ductal adenocarcinoma: a double-center, retrospective study.

Pancreatic cancer remains a highly devastating malignancy with limited therapeutic options. According to the latest statistics of the American Cancer Society, pancreatic cancer is estimated the fourth and third leading cause of cancer-related death in the United States among males and females, respectively, with the poorest 5-year survival of merely 13% ( 1 ). In China, the health burden of pancreatic cancer is also not optimistic, with annual mortality nearly equals to new incidence ( 2 ). Among all pancreatic malignancies, pancreatic ductal adenocarcinoma (PDAC) accounts for over 90% and has almost become synonymous with pancreatic cancer ( 3 ). Surgical resection remains the only potential curative method. Unfortunately, only fewer than 20% of PDAC patients are eligible for surgical intervention since it is usually detected at late stages ( 4 ). As we all know, there are three main surgical approaches for PDAC generally based on the lesion location, pancreaticoduodenectomy (PD), distal pancreatectomy (DP), and total pancreatectomy (TP). Among these three surgical methods, PD and DP have been commonly applied in clinical practice with confirmed short- and long-term outcomes ( 5 , 6 ), while the role of TP in the treatment of PDAC has been controversial. TP was first performed by Rockey in 1943 for PDAC, but the patient died 15 days after operation due to severe bile leakage ( 7 ). In the early time, TP was carried out to prevent potential postoperative pancreatic fistula (POPF) and POPF-related complications, and also to pursue a so-called radical resection to minimize the risk of tumor recurrence based on the theory that the cancer might be multi-centric in the pancreatic gland ( 8 ). However, the expected benefits were limited. TP was reported to be associated with even higher perioperative morbidity and mortality compared with PD ( 9 ). Furthermore, TP caused metabolic disorders, permanent pancreatic endocrine and exocrine dysfunction, which severely influenced patients’ quality of life (QoL) ( 10 ). Some studies also reported that TP was even associated with poorer long-time survival for PDAC compared with PD ( 9 , 11 ). With improvements in surgical techniques and perioperative management, developments in researching synthetic insulin and pancreatic enzymes, some scholars have argued that TP can be now safely performed, while the endo-exocrine insufficiency and QoL have greatly improved ( 12 , 13 ). Long-term survival of patients with PDAC following TP has also been reported similar to those following PD ( 14 , 15 ). However, other studies have reported that compared with PD, TP resulted in higher perioperative morbidity and mortality, or worse long-term survival in patients with PDAC ( 16 , 17 ). Taking these conflicting results into consideration, TP has been controversial and cannot be routinely recommended for patients with PDAC. Therefore, the present study aimed to comprehensively investigate the perioperative outcomes, long-term survival and postoperative QoL of PDAC patients following TP by comparing with those following PD. We present this article in accordance with the STROBE reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-2025-aw-544/rc ).

This was a double-center, retrospective, observational study. We reviewed the medical records of 68 patients with PDAC who underwent elective TP in the Department of Hepatobiliary Surgery at Beijing Chao-Yang Hospital and Beijing Hospital from January 2021 to December 2023. Meanwhile, we also randomly selected another 68 patients with PDAC who received PD during the same period for comparison. The indications for TP in the study were mainly multi-centric tumors, and tumors grown from pancreatic head to neck or body which resulted in repeated positive margins in intraoperative frozen section examination, mainly based on preoperative imaging evaluation, intraoperative exploration and intraoperative pathological examination. If PD was appropriate, we avoided TP. The exclusion criteria were as follows: (I) distant metastasis; (II) emergency or laparoscopic surgery; (III) neoadjuvant chemotherapy or radiotherapy; (IV) incomplete clinicopathological data. All the patients included were diagnosed with histologically proven PDAC. We collected data including the patients’ demographic, preoperative clinical features, intraoperative and pathological details, total hospitalization costs, perioperative outcomes, and overall survival (OS). Survival status of the patients was monitored by regular follow-up, and the end date of follow-up was December 31, 2024. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Beijing Chao-Yang Hospital (ethics approval No. 2024-D-512) and the requirement for informed consent was waived owing to the retrospective nature of the study. Beijing Hospital was also informed and agreed to the study. Additionally, we investigated the life state and QoL of the patients who were still alive by telephone conversation or sending them questionnaires online. A non-validated, self-designed concrete question sheet was used to acquire endocrine and exocrine insufficiency information, including the Bristol Stool Form Scale (BSFS) ( 18 ). The European Organization for Research and Treatment of Cancer Quality of Life Questionnaire C-30 (EORTC QLQ-C30), the European Organization for Research and Treatment in Cancer Pancreas 26 (EORTC QLQ-PAN26), the Problem Areas in Diabetes Questionnaire (PAID20) and the Diabetes Treatment Satisfaction Questionnaire, Status (DTSQs) were used to assess QoL of the survivors ( 19 - 22 ). The EORTC QLQ-C30 is a multi-dimensional questionnaire that contains 30 items covering global health status, functioning and symptom scales to assess the influence of cancer on QoL, while the QLQ-PAN26 module is designed to investigate pancreatic cancer-specific symptoms as a supplement to QLQ-C30. Higher scores of global health status and functioning scales represent better QoL, whereas higher scores of symptom scales indicate poorer QoL. PAID20 briefly measures the problems and distress related to diabetes, with higher scores indicating severer relevant problems. DTSQs measures the satisfaction with diabetes treatment, in which higher scores correspond to higher satisfaction. All survivors with postoperative diagnosis of diabetes were invited to complete the PAID20 and DTSQs. Perioperative mortality was defined as deaths happening during hospitalization or within 30 days after surgery. Postoperative complications were classified referring to the Clavien-Dindo grading system ( 23 ); grade I and II were minor complications, while grade III and IV were considered as major complications, and grade V was death of the patient. R1 resection was defined as microscopically tumor residual ( 24 ). The T and N stage accorded with the 8th American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) TNM classification system ( 25 ). Continuous variables were expressed as medians and interquartile range (IQR), or means and standard deviations. Categorical variables were expressed as absolute numbers and percentages. Normally distributed continuous variables were analysed using the Student’s t-test while non-normally distributed continuous variables were compared using the Mann-Whitney U test. Comparisons of categorical variables were performed using the Chi-squared test or Fisher’s exact test. Survival was estimated using the Kaplan-Meier method, and the log-rank test was used for comparison between groups. Two-tailed P values <0.05 were considered statistically significant. The EORTC QLQ-C30 and QLQ-PAN26 raw scores were converted to standard scores from 0 to 100 by a liner transformation ( 19 , 20 ). Scores of six questions of the DTSQs (except questions 2 and 3) were added up to get a total score on diabetes treatment satisfaction ranging from 0 to 36 ( 21 ). Scores of 20 questions of the PAID20 were added up and multiplied by 1.25 to create an overall score ranging from 0 to 100 ( 22 ). Statistical analysis was performed using SPSS version 25.0 (IBM Corporation, Armonk, New York, USA).

Nine patients were lost to follow-up. Except the five patients who died perioperatively, 79 patients had passed away, while 43 were still alive. We sent questionnaires to the 43 survivors. Figure 1 briefly showed the overall research procedure. Flow diagram showing patients included in the study. PD, pancreaticoduodenectomy; PDAC, pancreatic ductal adenocarcinoma; TP, total pancreatectomy. Table 1 showed the baseline descriptive data of all the patients. There were no significant differences between the TP and PD groups in gender, age, ASA scores, body mass index (BMI), diabetes, jaundice, weight loss, tobacco and alcohol use, albumin and CA19-9 level. As a whole, the baseline data were balanced between the two groups. Data are expressed as median [interquartile range] or n (%). ASA, American Society of Anesthesiologists; BMI, body mass index; CA19-9, carbohydrate antigen 19-9; PD, pancreaticoduodenectomy; TP, total pancreatectomy. The median operation time of the TP group was significant longer than the PD group (P<0.001). Portal vein or superior mesenteric vein resection and reconstruction was much more common in the TP group than in the PD group (P=0.001). Obviously, no POPF occurred in the TP group, which was overwhelmingly superior to the PD group (P<0.001), and postoperative hospital stay of the TP group was significantly shorter (P<0.001). Three patients in the TP group and two in the PD group died perioperatively due to severe complications. Concerning the pathological outcomes, tumor size (P<0.001) and number of examined lymph nodes (P=0.001) were significantly larger in the TP group. In both groups, more than half patients suffered postoperative complications, but fortunately the majority were minor complications (grade I–II). R1 resection rate, tumor differentiation, and TNM stage also showed no significant differences between groups. Over 60% of patients in both groups received postoperative adjuvant chemotherapy containing gemcitabine or fluorouracil-based regimen following the oncologists’ decisions. Six patients were diagnosed with chronic pancreatitis by postoperative pathological analysis, but none had typical clinical symptoms of chronic pancreatitis before surgery. According to the Fistula Risk Score (FRS) ( 26 ), there were two negligible risk patients, 26 low risk, 37 intermediate risk, and three high risk of POPF in the PD group. Postoperative pathology showed that 23 of 27 (85.2%) patients who underwent venous resection had vascular invasion. Table 2 displays the clinical and pathological outcomes. Data are expressed as median [interquartile range] or n (%). DGE, delayed gastric emptying; PD, pancreaticoduodenectomy; POPF, postoperative pancreatic fistula; TNM, tumor-node-metastasis; TP, total pancreatectomy. The follow-up duration of all patients (except the five patients who died perioperatively) ranged from 4 to 48 months, with a median of 16 (IQR, 10–23) months. The median OS of the TP group was 20 months while 21 months of the PD group, and there was no significant difference (P=0.91) ( Figure 2 ). OS at 1, 2, 3 years postoperatively were 72.0%, 41.4%, 29.0% for the TP group, while 73.5%, 43.1%, 33.2% for the PD group, respectively. Overall survival of patients in TP and PD groups. PD, pancreaticoduodenectomy; TP, total pancreatectomy. In the TP group, 17 survivors lost weight with a median of 7.0 kg, and four patients gained weight with a median of 2.5 kg. All patients took pancreatic enzymes. Only two patients suffered worsen diarrhea, while others said that the frequency and severity of diarrhea kept stable or turned better as time passed and were mainly satisfied with their stool control. Results of the PD group concerning exocrine function were similar to the TP group, except that dose of pancreatic enzymes intake was significantly smaller (P<0.001) and merely 15 survivors needed pancreatic enzymes postoperatively. The upper part of Table 3 showed the concrete data of exocrine function items. Values are expressed as median [interquartile range]. BSFS, Bristol Stool Form Scale; PD, pancreaticoduodenectomy; TP, total pancreatectomy. Undoubtedly, in the TP group, all the survivors were diagnosed with diabetes postoperatively and required insulin injection. Thirteen patients experienced hypoglycemia with a median of one time within the past one month, but they could quickly recover by taking a piece of candy or chocolate and none suffered loss of consciousness. In the PD group, 12 survivors developed diabetes and used insulin. Compared with the PD group, the TP group needed a significantly larger dose of daily rapid-acting (P=0.003) and long-acting (P<0.001) insulin injection, and revealed higher glycosylated hemoglobin level (P<0.001). The lower part of Table 3 displayed specific information about endocrine function items. Survivors with postoperative diagnosis of diabetes were also asked to complete the PAID20 and DTSQs to furtherly evaluate the impacts of diabetes. As shown in Table 4 , the relevant symptoms and treatment items were acceptable on the whole in both groups, with a median overall score of 22.5 and 19.4, respectively. There were no significant differences in each item and the overall score of PAID20 between the two groups. In the mass, patients in both groups were satisfied with their diabetes treatment with a median DTSQs overall score of 28.0 and 29.5, respectively. Results did not show any significant differences in each question item or the overall score of DTSQs between the two groups ( Table 5 ). Among the 21 survivors in the TP group, nine patients had diabetes preoperatively and 12 patients didn’t have diabetes. We compared survivors with and without diabetes before TP, and there were no significant differences in the overall score of both PAID20 (21.3 vs. 22.5, P=0.28) and DTSQs (28.0 vs. 29.0, P=0.13) between the two subgroups. Original source: Polonsky et al. ( 27 ). Scores were expressed as median and interquartile range. Each question is ranged from 0 to 4; 0= not a problem; 1= minor problem; 2= moderate problem; 3= somewhat serious problem; 4= serious problem. The overall score ranges from 0 to 100 (no problems to maximum level of problems). PD, pancreaticoduodenectomy; TP, total pancreatectomy. License is obtained from Health Psychology Research Ltd (Ref: CB1793). Scores were expressed as median and interquartile range. The overall score ranges from 0 to 36 (lowest satisfaction of diabetes treatment to highest satisfaction of diabetes treatment). PD, pancreaticoduodenectomy; TP, total pancreatectomy. In the QLQ-C30, the median global health status scores in the TP and PD groups were both 58.3%. Functioning scales had median scores ranging from 66.7% to 83.3% in both groups, indicating a relative good QoL in physical, role, emotional, cognitive and social functioning. When it came to symptom scales, median scores were relatively low, which represented a passably acceptable QoL. All the items of the QLQ-C30 had no significant differences between the two groups ( Figure 3 , Table S1 ). We also compared survivors with and without diabetes before TP, and there was no significant difference in the global health status score (58.3% vs. 62.5%, P=0.62) between the two subgroups. Results of EORTC QLQ-C30. Bars are shown as means and standard deviation. PD, pancreaticoduodenectomy; TP, total pancreatectomy. In the QLQ-PAN26, there were also no significant differences in all the items between the two groups, including pancreatic pain, eating-related problems, indigestion, cachexia, flatulence, body image, side effects, fear of future health, ability to plan future, altered bowel habit, and health care satisfaction ( Figure 4 , Table S2 ). Results of EORTC QLQ-PAN26. Bars are shown as means and standard deviation. PD, pancreaticoduodenectomy; TP, total pancreatectomy.

In this study, we conducted comprehensive comparisons of TP versus PD for PDAC patients from two medical centers in Beijing, China, involving the perioperative outcomes, long-term survival, and postoperative QoL. The results showed that for PDAC patients, TP can be performed with similar postoperative morbidity and mortality, total hospitalization costs, OS, and QoL compared with PD, which suggested that TP may be a safe and reasonable treatment option for PDAC. To the best of our knowledge, the present study is the first to compare these two surgical procedures for PDAC from multiple perspectives mentioned above. PDAC is widely regarded as the king of cancers, which leads to a very poor prognosis. Surgery plays a vital and dominant role in the process of PDAC treatment, and only margin-negative resection can offer patients a straw to acquire long-term survival. PD remains the most common procedure, as the tumor often locates in the head of pancreas. However, for cases that the lesions involve the central part, multiple parts or even the entire pancreatic gland, PD is no longer sufficient and TP is generally proposed to achieve complete resection. In the present study, the main indications for TP were neck-positive in intraoperative frozen section examination and multi-centric PDAC. In the early time, TP was advocated as a measure to treat potential multi-centric tumors and meanwhile, avoid POPF. Nevertheless, the disadvantages of TP became obvious after the initial enthusiasm. TP failed to improve the prognosis of PDAC patients but was associated with unfavorable higher perioperative morbidity and mortality as well as poorer long-term survival compared with PD ( 16 ), and often resulted in heavy metabolic disorders and poor QoL due to permanent pancreatic dysfunction ( 10 , 28 ). As time goes on, several studies in the new era have argued that TP can be performed with acceptable morbidity and mortality similar to PD ( 13 - 15 , 29 - 31 ). On the contrary, data of the National Surgical Quality Improvement Project (NSQIP) from America showed inferior perioperative outcomes for TP compared to PD, and thus concluded that TP cannot be routinely recommended ( 17 ). In some of those studies, the indications for TP included PDAC and some other pancreatic diseases. Stoop et al. ( 32 ) argued that an increase in surgical volume of TP was associated with improved perioperative outcomes. In the present study, we focused on PDAC patients. The patients had similar baseline characteristics, which guaranteed the comparability of the two groups. The results showed that the TP group needed a longer operation time and acquired more lymph nodes resection, since TP was a more extensive surgery and vascular resection and reconstruction was more often operated. Larger tumor sizes and more vascular resection were seen in the TP group, which was in accordance with previous studies ( 14 , 33 ). Higher rate of vascular resection can be partly explained by the location of the tumors. The overall morbidity and mortality were similar between the two groups. Although over half of patients in both groups suffered postoperative morbidity, most were minor complications which can be handled smoothly. POPF was eradicated in the TP group, which was an apparent advantage over the PD group, and thus contributed to a shorter postoperative hospital stay, since prolonged stay of the PD group was partly due to pancreatic fistula treatment. Three patients in the TP group who received vascular resection and reconstruction died perioperatively, resulting in a perioperatively mortality of 4.4%, as was reported of 0–8.5% in other studies ( 6 , 34 - 36 ). Total hospitalization costs, an economic index, can evaluate the cost-effect value of the clinical practice ( 13 ). Our data showed that the total hospitalization costs of the two groups were close, which indicated that TP was acceptable in terms of cost-effectiveness. Several studies in recent years have compared the long-term survival between TP and PD. Some scholars reported that the long-term survival of patients who underwent TP for PDAC was similar to those who underwent PD ( 14 , 15 , 30 ). A systematic review analysed six studies published from 2009 to 2016 including 316 PDAC patients and concluded that long-term survival of TP was comparable with survival after partial pancreatectomy ( 37 ). Using data from the Surveillance, Epidemiology, and End Results (SEER), a population-based database, Nathan et al. ( 38 ) also reported the same conclusion with above. On the contrary, Karpoff et al. ( 16 ) reported a worse survival following TP than PD for PDAC patients. In the present study, although the TP group was characterized by larger tumor sizes and more vascular resection, no significant difference was found in OS compared with the PD group, which supported the use of TP for PDAC when oncologically appropriate ( 15 , 38 ). Relevant studies have shown that venous resection and reconstruction has not been a contraindication for pancreatic surgery ( 39 , 40 ). In order to strive for R0 resection, vascular resection and reconstruction is performed in some cases. For cases with long resection vessels and difficult to directly reconstruct, allogeneic vascular reconstruction is feasible ( 41 ). Although the TP group had larger tumor size and more vascular resection, there was no significant difference in TNM stage between the two groups. R0 resection and postoperative adjuvant therapy might result in similar OS between the two groups. Attentions have also been paid to QoL after TP since the procedure causes significant anatomical and physiological changes that greatly interfere QoL of the patients. Exocrine insufficiency management has improved due to the progress in compound pancreatic enzymes. Weight loss and diarrhea can be often observed in patients following TP ( 34 , 42 , 43 ). The present study assessed the exocrine insufficiency of the survivors. Both groups had a high proportion of weight loss, but fortunately most of them said the diarrhea symptom kept stable or turned better as time went on. No surprisingly, patients in the TP group needed more intake of pancreatic enzymes. Weight changes, diarrhea, night stools, and BSFS score were similar between the two groups. In addition to exocrine insufficiency, the resection of the nervous plexus around the coeliac trunk and superior mesenteric artery may also aggravate the malabsorption after TP ( 44 ), and relative researches are needed to solve this issue in the future. Endocrine insufficiency after TP causes permanent pancreatogenic diabetes, which has been a barrier to implementing TP. Due to the increased peripheral receptor sensitivity to insulin and glucagon deficiency, patients after TP are more prone to hypoglycemic attacks ( 45 ), which increases the difficulty in blood glucose control. In this study, over 60% of survivors in the TP group experienced episodes of hypoglycemia with a median of one time in the past one month, but luckily none suffered loss of consciousness. TP patients are routinely referred to the endocrinology department in these two hospitals for better blood glucose management. As a result, the median Hb A1C level of the TP group was acceptable, as was reported from 7.4% to 8.0% in other studies ( 28 , 29 , 46 ), although it was significantly higher than 7.1% of the PD group in this study. PAID20 and DTSQs outcomes mirrored difficulties and problems in diabetes control, and degree of satisfactory in diabetes treatment. According to these two questionnaires, we found that the impacts of diabetes were comparable on the TP and PD groups, and patients in both groups were mainly satisfied with their diabetes treatment. We deem that robust consultation and guidance from the endocrinologists, good diabetic education, and enhanced self-management are strongly recommended. Three studies have compared postoperative QoL of patients who underwent PD and TP ( 13 , 29 , 33 ), but none were specifically for PDAC. These three studies drew a similar conclusion that QoL of patients after TP was acceptable and almost comparable to that of PD patients. Scholten et al. ( 47 ) performed a Dutch nationwide, retrospective study among patients who underwent TP in 17 medical centers from 2006 to 2016, and concluded that long-term QoL was lower compared to the general population with small differences. In the present study, by using the validated questionnaires, EORTC QLQ-C30 and QLQ-PAN26, we also found no significant differences in all functioning, symptom scales and global health status between the two surgical procedures for PDAC patients. Improvements in overall QoL may owe to the advancements in pancreatic endocrine and exocrine insufficiency management. The bad reputation of TP concerning its great adverse impact on QoL should be reconsidered with time going by. The results of this study should be interpreted cautiously due to several shortcomings. First, the design of this study was retrospective and cross-sectional, so we could not assess the life status of the patients who had died during follow-up. Second, the sample size of this study was relatively small and follow-up period was short. Moreover, life status after surgery may vary with time, but the QoL data we collected were at various time nodes. As for strengths, data of the study were from two surgery centers and were relatively comprehensive. We expect further prospective studies with larger sample sizes to prove and justify our research perspectives.

This study demonstrates that for patients with PDAC, the postoperative morbidity and mortality, total hospitalization costs, OS, and QoL including endo-exocrine insufficiency management results are similar and comparable between the TP and PD procedures. These data indicate that TP may be as safe, feasible and efficacious as PD in the treatment of PDAC.

The article’s supplementary files as