Reduction of Postoperative Complications in Pancreatic Surgery by Standardizing Perioperative Management: An Observational Cohort Study

Abstract

Introduction

Resection for pancreatic malignancy remains the gold standard for cure. Postoperative morbidity continues to be high even after technical innovations. This study evaluates the effectiveness of a standard perioperative pancreatic oncological surgery step-by-step protocol in reducing organ-specific complications.

Methods

In this observational cohort study, we analyzed the outcomes of oncologic pancreatic head resections from 2015 to 2022 after the implementation of a standard perioperative fail-safe protocol and compared these data with a historical cohort (2013–2014). In the study group, all patients were treated with preoperative limited mechanical bowel preparation, administration of a somatostatin analog, and a “pancreatic duct tube” in pancreatoduodenectomy. The primary outcome measure was the occurrence of postoperative organ-specific complications.

Results

A total of 151 patients were included in this study. The rate of postoperative pancreatic fistula (grade B and C) in the fail-safe group was 4.2%. Other organ-specific complications as postoperative hemorrhage (2.5%) and delayed gastric emptying (9.2%) also occurred less frequent than before implementation of the fail-safe protocol.

Conclusion

The use of this standardized fail-safe protocol for oncologic pancreatoduodenectomy can lead to a low postoperative morbidity with improved surgical outcomes.

Introduction

Pancreatic surgical procedures remain the only curative remedy for the pancreatic malignancy. However, the postoperative complications after pancreatic oncological procedures remain as high as 41% [1]. The rates of postoperative medical complications such as strokes, respiratory distress, kidney injury, pneumonia, pulmonary embolism, and metabolic abnormalities range from 4% to 19% [2]. The medical complication rates have been dramatically reduced due to improved perioperative intensive care measures [3]. However, the rate of surgical complications after pancreatic surgery remains high. The four most frequently reported procedure‐related complications after pancreatic oncological surgery include postoperative pancreatic fistula (POPF), delayed gastric emptying (DGE), sepsis, and postoperative hemorrhage (POH) [4]. One of the most devastating complications after pancreatic surgery is the occurrence of POPF, which ranges from 24% [5] to 50% [1]. POPF itself can lead to a wide range of complications such as POH, intra-abdominal infection, sepsis, and surgical site infection (SSI), while increasing the postoperative mortality up to 25% [6]. Many studies have investigated the possible risk factors for POPF, including patient-dependent factors and different operative techniques [5, 7, 8], but a consensus on a standard approach to curtail the rates of POPF is still lacking. The rates of other organ-specific complications following pancreatic surgery such as postoperative DGE and POH have been reported to be as high as 40% [9] and 16.1% [10], respectively. In addition to the organ-specific complications after pancreatic surgery, infectious complications play a major role in contributing toward the morbidity and potential mortality. SSI is analyzed in numerous studies and occurs in up to 30% of cases, with an incisional SSI rate of 9% [11, 12]. The occurrence of SSI is directly associated with organ-specific complications such as POPF [12]. All complications, whether organ-specific or otherwise, are commonly graded according to the International Dindo-Clavien grading system [13]. The section of the International Dindo-Clavien grading system, focusing on pancreatic surgery [14], has explicitly elucidated a high correlation between postoperative pancreatic complications and clinical outcomes [15]. Historically, every single postoperative complication leads to an extended hospital stay with an estimated increase in postoperative mortality. Henceforth, an early recovery after pancreatic surgery is crucial. In this perspective, surgical pathways including the fast-track surgery, first described by Kehlet et al. [16] in 1997, have been introduced to reduce postoperative length of stay and thereby systemic response to the surgical trauma. This ERAS pathway was adopted step by step for different organs and the pancreatic surgery pathway was published in 2020 [17]. While the ERAS pathway is focusing on a short length of stay, a standardized protocol reducing postoperative complications is missing. Therefore, a fail-safe protocol was established in the study center for colorectal surgery showing promising results [18].

The aim of this study was to evaluate the impact of a structured and standardized perioperative fail-safe protocol for oncologic pancreatic head resection following the established fail-safe approach [18]. This can potentially improve surgical outcomes by reducing postoperative organ-specific complications such as POPF.

Methods

In this observational cohort study, we collected the data of all patients undergoing pancreatic surgery for pancreatic cancer at Reinbek Hospital St. Adolf-Stift, Germany, from January 2013 till December 2022. In 2015, a standard perioperative surgical pathway was established, which is summarized in Table 1. We prospectively collected the data of patients undergoing pancreatic surgery for cancer between January 2015 and December 2022 (fail-safe group). Therefore, patients undergoing pancreatoduodenectomy for malignancy were included to determine the implication of the standardized surgical fail-safe pathway on organ-specific complications. During the preoperative work-up, we recruited only those patients, who were diagnosed with resectable pancreatic lesions. This also included locally advanced cases with the need of extended resections. We excluded patients with initial total pancreatectomy (planned or within the primary procedure due to tumor extend) because of a different spectrum of postoperative complications without the risk of POPF and patients with a distal splenopancreatectomy to enhance comparability. Patients with a rescue pancreatectomy due to bleeding or fistula after an initial pancreatoduodenectomy were included. All procedures were performed or supervised by two experienced pancreatic surgeons. Both surgeons have completed their learning curve prior to this study period and have performed more than 40 pancreatic resections per year continuously. The study center provides all necessary facilities for a multidisciplinary treatment of pancreatic cancer. This includes but not limited to a well-trained interventional radiology department performing around 1,000 interventions annually within the whole study period. There is also an intensive care unit with continuous up-to-date intensive care provided by an anesthetist specially trained in intensive care treatment. The study center also has an endoscopy unit with the possibility for endoscopy and interventions and is certified by the German Cancer Society.

Table 1.

Perioperative pathway for pancreatoduodenectomy implemented in 2015 and used in the study group

	Study group
Before surgery
Oral bowel preparation	Yes (1L Endofalk®)
Overnight fasting	Yes
Antibiotic prophylaxis	Yes
Mid-thoracic epidural anesthesia	Yes
Intraoperative
Somatostatin analog	Yes (octreotide 0.2 mg intravenously)
Lost pancreatic tube	Yes (6-Fr or 7-Fr silicon tube)
Pancreaticojejunostomy	6-0/5-0 singular and 4-0 running double layer
Devices conducting heat during resection	No
Lymphadenectomy	DII
Bowel reconstruction	Single loop
Restrictive blood transfusion	Yes (patient blood management)
Drainage	Yes (4 silicone tubes)
Nasogastric tube	Yes
After surgery
Somatostatin analog	Yes, 0.1 mg every 8 h for 5 days
Day 1	Fasting
Day 2	Sips of water
Day 3	Water
	Removal of urinary catheter
Day 4	Oral liquids
	Removal of nasogastric tube
Day 5	Soft solid food
	End of somatostatin analog
Day 6	Regular diet

Patients’ Characteristics and Definitions

The medical records were reviewed for the patients’ demographics, body mass index, the American Society of Anesthesiologists (ASA) classification of physical health, surgical data including approaches and procedures, length of surgery, length of hospital stay, complications according to the Dindo-Clavien’s classification [13], and mortality. Extended pancreatic surgery was defined in accordance to the International Study Group of Pancreatic Surgery (ISGPS) [19]. This included but was not limited to an extended gastric resection, resection of the colon and/or mesocolon, small bowel, adrenal gland, kidney, liver, or diaphragmatic crura. In addition, as defined by the ISGPS vascular resections like portal or mesenteric vein, inferior vena cava or hepatic artery, celiac trunk or superior mesenteric artery were counted as extended resections as well. The postoperative morbidity was defined as any complication occurring within the same hospital stay. POPF was defined and graded according to the updated definition by the ISGPS [20]. Accordingly, a grade A fistula was described as a biochemical leak, defined as an asymptomatic fistula with elevated drain amylase. Grade B fistula was considered when it needed a change of management including percutaneous drainage, while grade C was defined as fistulas associated with an organ failure, reoperation, or death [20]. In our study, the elaborated definition of the ISGPS for DGE was used whereas grade B and C was combined as “DGE yes” presenting vomiting or gastric distension [21]. Similarly, the definition for POH, as proposed by the ISGPS, was used as “hemorrhage yes” in case of grade B and C. Grade A shows no therapeutically modified consequence (Table 2) [22]. We described the postoperative pathological staging of the tumor using the UICC classification [23].To compare the results of the standardized fail-safe approach, we analyzed a historical cohort from the study center before implementation of the fail-safe protocol (2013–2014).

Table 2.

Organ-specific complications according to the International Study Group of Pancreatic Fistula and the International Study Group of Pancreatic surgery

Grade	POPF	DGE	POH
A	Asymptomatic fistula with elevated drain amylase (biochemical leakage)	Need of nasogastric tube for 4–7 days or reinsertion after POD 3	Early onset, mild bleeding, good clinical conditions, no therapeutical consequence
B	Change in management needed (percutaneous drain)	Need of nasogastric tube for 7–14 days or reinsertion after POD 7	Early or late onset, severe bleeding, well to intermediate clinical conditions, needs transfusion or intervention
C	Organ failure, reoperation, death	Need of nasogastric tube more than 14 days or reinsertion after POD 14	Late onset, severe or life-threatening clinical conditions, intervention, or re-operation

POD, postoperative day.

Control Group

We retrospectively analyzed all patients with oncologic pancreatic head resection who underwent duodenopancreatectomy from January 2013 to December 2014. No previous data were available due to a different documentation system in the study center. In this time period, no structured perioperative protocol was used and the only standardization was the performance by an experienced surgeon. Some parts of the afterward implemented fail-safe protocol have been used already.

Study Group

The standardized approach in pancreatic surgery was changed in some key elements. This included a limited preoperative colonic irrigation with 1 L of Endofalk^® for mechanical bowel preparation (MBP) to mitigate the risk of postoperative bowel obstruction due to an increased use of opioids and postoperative ileus. Also a new component was the immediate administration of a loading dose of 0.2 mg of somatostatin analog (octreotide) intravenously before the start of the pancreatic resection, in order to block the exocrine pancreatic secretion. Patients continued to take another dose of 0.1 mg octreotide every 8 h subcutaneously for the first 5 postoperative days (PODs) to prevent a stimulated exocrine secretion. PD was done following the pyloric preserving technique according to Traverso and Longmire [24]. This technique includes a single loop reconstruction with a reduced number of anastomoses in comparison to the formerly used three-loop reconstruction. In case of pylorus resection, the reconstruction after PD was according to Whipple’s procedure. A double-loop technique as used in Roux-en-Y procedure was performed. We excluded all energy devices that could potentially conduct heat to the pancreatic gland during the resection. Monopolar or bipolar electric knifes and LigaSure^™ were used during the preparation. Finally, the pancreas was resected with a scalpel when performing PD. The technique of the anastomosis was changed within the protocol. Depending on the thickness and resilience of the tissue, any bleeding from the surface of the remnant pancreas was sutured with a singular 6/0 PDS^™ or 5/0 PDS^™. For the first anastomosis, the proximal jejunal loop was transpositioned through the mesocolon of colon transversum. We began the reconstruction of the pancreato-jejunostomy with a running 4/0 PDS^™ suture as a termino-lateral connection between the dorsal pancreatic tissue and the jejunal serosa. After a jejunal incision, at least 12 circular duct-to-mucosa, either 6/0 PDS^™ or 5/0 PDS^™, single sutures were placed. Another key element in the standardized approach is a 6-Fr or a 7-Fr silicon tube (commonly using the end of a double-J ureter tube), depending on the diameter of the pancreatic duct, which was inserted into the pancreatic duct and the jejunum, covering the single-sutured anastomosis and insuring the initial pancreatic drainage during the first PODs. The tube was fixed with a 5/0 vicyl rapid^™ single suture to the mucosa (Fig. 1). After the closure of the pancreatic duct, a running 4/0 PDS^™ suture between the ventral aspect of the pancreatic tissue and jejunal serosa completed the anastomosis (Fig. 1). Using the same loop, a termino-lateral choledocho-jejunostomy with a circular running 5/0 PDS^™ suture would follow 15 cm behind the former pancreato-jejunostomy. After preserving the pylorus during resection, a duodeno-jejunostomy was performed with a running 4/0 PDS^™ circular suture. In a case of pylorus resection, the gastro-jejunostomy would be performed using a separate loop and a Roux-en-Y jejuno-jejunostomy at 60 cm distance to gastro-jejunostomy, avoiding a possible symptomatic reflux. After reconstruction, one drain was placed near the pancreatic anastomosis while the other drain was positioned near the hepatico-jejunostomy. Drains were removed depending on the volume and lipase activity after oral nutrition. Postoperatively, oral nutrition was resumed on the third postoperative day.

Fig. 1.

Lost pancreatic tube within the pancreatic duct. a Inserting the lost tube (blue arrow) into the pancreatic duct. b Pancreatic anastomosis with the 5/0 vicyl rapid^™ single suture fixing the tube (white arrow). c Performed anastomosis. P, pancreas; J, jejunum. Dashed arrows: sutures for the duct-to-mucosa-anastomosis.

Statistical Analysis and Ethical Approval

Statistical analyses were done using IBM SPSS Statistics version 25 (IBM Co., Armonk, NY, USA) and a p value <0.05 was considered statistically significant. Nominal variables were listed as mean with standard deviation and categorical variables as numbers with percentages.

In our study, the data are reported in concordance with the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) criteria [25]. This study was conducted in accordance with the Declaration of Helsinki [26] and was approved by the Ethics Committee of the Medical Association Schleswig-Holstein and registered at the German Clinical Trial Register (Register-ID DRKS00029753).

Results

Within the study period, 342 patients had surgical resections of the pancreas between January 2013 and December 2022. Of these, 86 patients were operated due to benign lesions, 64 patients had a complete pancreatectomy, and 41 had a distal pancreatectomy. These patients were excluded from the study cohort. The remaining 151 patients met the inclusion criteria for our observational cohort study. A total of 32 patients were treated before the implementation of the standardized protocol and form the control group. A total of 119 patients were treated in the standardized fail-safe approach and included in the study group. All these 119 patients received all protocol components as mentioned in the method section (Table 1). A detailed overview of the characteristics of 151 patients is shown in Table 3. There were 86 men and 65 women with a mean age of 70.78 ± 10.41 years. The average body mass index of the patients was 25.34 ± 4.20 and a majority (47.9%) was classified as ASA 2. In this cohort, 125 (82.8%) patients were managed by a pyloric preserving technique. In the fail-safe group, extended resections according to the ISGPS were performed in 30.3% of the cases. This included 15.1% vascular resections and 22.7% multivisceral resections, whereas these included 5 extended gastric resections, 7 resections of the colon, mesocolon, or small bowel, 1 kidney resection, and 15 liver resections. The overall rate of POPF (grade A–C) in the study group was 9.2%. Excluding the biochemical leakage (formerly classified as POPF grade A), the reported fistula rate was with 4.2% grade B and C POPF even below this. Within the reported standardized fail-safe approach, the rate of POH was reduced to 2.5%. DGE was diagnosed in 11 patients in the study group (9.2%). Following the Dindo-Clavien’s classification, major complications (Dindo-Clavien ≥3) were reported in 28 (23.5%) patients in the fail-safe group. These included 14 (11.8%) re-operations. Table 4 provides a detailed overview about the perioperative parameters and the postoperative outcome. The postoperative 30-day-mortality in the study cohort was 4.2% (Table 5).

Table 3.

Demographic characteristics, pathological classifications, and surgical resections of pancreatic malignancy in the study (n = 151)

Grade	Total (n = 151)	Control group (n = 32)	Fail-safe group (n = 119)	p value
Age, M±SD, years	70.7±10.41	70.69±8.60	70.81±10.88	0.814a
Sex				0.689b
Male	86 (57.0)	17 (53.1)	69 (58.0)
Female	65 (43.0)	15 (46.9)	50 (42.0)
ASA classification (missing data in 5 cases)				0.483c
ASA 1	5 (3.4)	0 (0.0)	5 (4.2)
ASA 2	67 (45.9)	12 (44.4)	55 (46.2)
ASA 3	70 (47.9)	15 (55.6)	55 (46.2)
ASA 4	4 (2.7)	0 (0.0)	4 (3.4)
BMI, M±SD	25.34±4.20	25.83±4.28	25.71±4.21	0.951a
Malignancy				0.172c
Adenocarcinoma	135 (89.4)	31 (96.9)	104 (87.4)
Neuroendocrine tumor	4 (2.6)	1 (3.1)	3 (2.5)
Otherd	12 (7.9)	0 (0.0)	12 (10.1)
Pathological tumor classification, n (%)				0.018c
pT1	16 (10.6)	4 (12.5)	12 (10.1)
pT2	45 (29.8)	4 (12.5)	41 (34.5)
pT3	81 (53.6)	24 (75.0)	57 (47.9)
pT4	9 (6.0)	0 (0.0)	9 (7.6)
Neoadjuvant treatment, n (%)	7 (4.6)	2 (6.3)	5 (4.2)	0.640b
R0 resection, n (%)	128 (84.8)	26 (81.3)	102 (85.7)	0.581b
Extended pancreatoduodenectomy	44 (29.1)	7 (21.9)	37 (31.1)	0.384b
Vascular resection, n (%)	24 (15.9)	6 (18.8)	18 (15.1)	0.594b
Multivisceral resection, n (%)	31 (20.5)	4 (12.5)	27 (22.7)	0.323b

M, mean; SD, standard deviation; ASA, American Society of Anesthesiologists; BMI, body mass index.

^aMann-Whitney U test.

^bFisher’s exact test.

^cChi-square test.

^dIncluding sarcoma, distal cholangiocarcinoma, rare cancers.

Table 4.

Surgical outcome within the study period (n = 151)

	Total (n = 151)	Control group (n = 32)	Fail-safe group (n = 119)	p value
Length of hospital stay, M±SD, days	18.29±12.46	26.13±17.92	16.18±9.58	<0.0001a
Time on ICU, M±SD, days	4.24±4.97	6.44±7.86	3.65±3.67	0.00012a
Intraoperative transfusion pRBC, n (%)	16 (10.6)	6 (18.8)	10 (8.4)	0.108c
Operation time, M±SD, min	367.97±83.75	325.66±77.29	379.34±82.04	0.004a
Number of resected lymph nodes, M±SD	25.90±10.33	20.43±6.91	27.29±10.62	0.001a
Intraoperative crystalloids, M±SD, mL	4,232.82±1,452.18	5,250.00±1,777.38	4,130.25±1,383.10	0.034a
Complications
Dindo-Clavien classification, n (%)				0.056b
0	64 (42.4)	7 (21.9)	57 (47.9)
I	2 (1.3)	0 (0.0)	2 (1.7)
II	46 (30.5)	16 (50.0)	30 (25.2)
III	28 (18.5)	6 (18.8)	22 (18.5)
IV	5 (3.3)	2 (6.3)	3 (2.5)
V	6 (4.0)	1 (3.1)	5 (4.2)
Major complication DC ≥3, n (%)	37 (24.5)	9 (28.1)	28 (23.5)	0.645c
POH, n (%)	8 (5.3)	5 (15.6)	3 (2.5)	0.011c
DGE, n (%)	20 (13.2)	9 (28.1)	11 (9.2)	0.015c
POPF, n (%)	27 (17.9)	16 (50.0)	11 (9.2)	<0.001c
POPF grade A (biochemical leakage)	16 (10.6)	10 (31.3)	6 (5.0)	<0.001c
POPF grade B and C, n (%)	11 (7.3)	6 (18.8)	5 (4.2)	0.012c
POPF grade B	9 (6.0)	4 (12.5)	5 (4.2)	0.096c
POPF grade C	2 (1.3)	2 (6.3)	0 (0.0)	0.044c
Re-operation, n (%)	19 (12.6)	5 (15.6)	14 (11.6)	0.555c
Wound infection, n (%)	14 (9.3)	3 (9.4)	11 (9.2)	1.00c
Pneumonia, n (%)	22 (14.6)	5 (15.6)	17 (14.3)	0.785c
Mortality, n (%)	6 (4.0)	1 (3.1)	5 (4.2)	1.00c

M, mean; SD, standard deviation; pRBC, packed red blood cells.

^aMann-Whitney U test.

^bChi-Square test.

^cFisher’s exact test.

Table 5.

Postoperative mortality

Patient	Year	Procedure	Cause of death	Day of death
Control group
1	2013	PD	Multiorgan failure	44 POD
Study group
1	2016	PD + hemicolectomy, reconstruction mesenteric veins/portal vein	Hepatic failure	3 POD
2	2017	PD	Cardiac arrest	28 POD
3	2017	PD	Lung embolism	1 POD
4	2018	PD + right hemihepatectomy + portal vein reconstruction	Ventricular fibrillation	2 POD
5	2020	PD + right hemihepatectomy	Liver failure after hemihepatectomy	3 POD

PD, pancreatoduodenectomy; POD, postoperative day.

Discussion

Our study has shown that a fail-safe protocol in oncologic pancreatic head surgery can lower the rates of organ-specific pancreatic complications with an overall POPF grade B/C rate of 4.2% versus 18.8% (p = 0.012) before implementation of the fail-safe protocol. Following the low rate of POPF, the rates of other related complications such as POH (2.5%) and DGE (9.2%) were also low. These improved clinical indicators resulted in a shorter length of postoperative hospital stay with a consequent positive impact on ERAS. This is in line with the results of such a perioperative fail-safe approach in colorectal cancer surgery [18].

Preoperative Set-Up

Among a wide spectrum of the preoperative work-up in pancreatic surgery, MBP using 1 L of Endofalk^® is generally considered to reduce the risk of postoperative ileus and delayed mobilization. Only one study has compared the impact of preoperative MBP with liquid diet without outright advantages of MBP [27]. Our study has shown a low rate of DGE and highlights the effectiveness of a targeted and focused MBP in pancreatic surgery. The use of perioperative antibiotics is still debatable. Degrandi et al. [28] showed a low rate of postoperative SSIs and pulmonary complications in PD by using a perioperative antibiotic protocol. In our study, the rate of SSI remained as low as 9.2%. This might be attributed to the use of a single shot antibiotic, close attention to hemostasis, and the quality of barrier nursing as proposed by the ERAS protocol for pancreatic surgery [17].

Surgical Approach

A large body of literature has deduced that the outcome of pancreatic surgery is strongly correlated with the surgeons’ experience. This underpins the usefulness of centralization in specialized centers [1, 29]. In our study, all pancreatic procedures were performed or supervised by an experienced pancreatic surgeon, who had already completed his learning curve prior to this study. A distinguishing feature of our fail-safe protocol is the intraoperative administration of octreotide as a somatostatin analogs. Somatostatin analogs potentially reduce the rate of POPF by exhibiting their inhibitory effects on the exocrine secretion of the pancreas. In their randomized controlled trial, Gouillat et al. [30] have found a significant reduction in the rates of pancreatic fistula using octreotide. A meta-analysis, comparing the results of 15 studies with 1,352 patients, showed a positive effect of somatostatin analogs in preventing POPF, but the results were inconsistent [31]. El Nakeeb et al. [32] did not find any benefit of the prophylactic use of octreotide in their randomized controlled trial. Our study showed a POPF rate of 4.2% that might have been influenced by the combination of the intra- and postoperative use of octreotide. The most popular surgical technique in PD is the creation of pancreaticojejunostomy anastomosis [33]. We also performed pancreaticojejunostomy in our series, though a pancreaticogastrostomy anastomosis might be an alternative. Kleespies et al. [34] compared a range of surgical techniques and have argued that the most important discriminator in reducing the rates of POPF was the surgeons’ experience instead of the technique itself. On the other hand, a study showed no difference in the POPF rates when comparing both types of reconstruction [35]. Apparently, this seems to be the surgeon’s preference rather than a scientific reason for the choice of surgical reconstruction in pancreatic surgery. Intraoperative stenting of the pancreatic duct is a commonly practiced surgical step, which aims to prevent obstruction of the pancreatic duct by suturing or postoperative drainage issues. In a propensity score-matched analysis by Kawaida et al. [4], the authors have argued that external stenting was associated with a higher rate of POPF. Even internal stents had to be removed within first few months after surgery. The benefits of a short internal stent was first reported by Hirashita et al. in 2020 [36]. The technique of internal stenting was adopted in our procedure using a “lost” pancreatic duct tube fixed by a rapidly absorbable suture. Using a rapidly absorbable suture with internal stenting prevents additional postoperative requirements to remove stents. Such stenting might lead to the further reduction of the rates of POPF compared to other clinical pathways in pancreatic surgery as reported by Téoule et al. [37].

Postoperative Settings

In our study group, enteral nutrition was started on the third POD. Optimal nutrition route in the postoperative phase is still under discussion and ranges from early oral intake to feeding jejunostomy [38]. In accordance with the ERAS guidelines in pancreatic surgery, early nutritional support activates the neuro-hormonal signals for the motor and secretory functions of the gastrointestinal tract, which could possibly enhance early recovery [17]. Drainage tubes are routinely placed during pancreatic surgery, to be removed depending on the volume (≤100 mL in 24 h), and low lipase levels usually after first oral nutrition. The placement and time point for the removal of drains is still under intense discussion. As these tubes drain blood, bile, and pancreatic secretion, they can lead to an intra-abdominal infection due to a retrograde contamination [4]. A host of studies have failed to find any advantage of the routinely placed intraoperative drains in elective pancreatic surgery [39–41]. In contrast, the randomized controlled trial of Van Buren et al. [42] was suspended due to an increased mortality rate from 3% up to 12% in the non-drainage group. However, we did not find any issue with intra-abdominal drains and, rather, drains guided us about the nature and volume of contents. A recently published work from Adamenko et al. [43] described the use of a prophylactic irrigation near the anastomosis. This showed a decreased rate of POPF, which remained with 12.7% more than three times higher than the data reported here [43].

In the study group, the rate of extended pancreatic resections according to the ISGPS seems 30.3% relatively high including a multivisceral resection rate of 22.7%. This is in line with the literature reporting multivisceral resections of up to 50% [44–47]. This topic is still under intense discussion, even so several reviews encourage a multivisceral resection in selected patients [48–50].

In our study, the postoperative in-house-mortality, as a major outcome parameter, was reported to be 4.2%. Literature has shown a mortality rate of approximately 3% in oncological pancreatic surgery [42]. In contrast, data from large German registry showed a nationwide mortality after pancreatic oncological surgery to be about 10% [51, 52]. Nimptsch et al. [51] reported a German in-hospital mortality after oncological pancreatic head resection of 8.2%. Even mortality rates from high volume centers in Germany are reported to be 7% in pancreatic oncological resections [53]. As Adamenko et al. [43] presented a mortality rate of 6.8%, this seems to be not limited to Germany. The perceived increased postoperative mortality in our study group may be attributable to the high rate of multivisceral resections and more advanced tumors treated in the fail-safe group. In this group, 2 patients died after PD and simultaneous hemihepatectomy. Nimptsch et al. [51] reported an in-house mortality of 17.6% after simultaneous resections of a visceral artery and 25.4% after simultaneous hepatic resections. Therefore, extensive multivisceral resections should be used only in selected patients within an optimal multimodal concept and a two-stage procedure to evaluate the response to preoperative chemotherapy should be considered [54].

Study Limitations

Major limitation of our study remain its retrospective design in a single center. This retrospective design prevented us to report long-term-follow-up data for disease-free and overall survival. A small group of patients and a heterogenous representation of malignant lesions is another limitation of our study. The statistical analysis was limited by the comparison to a historical cohort from the same institution. As this is an observational study, we report data from daily clinical practice. In this cohort, decision-making was based on standardized operating procedures and individual case discussion within the multidisciplinary team. This also plays a key role in algorithm-based decision-making as presented in the PORSCH.Trial [55]. Our study findings need to be validated by large multi-center cohort studies. We postulate that the marked reduction in the rate of POPF may also be influenced by a single surgeon’s expertise, the growing experience of the surgical team, or using high-quality surgical instruments. Not all aspects of our standardized approach are evidence based and several components of this protocol are still under trial. As it is already known from other standardized surgical approaches, the positive effects of the fail-safe approach reported in this study cannot be attributed to a single intervention in the bundle; therefore, it remains unclear if a reduction in complications was caused by one component or by a combination of components in the protocol collectively as this was also discussed for the ERAS approach [56].

Conclusions

This cohort study evaluates a single-institution’s analysis of the impact of a standardized perioperative fail-safe approach in pancreatic oncological surgery. This approach encompasses pre-, peri-, and postoperative management steps that led to a low rate of POPF, POH, and DGE following pancreatic surgery. The relatively short length of hospital stay in the study cohort might contribute to an improvement of short- and long-term oncological outcomes and encourage the use of a fail-safe approach in pancreatic surgery.

Statement of Ethics

This study protocol was reviewed and approved by Ethics Committee of the Medical Association Schleswig-Holstein and registered at the German Clinical Trial Register (Register-ID DRKS00029753). In accordance to the Ethics Committee of the Medical Association Schleswig-Holstein, written informed consent was not required due to the retrospective study design.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This study was not supported by any sponsor or funder.

Author Contributions

J.H. and H.H. made the concept for this research project. J.H. performed the data curation and investigation. J.H. and S.Y.G. made the methodology and formal analysis. T.S., P.P., S.Y.G., and H.H. supervised the project. J.H., S.Y.G., and H.H. wrote the initial manuscript. T.S., J.H., T.S., M.A., P.P., S.Y.G., and H.H. reviewed and edited the manuscript. All authors approved the final version of the manuscript.

Funding Statement

This study was not supported by any sponsor or funder.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.