Abstract
Aim: The aim of the current study was to investigate whether the artery-first approach (AFA) improved surgical outcomes of pancreaticoduodenectomy (PD) at our non-high-volume center. Patients and Methods: We retrospectively reviewed data on 121 consecutive patients who underwent PD between January 2009 and December 2018. The perioperative data of 49 patients who underwent conventional PD (conventional group) and 72 patients who underwent PD via artery-first approach were analyzed and compared to assess the effectiveness of the AFA. Results: Although no significant difference was observed between the two groups overall, in those with pancreatic cancer, the duration of surgery, intraoperative blood loss and transfusion rate in the AFA group (n=33) were significantly lower than those for the conventional group (n=11) (p=0.011, p=0.021 and p=0.038 respectively). Conclusion: AFA can be used to reduce the operative time, intraoperative blood loss, and transfusion rate in patients with pancreatic cancer.
Pancreaticoduodenectomy (PD) is one of the most challenging and complex procedures among abdominal surgeries. It is commonly performed not only for malignant tumors of the pancreatic head, ampulla, and distal bile duct but also for benign tumors and peripancreatic trauma (1-5). This procedure is associated with a high rate of major complications, including postoperative pancreatic fistula (POPF), bleeding, intra-abdominal collection, and delayed gastric emptying (DGE), resulting in significant postoperative morbidity in 22-60% patients (6-9). A previous study evaluating the retrospective data on 8,575 PDs performed in 1,167 hospitals showed a 30-day mortality rate of 1.2% and an in-hospital mortality rate of 2.8%, which were much higher than those of reported for general abdominal surgeries (10).
Multiple retrospective studies using data from the United States have demonstrated better surgical outcomes of major pancreatic resections at high-volume centers than at non-high-volume centers (11-13). Volume may be a surrogate marker for high-quality care systems, surgical training, and experience. However, more recent analyses of the volume– outcome relationship have demonstrated that in addition to hospital volume, the characteristics of surgeons and system resources affect these qualities (14). Several studies have reported that the outcomes of hepatic and pancreatic resections at low-volume centers are comparable to those at high-volume centers (15, 16). We hypothesized that use of an appropriate surgical technique by well-experienced surgeons leads to sufficient outcomes after PD both in non-high-volume and high-volume centers.
In the artery-first approach (AFA), a prompt dissection is made between the pancreas and the superior mesenteric artery (SMA) ahead of the point of no return, such as pancreatic transection or bile duct division (17, 18). The AFA is recommended for the early determination of resectability in cases of borderline resectable cancer (19), ensuring SMA margin in advanced pancreatic cancer (20), reduction of intraoperative blood loss due to early ligation of the inferior pancreaticoduodenal artery (IPDA) (21), and identification and safeguarding of a replaced right hepatic artery (22). Inoue et al. reported that the blood loss, operative time, and POPF and DGE rates were significantly lower in the AFA-PD group (n=82) than in the conventional PD group (n=80) (17).
In this retrospective study, we have reported on the efforts taken at a non-high-volume center for performing PD safely and improving surgical outcomes using the AFA.
Patients and Methods
Patients. Between January 2009 and December 2018, 121 consecutive patients who underwent PD at Japan Community and Health Organization Kumamoto General Hospital were included in this study. Two expert hepatopancreatobiliary surgeons performed PD as an operator or a lead assistant. In total, 49 patients underwent conventional PD (control group) between January 2009 and January 2013 and 72 patients underwent PD via the AFA between February 2013 and December 2018. Written informed consent was obtained from all patients, and this study was approved by the Ethical Committee (Registry Number 1291).
Surgical technique. After making a reverse T abdominal incision, the peritoneal cavity was explored to confirm the tumor stage and operability. Following a wide Kocher maneuver, the para-aortic lymph nodes were explored and resected if necessary. The gastrocolic ligament and great omentum were dissected until the pancreatic head was well exposed. In the case of carcinoma of the pancreatic head, we chose the mesenteric approach to ensure a forefront margin. The superior mesenteric vein was taped and its branches dissected, including Henle`s gastrocolic trunk, inferior mesenteric vein and middle colic vein when there was no tumor invasion. After the SMA was recognized by palpation and its surrounding tissues were dissected, the procedure was switched to the left side. With the transverse colon retracted cranially, the proximal jejunum and mesojejunum corresponding to the sacrificed jejunum arteries were divided and dissected toward the left aspect of the SMA. The ligament of Treitz and the left side of the SMA were dissected. After completion of the left-side approach, the procedure was switched to dissection of the right side of the SMA. The superior mesenteric vein was retracted to the right and the plexus of the SMA was retracted to the left side. By retracting the smooth muscle actin plexus, the SMA was then easily rotated around the pinpoint to facilitate exposure of its branches. Then, the common trunk of the IPDA and jejunal artery was ligated (Figure 1A), and the first and second nerve plexus of the pancreatic head were dissected (Figure 1B). When the tumor was found to have invaded the right side of the smooth muscle actin plexus, the right hemicircumferential SMA plexus was chosen. The stomach was resected from the subtotal stomach-preserving PD line. After the hepatoduodenal ligament was dissected, the pancreas was transected. Then only the portal vein (PV) was connected to the specimen. When there was no tumor invasion, the branches of the PV to the pancreatic head were resected, otherwise the PV was resected and anastomosed.
The strategy of artery first approach and polyglycolic acid felt pasting method. A: The superior mesenteric vein was retracted to the right side and the superior mesenteric artery (SMA) was retracted to the left side. The common trunk of the inferior pancreaticoduodenal artery and jujenal artery (arrowhead) was ligated. B: The superior mesenteric vein was retracted to the right side and the SMA was retracted to the left side. The first and second nerve plexus of the pancreatic head (arrowhead) were dissected. C: The cut end of the pancreatic remnant was pasted with polyglycolic acid felt (arrowheads) using a fibrin sealant. PV: Portal vein.
Reconstruction was performed using a modified Child’s procedure, and end-to-side pancreaticojejunostomy (PJ) anastomosis was performed with a two-layer anastomosis technique. The cut end of the pancreatic remnant was joined together with polyglycolic acid (PGA) felts using a fibrin sealant (Figure 1C). We had introduced the PGA felt technique in January 2012. A pancreatic and biliary duct tube was placed as an external drainage stent. The gastrojejunal anastomosis was placed in an ante-colic position and the gastric remnant was fixed to the transverse mesocolon. The stump of the gastroduodenal artery (GDA) was covered by the round ligament of the liver (Figure 2A and B). Drains were routinely placed close to the PJ and biliary–jejunal anastomosis.
The strategy used to prevent postoperative bleeding. A: Stump of the gastroduodenal artery (arrowhead) and round ligament. B: The stump of the gastroduodenal artery was covered by round ligament (arrowhead). C: Remnant round ligament was placed between superior mesenteric artery and pancreaticojejunostomy anastomosis. D: Drain was placed on the dorsal side of pancreaticojejunostomy anastomosis.
Data recording. The intraoperative variables included operative time, blood loss, blood transfusion (with vs. without), texture of the pancreas, and pancreatic duct size (≤3 vs. >3 mm). The texture of the pancreas was classified as soft or hard based on the judgement of the operating surgeon, and the pancreatic duct size was measured at the cut surface of the remnant pancreas.
Postoperative variables included measurement of the amylase concentration in serum and drainage fluid on postoperative days 1, 3, 5, and 7. POPF was defined and graded according to the criteria proposed by the International Study Group on Pancreatic Fistula (23). POPF was defined as a drain output of any measurable volume of fluid on or after postoperative day 3 with an amylase concentration greater than three times the serum amylase concentration. The severity of POPF was categorized into three grades according to its clinical impact: A, no clinical impact; B, minor adjustment of the clinical pathway; and C, major change in clinical management. Computed tomography was performed on postoperative day 7, and the peripancreatic drains were removed when there were no abnormal findings. The pancreatic duct tube was removed at our outpatient clinic approximately 4 weeks after surgery.
Statistical analysis. All statistical analyses were performed using R (3.3.3, R Development Core Team, https://cran.r-project.org/) and GraphPad Prism V5.0 (GraphPad Software, San Diego, California, USA). The Mann-Whitney U-test, Wilcoxon’s signed-rank test, and Fisher’s exact test were used to analyze clinicopathological data. Associations between morbidity and clinicopathological factors were evaluated using univariate and multivariate logistic regression analyses with odds ratios (ORs). Logistic regression was performed using the “stats” package in R. ORs were calculated using the “vcd” package in R.
Results
Clinicopathological characteristics and intraoperative factors are shown in Table I. There were 67 men (55.4%) and 54 women (44.6%) with a median age of 72 years (range=34-88 years). The pathological diagnoses were pancreatic (n=44) or bile duct cancer (n=34) in the majority of cases. For patients with jaundice, preoperative biliary drainage was performed endoscopically. The median operative time was 383 min (range=245-625 min). The median blood loss was 579 g (range=143-2,675 g).
Clinical, epidemiological, and surgical factors of patients undergoing pancreaticoduodenectomy (n=121).
The incidence of postoperative complications is shown in Table II. The overall complication rate (Clavien-Dindo classification ≥II) was 21.5%. The overall POPF and POPF rates were 39.7% and 11.5%, respectively. Postoperative bleeding occurred in one (0.8%) patient and was treated with interventional radiology. One patient (0.8%) underwent reoperation due to leakage of the gastrojejunal anastomosis. No perioperative deaths were reported during the study period.
Short-term outcomes after pancreaticoduodenectomy (n=121).
Table III shows the comparison of perioperative characteristics between patients who underwent conventional PD (n=49) and those who underwent conventional PD via the AFA (n=72). No significant difference was observed between the groups in terms of operative time, blood loss, transfusion rates, POPF, and DGE. In pancreatic cancer, plexus dissection around the SMA is technically challenging, and the AFA may reduce the difficulty of the operative procedure (17). Therefore, perioperative characteristics were further compared between patients with pancreatic cancer in the conventional group (n=11) and those in the AFA group (n=33) (Table IV). The operative time, blood loss, and transfusion rate were significantly lower in patients with pancreatic cancer in the AFA group than in those in the conventional group (p=0.011, p=0.021, and p=0.038, respectively). No significant differences were observed between the groups in terms of postoperative complications.
Comparison of perioperative factors between patients undergoing pancreaticoduodenectomy by conventional and artery-first approach (AFA) (n=121).
Comparison of perioperative factors in patients with pancreatic cancer between those undergoing pancreaticoduodenectomy by conventional and those treated using an artery-first approach (AFA).
Discussion
PD is a complex procedure with high complication rates (24). Although morbidity and mortality have significantly improved in the past decade due to improvements in intensive care management, surgical techniques and devices, the rates of postoperative complications are high (25).
In this study, the overall complication rate was 21.5%, comparable to that reported in previous studies (22-60%) (6-9). Japanese retrospective data on 8,575 PDs showed that the rate of grade B and C POPF was 13.2% (10), similar to that reported in our study (11.5%). At our Institution, the mortality rate was 0%, which was lower than that reported in a Japanese study (2.8%) (10). Taken together, the postoperative outcomes of PDs performed at our Institute are considered acceptable compared to those reported in previous studies.
In a meta-analysis of AFA versus standard PD evaluating perioperative outcomes and survival, intraoperative blood loss and transfusion rates were significantly lower in the AFA group than in the standard PD group (26). Although the rates of perioperative mortality were comparable between the groups, perioperative morbidity and the incidence of grade B and C POPF were significantly lower in the AFA group (26). In this study, there was no significant difference in the operative time, blood loss, transfusion rates, and postoperative morbidity between groups. However, the operation time, blood loss, and transfusion rates were significantly lower in patients with pancreatic cancer in the AFA group than in those in the conventional group. There was no significant difference in complication rates, including the POPF rate, between groups. In pancreatic cancer, the SMA margin, i.e., the soft tissue directly adjacent to the proximal 3-4 cm of SMA, is the most common cause of R1 resection (positive for cancer cells at the margin), and dissection of the nerve plexuses of the pancreatic head is the most challenging process (27). In the AFA, the plexuses are dissected before transection of the neck of the pancreas, allowing early division of the IPDA and subsequent meticulous dissection along the SMA borders (26). These factors might have resulted in the reduced operation time, blood loss, and transfusion rates in patients with pancreatic cancer in the AFA group in this study.
POPF is the main trigger for other morbidities, such as pseudoaneurysm of the gastroduodenal artery stump, intraabdominal abscess, and sepsis, which are significantly associated with high mortality rates (28). The morbidity rate of POPF is 10-40% after PD (29-31) but there is no consensus regarding the best way to manage pancreaticoenteric anastomosis. PGA is a felt-like absorbable suture-reinforcing material generally used to reinforce sutures of fragile tissues such as the lung, bronchi, liver, and gastrointestinal tract, as well as a wide range of tissue defects (32). Regarding pancreaticojejunostomy using PGA felt, the incidence of POPF formation was lower than in the control group in some retrospective studies (33, 34). Furthermore, we have reported the usefulness of the PGA felt pasting method to reduce POPF in patients with soft pancreas. The overall incidence of POPF in the soft pancreas was significantly lower in the PGA group than in the control group (39.1% vs. 70.0%, p=0.042). We concluded that PGA felt should be used not to tightly seal the PJ anastomosis and block the liquid oozing from the cut end of the pancreas but should be used to reinforce the surface of the pancreatic remnant for tight fixation to the jejunum (35).
In this study, postoperative bleeding occurred in one (0.8%) patient treated with interventional radiology, a rate lower than that reported in a previous study (5.1%) (10). In this study, the stump of the GDA was covered by the round ligament of the liver to prevent erosion hemorrhage (Figure 2B). Furthermore, we placed the remnant round ligament between the SMA and PJ anastomosis to separate the stump of the IPDA from the PJ anastomosis (Figure 2C). The drain was placed on the dorsal side of the PJ anastomosis to absorb pancreatic enzymes when POPF occurred (Figure 2D). A previous study has shown the effectiveness of a falciform ligament wrap in preventing bleeding due to the rupture of the GDA stump after PD (36). Separation of the skeletonized visceral vessels from pancreatic enzymes results in a lower rate of pseudoaneurysm rupture. We believe that this procedure is simple and useful for preventing erosion hemorrhage.
The present study had several limitations. Firstly, this study was a single-institution, retrospective study. Secondly, the observation period was long, which may have resulted in historical bias regarding treatment strategy, perioperative management, and surgical devices that might affect short-term outcomes after surgery. Thus, further multi-institutional prospective investigations with a larger cohort are required to reconfirm the usefulness of the AFA in patients with pancreatic cancer in terms of reducing operative time and blood loss.
In conclusion, the AFA can be used to reduce the operative time, blood loss, and transfusion rate in PD even at non-high-volume centers, and various efforts, such as the induction of the AFA, may lead to comparable outcomes after PD.
Acknowledgements
The Authors thank all the people who contributed to this work.
Footnotes
Authors’ Contributions
T. Yamane wrote the article. D. Izumi and K. Horino planned the study concept and design. S. Kinoshita, C. Shirakami, K. Morita, and S. Ikeshima collected the clinical data. S. Shimada conducted critical revision of the article. H. Baba coordinated the study, oversaw collection and analysis of the results. All Authors discussed the data and commented on the article.
Conflicts of Interest
The Authors declare that they have no conflicts of interest in regard to this study.
- Received June 16, 2021.
- Revision received August 15, 2021.
- Accepted August 24, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
