Chyle leak following root of mesentery dissection in pancreaticoduodenectomy with inferior infracolic superior mesenteric artery first approach

Abstract

BACKGROUND

The root of mesentery dissection is one of the critical maneuvers, especially in borderline resectable pancreatic head cancer. Intra-abdominal chyle leak (CL) including chylous ascites may ensue in up to 10% of patients after pancreatic resections. Globally recognized superior mesenteric artery (SMA) first approaches are invariably performed. The mesenteric dissection through the inferior infracolic approach has been discussed in this study emphasizing its post-operative impact on CL which is the cornerstone of this study.

AIM

To assess incidence, risk factors, clinical impact of CL following root of mesentery dissection, and the different treatment modalities.

METHODS

This is a retrospective study incorporating the patients who underwent dissection of the root of mesentery with inferior infracolic SMA first approach pancreatoduodenectomy for the ventral body and uncinate mass of pancreas in the Department of Gastrointestinal and General Surgery of Kathmandu Medical College and Teaching Hospital from January 1, 2021 to February 28, 2024. Intraoperative findings and postoperative outcomes were analyzed.

RESULTS

In three years, ten patients underwent root of mesentery dissection with inferior infracolic SMA first approach pancreatoduodenectomy. The mean age was 67.6 years with a male-to-female ratio of 4:5. CL was seen in four patients. With virtue of CL, Clavien-Dindo grade II or higher morbidity was observed in four patients. Two patients had a hospital stay of more than 20 days with the former having a delayed gastric emptying and the latter with long-term total parenteral nutrition requirement. The mean operative time was 330 minutes. Curative resection was achieved in 100% of the patients. The mean duration of the intensive care unit and hospital stay were 2.55 ± 1.45 days and 15.7 ± 5.32 days, respectively.

CONCLUSION

Root of mesentery dissection with lymphadenectomy and vascular resection correlated with occurrence of CL. After complete curative resection, these were managed with total parenteral nutrition without adversely impacting outcome.

Key Words: Chyle leak; Chylous ascites; Clavien-Dindo; Pancreatoduodenectomy; Root of mesentery dissection; Superior mesenteric artery first approach; Total parenteral nutrition; Uncinate mass; Vascular resection; Ventral body mass

Core Tip: Pancreatoduodenectomy is invariably discussed in terms of its weightage of surgery and frequent complications like post-operative pancreatic fistula, post-pancreatectomy hemorrhage, and delayed gastric emptying. However, chyle leak is rarely discussed. This article is a testament to the outcomes of the root of mesentery dissection with the virtue of achieving complete oncological resection. This article discusses the incidence, clinical impact, and risk factors of chyle leak following the root of mesentery dissection in pancreaticoduodenectomy, emphasizing effective management without re-exploration but utilizing total parenteral nutrition and dietary restrictions without adverse effect on the outcome.

INTRODUCTION

Achieving a complete removal of the malignancy (R0 resection) is essential in surgeries for most surgeries, especially pancreatic cancer. Recent advances in surgical techniques highlight the importance of thorough dissection, especially around critical areas like the retroperitoneal and vascular margins, to improve patient outcomes. Nakao[1] introduced the “mesenteric approach” during pancreaticoduodenectomy (PD), a procedure for intervening in pancreatic head cancers. This technique avoids direct contact with the tumor initially, helping to prevent any spread of the cancer cells. Instead, the surgeon starts by carefully dissecting the tissue along the mesentery, targeting areas like lymph nodes, arteries, veins, and nerve plexuses surrounding the superior mesenteric artery (SMA). In 2010, Weitz et al[2] coined the term “artery-first approach”. This approach involves working from the base of the mesocolon to access the SMA from an inferior angle. By starting here, the surgeon gains better visibility and control around major blood vessels and lymph nodes, potentially lowering the risk of bleeding and allowing early identification of nearby structures like the right hepatic artery. However, this method can be more challenging for patients who are obese or have anatomical variations[3-5].

One of the known complications after pancreatic surgery is chyle leak (CL), which happens when lymphatic fluid leaks into the abdomen. The chances of CL can increase with more extensive resections, with rates of occurrence in pancreatic surgeries ranging from 0.6% to 16.3%[6]. The International Study Group of Pancreatic Surgery established a standard definition of CL in 2017[7]. According to this definition, CL is identified by milky fluid coming from a drain or wound on or after the third day after surgery, with a triglyceride level above 110 mg/dL. Grade A is categorized based on minimal clinical impact, requiring little or no intervention. Grade B and C require nutritional support, which might involve feeding tubes or total parenteral nutrition (TPN). Other treatments, like specific dietary fats, medium-chain triglyceride, and sometimes percutaneous procedures, are considered if needed. The most severe, grade C may involve intensive care and even risk mortality[7,8]. Understanding the causes and effects of CL can help guide treatment and improve patient recovery. This study aims to examine how often CL occurs after PD, its clinical impact, and the risk factors associated with it.

MATERIALS AND METHODS

This retrospective study included patients who underwent a specific pancreatic surgery called the “inferior infracolic SMA-first approach PD” to treat tumors in the front part of the pancreas and the uncinate process. The surgeries were performed at the Department of Gastrointestinal and General Surgery at Kathmandu Medical College and Teaching Hospital from January 1, 2021 to February 28, 2024. All procedures were carried out by the same team of surgeons, and ethical approval was obtained from the Institutional Review Board. Patients who had different types of PD surgeries were not included in this study. Before surgery, patients underwent routine tests, such as blood counts, kidney and liver function panels, and abdominal ultrasounds. A computed tomography scan with contrast helped assess the cancer’s stage and whether surgery could fully remove the tumor, that is R0, considering any local involvement of blood vessels.

During the procedure, the surgeons carefully assessed and dissected the SMA and surrounding structures, ensuring a safe and thorough approach to removing the tumor while protecting important blood vessels. The main steps included the mobilization of nearby structures, exposing the SMA, and identifying key vessels around the tumor. The meticulous approach helped improve visibility and access, allowing surgeons to work with greater precision. Key factors measured during surgery included operation time, blood loss, need for transfusions, extent of lymph node removal, and any vascular reconstruction needed. After surgery, complications were tracked and classified based on the Clavien-Dindo system, with CL severity graded according to international guidelines. The length of each patient’s hospital stay was recorded, and data were analyzed for patterns in outcomes. Continuous data were presented as averages with standard deviations, while categorical data were shown as percentages. The team used SPSS software for statistical analysis, and each patient provided informed consent before being part of the study.

RESULTS

During the study period of three years, a total of ten patients underwent dissection of the root of the mesentery with inferior infracolic SMA first approach PD. Four patients had a ventral body mass while six had an uncinate mass (Figure 1). Three patients in the uncinate group had undergone neoadjuvant chemotherapy. Four patients developed CL, one in the ventral group and three in the uncinate group. One patient had a grade A CL, while three had a grade B leak. The diagnosis of CL was made before the seventh post-operative day and the abdominal drain was placed for daily monitoring of the volume of CL. Four patients had the American Society of Anesthesiologists grade of either one or two, while the rest had a grade of three or four. The patient’s preoperative comorbidities were comparable. The hospital length of stay, intraoperative information, and patient demographics are mentioned in Tables 1 and 2.

Figure 1 Consort diagram for the patients enrolled in the study.

Table 1 Baseline characteristics with intraoperative findings.

Patient characteristics	Ventral (n = 4)	Uncinate (n = 6)
Age (year), mean ± SD	68.1 ± 8.2	67.6 ± 9.2
Gender
Male	2	2
Female	2	4
BMI (kg/m2), mean ± SD	24.2 ± 2.4	25.4 ± 2.1
ASA score
1-2	1	3
3-4	3	3
Neoadjuvant therapy	0	3
Preoperative resectability
Resectable	3	1
Borderline resectable	1	5
Vascular resection and reconstruction
No	3	1
Venous resection ISGPS type 1	1	3
Venous resection ISGPS type 2	0	0
Venous resection ISGPS type 3	0	2
Venous resection ISGPS type 4	0	0
Both arterial and venous resection	1	5
1st Jejunal branch sacrifice	0	5
Gastrocolic trunk of Henle involvement
Yes	3	1
No	1	5
Margin positive	1	1

ISGPS: International Study Group of Pancreatic Surgery; ASA: American Society of Anesthesiologists; BMI: Body mass index.

Table 2 Postoperative and pathological outcomes.

Postoperative and pathological outcomes		Ventral (n = 4)	Uncinate (n = 6)
Number of lymph nodes harvested		12	14
Clavien-Dindo (consequence of chyle leak)
2		1	1
3		0	2
4		0	0
5		0	0
Chyle leak		1	3
Grade A		0	1
Grade B		1	2
Grade C		0	0
Re-exploration		0	1 (reactionary hemorrhage)
Hospital stays	Total
1-10 days	4	2	2
10-20 days	4	1	3 [2 + (1 was reexplored for reactionary hemorrhage)]
> 20 days	2	1	1

The body mass index (BMI) of the uncinate group, in comparison to the ventral group, was higher in the patients developing CL (25.4 ± 2.1 and 24.2 ± 2.4). Six patients had borderline resectable pancreatic mass. The first jejunal branch was sacrificed in five patients. Six patients underwent vascular resection and reconstruction (Figure 2). Similarly, the gastrocolic trunk was involved in four patients. One patient with CL was simultaneously diagnosed with grade B delayed gastric emptying. Median age was similar in ventral body mass group and uncinate mass group (68.1 ± 8.2 and 67.6 ± 9.2) with male to female ratio of 4:5. With the virtue of CL, Clavien-Dindo grade II or higher morbidity was observed in four patients. One patient had to be reexplored due to a reactionary hemorrhage. Eight patients had a hospital stay of fewer than 20 days, however, two patients had a hospital stay of more than 20 days with the former with a grade B delayed gastric emptying and the latter with a long-term TPN requirement.

Figure 2 Patient characteristics, surgical details. A: First jejunal branch of superior mesenteric vessels reconstruction; B: Unsalvageable first jejunal branch; C: Root of mesentery dissection showing branches of superior mesenteric artery and vein. The circles and arrows represent the root of mesentery dissection.

As the patients were diagnosed with CL, they were initially deprived of any oral intake. All four patients were started on intravenous. Octreotide 100 μg thrice daily. In two cases, the CL was resolved within seven days. Total parental nutrition was started in one patient with a grade B CL. These patients were evaluated by the dietician of our hospital and resumed normal diet within three weeks. Neither patients diagnosed with CL had a recurrence nor required a re-exploration. The abdominal drains were removed when less than 50 mL of drain content was observed in 24 hours and there were no mortalities within a month of surgery. The mean operative time was 330 minutes. R0 was achieved in 100% of the patients. The mean number of intensive care unit and hospital stay were 2.55 ± 1.45 days and 15.7 ± 5.32 days respectively.

DISCUSSION

CL after PD are relatively rare, with studies reporting an incidence ranging from 1.0% to 16.3%[6]. However, estimating the true incidence is challenging due to differing definitions among researchers. The occurrence of CL has been associated with more aggressive resections. Additionally, patients with a higher BMI tend to develop CL, possibly due to the increased amount of fat tissue around the pancreas, which can expose the cisterna chyli and its branches during retroperitoneal dissection. In a study by Assumpcao et al[9], which included over 3500 pancreatic resections, similar findings were observed. The study noted that simultaneous vascular resections were a key factor contributing to CL. Many patients with CL exhibited perineural and perivascular involvement, which is especially common when operating on borderline resectable malignancies[5]. In our own study, we found a significantly higher number of lymph nodes dissected, including those affected by the tumor. This aligns with findings from Kuboki et al[10] and Abu Hilal et al[11], who both identified similar risk factors for CL in large cohorts of patients undergoing major hepatobiliary and pancreatic surgeries. Abu Hilal et al[11] also suggested that manipulation of the paraaortic and SMA regions increased the risk of developing CL, a conclusion supported by a series from Pan et al[8], which included 2000 pancreatic resections. Interestingly, the length of hospital stay did not appear to be significantly influenced by the presence of CL[8].

The management of CL can vary, with higher-grade leaks often requiring nasoenteral nutrition, dietary restrictions, and sometimes TPN, percutaneous drainage, or the use of medications such as octreotide[7]. Shadhu and Ramlagun[12] also suggested that medium-chain triglyceride diets or TPN could help decrease lymph flow, while a low-fat elemental formula was beneficial for resolving both CL and chylous ascites. Timing for the initiation of enteral feeding is critical, and our study followed a similar protocol for managing grade B CL patients, some of whom required extended TPN and a fat-free diet[7,12]. Although studies like that of Abu Hilal et al[11] have shown a significant impact of CL on hospital stays, with delayed discharges, in our study, only one patient stayed longer than 20 days after surgery. Singh et al[13] reported an average hospital stay of 16.5 days to 19.1 days for CL patients, though their study involved only 159 patients undergoing hepatobiliary and pancreatic surgeries. No patients in our study with CL required re-exploration, which aligns with findings from other studies. Abu Hilal et al[11] found that most patients with CL following pancreatic resections did not require invasive interventions or re-exploration, supporting a non-surgical approach to managing these leaks. Several studies have proposed a range of potential risk factors for CL, including BMI, age, lymphadenectomy, surgical margins, surgery duration, histopathology, pancreatic amylase testing, early enteral feeding, vascular resections, comorbidities such as diabetes, and concurrent postoperative pancreatic fistula[6,10,14]. Despite this, the definitive risk factors for CL remain unclear and require further investigation.

CONCLUSION

The root of mesentery dissection harvesting a greater number of lymph nodes and concomitant vascular resection positively correlates with the occurrence of CL. Patients who developed CL had significantly higher BMI and longer duration of operation. However, having accomplished an R0 resection, these leaks were successfully managed with TPN with no adverse impact on the outcome. The impact of the CL is uncertain with limited studies and need long-term results to validate it. It is imperative to focus on the different SMA first approaches to dissect the location of the tumor and simultaneous patient characteristics like BMI with logical intraoperative decisions regarding lymphadenectomy, vascular reconstruction, and operating time.

ACKNOWLEDGEMENTS

We extend our heartfelt gratitude to all the patients who participated in this study.