Feeding jejunostomy (FJ) is critical for enteral nutrition after major upper gastrointestinal surgeries. In certain instances, FJ is performed in isolation to establish a source of enteral nutrition; hence, it is essential to create a well-formed FJ. Among the various methods of creating FJ, Witzel technique is the most widely practiced [1]. Saliently, this technique consists of securing the feeding tube by creating a serosal tunnel made of a bowel wall and then internally fixing it to the abdominal wall. This leads to decreased leak rates and complications [2,3].

Furthermore, laparoscopic FJ (LFJ) carries the advantages of a minimally invasive approach, like less pain, fewer postoperative morbidity, and shorter overall hospital stay. In contrast to open FJ, laparoscopic Witzel technique for FJ has limited reports in the literature and has not been discussed widely as a standalone procedure. Furthermore, LFJ by Stamm technique is also preferred at many centers but is associated with many complications [1,4]. Hence, there is a definite need to discuss the technical aspect of LFJ and its outcomes in detail.

We hereby discuss the technique of LFJ with Witzel technique in detail, along with its benefits. We have also compared the results of LFJ with the open approach to further consolidate the feasibility and outcomes of LFJ.

This retrospective comparative study was performed in the Department of Surgical Gastroenterology at our institute, from the database maintained from July 2018 to July 2022. All patients who underwent laparoscopic or open FJ by Witzel technique as a standalone procedure were included. Dysphagia due to esophageal carcinoma and acute corrosive injury or delayed corrosive stricture in the esophagus were the primary indications for the FJ. Informed and written consent was obtained from all patients. The details regarding patient demography, indication for FJ, clinical parameters, operative parameters, and postoperative complications of all included patients were retrieved from medical records.

Preoperative preparation

The patients were evaluated by detailed past medical history and physical examination. The relevant laboratory and imaging investigations were performed. Preoperative anesthetic check-ups and fitness for general anesthesia were taken. Marking of the future FJ site in the supine, sitting, and standing positions was done. The need for the procedure and the advantages of laparoscopy were explained to the patients. All patients in whom FJ was indicated were informed about both laparoscopic and open approaches for FJ. The selection of the procedure was based on the patient’s choice, but in instances where tolerability of pneumoperitoneum was in doubt, such as severe pulmonary disease, the open method was preferred.

Operative steps of laparoscopic feeding jejunostomy

Positioning of patient and ports:

The patient was placed in a reverse Trendelenburg position with the left side upwards and their arms tucked by their side. Pneumoperitoneum was created using a Veress needle via a 10-mm infraumbilical port. After that, four ports were placed for vision: one 10-mm camera port in the right anterior axillary line, placed between the right subcostal and anterior superior iliac spine; and two 5-mm working ports in the right mid-clavicular line at the right subcostal and iliac fossa around 7–10 cm between the ports (Fig. 1). One 5-mm port was inserted at the future FJ site on the left side of the abdomen. The operating surgeon stands on the patient’s right side with the camera assistant on the surgeon’s left side; one assistant surgeon stands on the patient’s left side and the laparoscopic unit stays at the patient’s left shoulder (Fig. 1).

Identification of the jejunal loop and anchoring to the parietal wall

After retracting the greater omentum and transverse colon superiorly, the duodenojejunal (DJ) flexure and ligament of Treitz were identified. A loop of jejunum, 20 to 30 cm distal to DJ flexure, was chosen and confirmed to ‘reach’ the anterior abdominal wall without tension. This jejunal loop was held near the left-sided 5-mm port (future jejunostomy site) using a grasper and sutured to the abdominal wall using a 3-0 polyglactin 910 suture (‘anchoring suture’). This anchoring orients the antimesenteric border of the jejunum face toward the right side, thus facilitating the whole procedure.

Insertion of the tube and purse-string suture

An enterotomy was made 6 to 8 cm distal to the anchoring site (the future tube insertion site) into the jejunum via a harmonic scalpel (Ethicon). Then, a 14-French (Fr) Ryles tube (RT) for feeding was inserted through the left-sided 5-mm port, after which it was removed. The lubricated tip of the RT was gently inserted into the enterotomy site until all of the side holes of the RT were inside the bowel. The assistant then slowly pushed normal saline through the RT to aid in the advancement of its tip for a length of around 30 cm. The position of the RT tip and its smooth lie inside the bowel, along with bowel wall laxity, was ensured. The insertion site of RT in the jejunum (RT-J) was secured by placing a purse-string suture using 3-0 polyglactin 910. The anchoring suture and the purse-string suture kept the jejunum loop and RT steady along the antimesenteric border (Fig. 2).

Tunneling

After this, interrupted Lembert sutures were placed using 3-0 polyglactin 910 to form a serosal tunnel around RT. For this, a first forehand bite was taken on the left of the RT, the needle passed in front of RT and then a similar bite was taken on the right of RT at the same level. The suture was then tied, so the RT was fully covered, thus creating a tunnel over it. Multiple interrupted sutures were placed in the same manner at 5 to 7-mm intervals between the RT-J site and the anchoring site (Fig. 2). These sutures complete the exclusion of RT from the peritoneal cavity.

Fixation to the anterior abdominal wall

The bowel was then secured to the abdominal wall at two places to avoid its rotation/volvulus (‘three-point fixation’). The RT-J site was sutured first using a seromuscular bite. During this step, the assistant applied external pressure over the abdominal wall to bring it close to the jejunum before securing the knot. This ‘maneuver’ helped a loose knot to be avoided as lifting the jejunal loop toward the abdominal wall is difficult and technically challenging. Furthermore, a similar suture was taken either proximal to the anchor suture or distal to the RT-J suture to finish the three-point fixation (Fig. 3). It is crucial to ensure that there is no kinking of the bowel loop among these fixation points.

The patency of FJ was ensured again by the free flow of the saline via RT. The FJ was fixed externally at the skin and secured in a circular fashion to avoid any pull on the tube from transmitting a skin fixation suture to the abdominal entry site.

Operative steps of open feeding jejunostomy

The open Witzel FJ technique is similar to the laparoscopic approach. Briefly, the abdomen is opened by a supraumbilical vertical midline incision. DJ flexure identified and a suitable site 20 to 30 cm distal to it is identified. A small enterotomy is made, 14-Fr RT inserted up to 30 cm and secured with a purse-string suture with absorbable polyglactin 3-0. The RT is buried in a serosal tunnel for 5 cm using interrupted 3-0 polyglactin sutures. The RT was brought out through the abdominal wall, then the exit site was fixed internally (jejunum to the abdominal wall) with three to four polyglactin 3-0 sutures and lastly, the three-point fixation was done. The RT is checked for patency using normal saline, after which it is fixed externally to the skin at the exit site as well as in a circular fashion.

Postoperative management

All patients were started on the first postoperative day (POD) with normal saline at the rate of 50 mL/hr, followed by half-strength and full-strength enteral feed on POD 2 and POD 3, respectively. Most patients tolerated 1.5 to 2 L of full-strength FJ feed by POD 4.

A total of 41 patients underwent FJ by Witzel technique in our department during the study period; 20 patients by the laparoscopic method and 21 by the open method. All patients tolerated the procedure well. The baseline characteristics of the patients were comparable between laparoscopic and open FJ groups (Table 1). Median preoperative hemoglobin (11.4 g/dL vs. 9.2 g/dL) and albumin (3.3 g/dL vs. 3.1 g/dL) were also lower in the open FJ group, albeit statistically insignificant. One patient with previous abdominal surgery and one who underwent the procedure under local anesthesia were in the open FJ group. Malignancy was the most common indication (n = 26) for the procedure in both groups, including carcinoma of the esophagus (n = 19), postcricoid (n = 4), and stomach (n = 3). Among the benign causes (n = 15) were acute corrosive injury of the esophagus (n = 7), corrosive stricture of the esophagus (n = 6) and acute pancreatitis (n = 2).

The median operative duration was more in the LFJ group (180 minutes vs. 60 minutes, p = 0.01). The median postoperative length of hospital stay was shorter in the LFJ group (3 days vs. 4 days, p = 0.08). Four patients suffered from immediate postoperative complications in the open group: two patients had a surgical-site infection (SSI), one patient required reoperation for intussusception, and the remaining patient had percutaneous drainage of pelvic collection. There was no 30-day mortality in either group. Fewer patients in the purely laparoscopic group suffered complications such as tube dislodgement (4 vs. 9), tube clogging (13 vs. 15), and SSI (0 vs. 1) compared to the open group within 30 days of discharge. At a median follow-up of 10 months (range, 4–32 months), one patient in the laparoscopic group and two in the open group presented with a small bowel obstruction.

Establishing a source of enteral feeding at the end of the surgical procedure is an essential part of complex gastrointestinal and pancreatic surgeries. Apart from this, there are conditions where FJ is required as a palliative measure or in the preoperative setting to optimize the patient’s general condition. Usually, an open approach is used to create an FJ in such scenarios, which is associated with its inherent disadvantages. Among the various techniques, Witzel technique is used for the creation of FJ. However, despite its advantages, using the minimally invasive approach is not widely popular compared to Stamm method.

With an increased experience in laparoscopic surgery, surgeons have established benchmark outcomes in terms of postoperative pain, less wound infection, the early resolution of an enteral diet, and early discharge without compromising the radicality or quality of surgery. Hence, LFJ should be considered as an alternative to open FJ. Although there are many case reports and series on LFJ in the literature, most either deal with hybrid techniques, the use of costly catheters, or direct fixation of the jejunum to the abdominal wall without creating a tunnel [5–10]. To the best of our knowledge, only one series has evaluated the outcome of the Witzel technique [6]. Although their results show a considerable morbidity of 12.4% and a mortality of 0.3%, it was associated with procedural heterogeneity (both purse-string fixation and Witzel tunnel) and morbidity of the accompanying major procedures such as esophagectomy.

In our series, LFJ was created with Witzel technique as a standalone procedure and by using standardized steps. Our technique of LFJ is safe, has several advantages, and can be considered an alternative to its open counterpart. Firstly, the median hospital stay was less in the laparoscopic group (3 days vs. 4 days), which can be attributed to less pain and early ambulation leading to early bowel activity. Secondly, no patients in the laparoscopic group encountered immediate postoperative complications such as SSI, intussusception, and respiratory infection. Furthermore, fewer patients suffered from early (<30 days) and late complications after discharge from the hospital. Finally, an easily available nasogastric tube (RT), routine laparoscopic instruments and sutures were used compared to the open counterpart. The utilization of braided multifilament absorbable sutures like polyglactin to create a purse string, tunneling and three-point fixation to the anterior abdominal wall is useful during laparoscopic suturing. Studies have shown that barbed suture used for fixation may cause small bowel obstruction and ileus [11–13]; thus, absorbable polyglactin sutures is a good alternative to avoid such complications.

Catheter site infection is a major morbidity in LFJ, ranging from 0% to 12.5% in the literature [1,5,6,14]. Young et al. [2] reported severe SSI that prompted them to alter their technique to create a Witzel tunnel in the later part of their surgical period, after which they did not encounter any such complication. Care must be taken to avoid excess suturing of the jejunum to avoid luminal narrowing. Further, the cost of laparoscopic procedures has always been debated over their open counterparts and it is usually perceived that minimally invasive procedures are more costly. Although the cost has not been studied separately in this study, we would like to highlight that no expensive introducer kits or jejunostomy catheters were used in our technique. Instead, we recommend easily available and cost-effective RT to be used as a jejunostomy catheter and polyglactin for suturing and tunneling in place of the barbed suture, which is more costly. Also, all the existing instruments were used in the procedure, which was routinely used in other laparoscopic procedures such as cholecystectomy and colectomy.

Recently, percutaneous endoscopic gastrostomy has been propagated as one of the superior routes of enteral nutrition due to its less invasive nature and lesser need for general anesthesia [15,16]. Nevertheless, it is also associated with complications of bleeding, perforation, peritubular leak, necrotizing fasciitis, and major complications like peritonitis [16]. Moreover, it also requires costly catheter tubes and fixation devices and the availability of an expert endoscopist. On the other hand, LFJ is useful in cases where the endoscopic route is not available or amenable. It has several advantages in terms of visualizing the peritoneal cavity for the extent of disease, in cases of malignancy, or assessing the status of the stomach, for the future conduit creation in corrosive injuries, and lastly, fewer adhesions for future procedures to be performed for definitive management.

Our study has limitations due to its retrospective nature, selection bias, and small sample size. Although the differences in the two groups were not statistically significant, there was a tendency to select more fit patients for LFJ. However, it was predominantly the patient’s choice for the selection of the procedure. Further, a longer operating time was required for LFJ. It is to highlight that LFJ is usually performed in a semi-emergency setting and requires advanced laparoscopic suturing skills. As mentioned, the median operative duration was 180 minutes, and we observed a shorter operative time during the late part of the study. Hence, we expect this median operative duration to decrease with increasing experience over time. Patients undergoing laparoscopic and open FJ as part of procedures such as esophagectomy, gastrectomy, pancreaticoduodenectomy, etc., were excluded, as the intent is to highlight the outcome of FJ as a standalone procedure which would otherwise be overshadowed by the inherent complications of the associated major procedures.

In conclusion, this study demonstrates the safety and feasibility of LFJ by the Witzel approach with comparable postoperative outcomes. LFJ may be considered in select patients with good performance status requiring enteral feeding. Future large-scale prospective and randomized studies are warranted to validate our findings as well as cost-effectiveness in terms of equipment benefits and other outcome variables.

Ethical statements

The ethical clearance for this study was obtained from the Institutional Ethics Committee of All India Institute of Medical Sciences (No. AIIMS/IEC/2023/4258) and a waiver of informed consent was taken.

Author’s contributions

Conceptualization: PV

Investigation, Data curation: All authors

Formal analysis: PV, VN, VKV, SS, SB

Methodology: PV, VN, VKV

Writing–original draft: PV, VN, VKV

Writing–review & editing: VKV, SS, SB, LA, AS

All authors read and approved the final manuscript.

Conflict of interest

All authors have no conflicts of interest to declare.

Funding/support

None.

The patients’ characteristics