Pre-Existing Mesh at the Hiatus in Revisional Surgery Does Not Result in Increased Morbidity: A Case-Control Evaluation

Abstract

Background:

Mesh is sometimes used to reinforce the hiatus during primary and reoperative fundoplication. This is a controversial practice as it is not clear that this leads to a decreased rate of failure of the hiatal closure, and concerns about morbidity related to the presence of mesh in this location exist. One of these concerns is that if reoperation is ever required (fundoplication herniates through the hiatus, for example), revisional surgery would be significantly more difficult and associated with a higher rate of morbidity than if mesh had not been placed at the hiatus in a previous procedure.

Methods:

A retrospective review was conducted of prospectively collected data on 104 patients to undergo surgery for a failed fundoplication between 2011 and 2015. Fourteen patients (13.5%) had previous operations where mesh had been placed at the hiatus and underwent a subsequent revisional procedure. Procedures performed were reoperative fundoplication and Roux-en-Y gastric bypass as a salvage procedure for a failed fundoplication, especially in the setting of obesity. These 14 cases were matched 1:2 with randomly selected control patients from the database who underwent revisional surgery in whom mesh had not been placed at the original operation. Cases and controls were paired based on the number of previous revision attempts and operation type. Perioperative outcomes were compared.

Results:

There was no statistically significant difference in 30-day morbidity, readmission, operative time, or length of hospital stay.

Conclusions:

In this retrospective case–control evaluation, mesh at the hiatus did have an impact on morbidity or operative time.

Introduction

Laparoscopic revisional surgery for failed fundoplication has a significantly increased rate of morbidity compared to primary surgery. Indications for reoperation include recurrent gastroesophageal reflux disease (GERD) and dysphagia.¹ Intraoperative complications have been reported to occur in up to 20% of cases, which most commonly includes esophageal or gastric perforation. Postoperative complications occur in ∼15% of cases and include infectious, pulmonary, and cardiac morbidities.²

Many surgeons will place mesh at the hiatus at the time of hiatal and or paraesophageal hernia repair with fundoplication depending on the size of the crural defect and integrity of the repair, although the effectiveness of this practice in reducing hiatal failures has been debated. There is also controversy surrounding the safety of placing hiatal mesh, especially if another operation proves to be necessary at the hiatus in the future. Studies have identified that patients with prior mesh at the hiatus have an increased risk of major resection during revisional operations.^3,4 We sought to determine the impact of pre-existing mesh at the hiatus on the morbidity of revisional surgery for failed fundoplication.

Materials and Methods

A retrospective review was conducted of a prospectively collected database of 104 patients to undergo surgery for a failed fundoplication at a single institution between the years 2011 and 2015. All patients included in the study had their original procedure(s) at a referring institution and the most recent revisional case with the senior author. Operative indications for the original fundoplication procedures were primarily GERD. A single surgeon performed all revisional procedures in this study.

Patients who had pre-existing mesh at the hiatus in place at the time of revisional surgery from a previous operation were identified. Indications and rationale for the use of mesh in the prior case could not be reliably determined from operative reports, as many of these cases had taken place years ago. Cases with pre-existing hiatal mesh were matched 1:2 with randomly selected control patients from the same database. A 1:2 distribution of cases and controls was selected because of the small number of patients in the series who had mesh in place at time of a revisional operation in an attempt to increase the sample size. Controls were selected based on the number of prior fundoplication procedures and the type of operation performed at the time of revision.

Operation types included laparoscopic takedown of a previous fundoplication and conversion to gastric bypass (Roux-en-Y Gastric Bypass [RYGB] study group) and laparoscopic takedown of fundoplication and reoperative fundoplication (reoperative fundoplication study group). Cases and controls were randomly paired based on the number of previous revision attempts and operation type using Stat Trek 2015. For example, a patient who had undergone a fundoplication without mesh at the hiatus, a reoperative fundoplication with mesh at the hiatus, and finally a takedown of a failed fundoplication and conversion to RYGB was paired with two control patients who had each undergone a fundoplication without mesh at the hiatus, a reoperative fundoplication without mesh, and a takedown of a failed fundoplication and conversion to RYGB. There were not enough patients with two prior failed attempts at fundoplication (neither attempt with mesh at the hiatus) who underwent a third time reoperative fundoplication to arrive at a full 1:2 comparison in this subset of the control group (one control patient short).

Perioperative outcomes included intraoperative perforation, readmission, 30-day morbidity, operative time, and length of stay. Reoperative fundoplication was the procedure of choice in most patients with a single prior failed fundoplication attempt. In patients with a body mass index (BMI) >35 kg/m², takedown of the fundoplication and conversion to RYGB was the procedure of choice. All cases and control patients in this series underwent the most recent operative revision case laparoscopically without conversion.

Reoperative fundoplication technique

Reoperative fundoplication involved taking down the previous fundoplication completely in all patients. Abdominal access for laparoscopy was attained with a Veress needle in the left subcostal area in the midclavicular line. Upon successful insufflation, a 5-mm optical viewing bladeless trocar was used to access the abdomen through the Veress site. A total of four 5-mm ports were used in the reoperative fundoplication cases. If mesh was removed or placed at a revision, the left subcostal port was upsized to 10 mm. A subxiphoid Nathanson liver retractor was placed in all patients.

Adhesiolysis involved a combination of sharp dissection with a scissors, dissection with an ultrasonic shear, and blunt dissection with a laparoscopic suction irrigator tip where appropriate. When technically feasible, the caudate lobe of the liver was identified first, which allowed for identification of the right crus of the diaphragm. Dissection typically started at the base of the right crus and proceeded anteriorly in the plane between the right crus and the fundus/esophagus if possible. Ultimately, the fundoplication or esophagus was circumferentially mobilized off the hiatal muscle—preserving as much muscle and fascia as possible on the diaphragm. In some cases with severe and dense adhesions, the anatomic landmark that proved easiest to find was the base of the left crus and this was after dividing branches or the gastroepiploic vessels to the body of the stomach and entering the lesser sac posterior to the stomach and somewhat removed from the operative site and hiatal adhesions. In cases with pre-existing mesh at the hiatus, the mesh was removed in its entirety when possible. Care was taken to preserve as much diaphragm muscle as possible when removing mesh. Care was also taken to ensure that mesh adherent to the fundus or esophagus could be safely removed without damage or perforation to the foregut. Mesh excision from the esophagus or fundus was almost always done sharply with scissors. Sometimes, more than one fresh scissors tip was needed in these cases. A sharp scissors tip is essential in these cases.

Once the fundus or esophagus was freed from the hiatus, the fundoplication was taken down. Usually, the fundoplication sutures could be visualized and a combination of dissection with an ultrasonic energy source or with scissors was utilized to dismantle the fundoplication. Ultrasonic shears were only used when the sutures and tissue planes could be clearly visualized. An advantage to using the shears in this context is better hemostasis but also a “cavitation” effect that can facilitate the dissection and help to identify tissue planes. A disadvantage to the ultrasonic shears is the fact that tissue injury and perhaps delayed perforation is a potential consequence. We did not observe this event in our experience. Fundus to fundus, fundus to esophagus, and ultimately fundus to retroperitoneum/diaphragm adhesions are carefully taken down until the greater curve of the stomach and angle of His is restored to its native anatomic position.

Circumferential and high mediastinal mobilization of the esophagus was performed until 2–4 cm of intra-abdominal esophageal length could be attained without tension. A Penrose drain around the esophagus was used for gentle retraction. The vagus nerves were identified and preserved when possible. Esophageal lengthening procedures were not necessary in any patient.

Primary crural repairs were performed in all patients. In patients with a dilated hiatus or loss of muscle at the hiatus from the dissection, a right-sided crural relaxing incision was performed to allow for primary closure. To perform a relaxing incision, a hook cautery or ultrasonic shears were used to incise the right crus of the diaphragm starting about 2–3 mm medial to the vena cava when possible. A full thickness muscle incision was made and if possible the right pleural cavity was not entered. In most cases good medialization of the right crura was possible without opening the pleura. In cases where the pleura was entered, the pleura on the right side was opened through the hiatus, and a 19F Silastic drain was placed through the hiatus anteriorly, into the mediastinum, and into the right chest. At the end of the case, this drain was externalized through a 5-mm port. The anesthesiologist was then asked to give several big Valsalva breaths with the ventilator as the drain is slowly removed under suction. Chest tubes or abdominal drains were not routinely left in place in these cases.

In patients undergoing reoperative fundoplication in whom the hiatal repair was determined to be under tension or for whom a crural relaxing incision was needed, a U-shaped piece of synthetic bioabsorbable mesh (GORE BIO-A, Flagstaff, AZ) was used to reinforce the hiatal repair and be sutured into place. In patients with pre-existing dysphagia or impaired esophageal motility (based on manometry), a posterior partial fundoplication (Toupet) was constructed. In patients to undergo reconstruction of a Nissen fundoplication, a 56F to 60F esophageal bougie was utilized. Gastropexy sutures among the fundoplication, the posterior hiatal repair, and the anterior left and right crural pillars were routinely placed.

Endoscopy is performed liberally during the procedure to help identify the anatomy and to ensure that there are no unrecognized perforations. Endoscopy is performed at the end of the procedure to ensure the same and that the fundoplication is in the right location relative to the gastroesophageal junction and that the geometry of the wrap on retroflexion is appropriate.

Conversion to Roux-en Y Gastric Bypass

In patients undergoing conversion to RYGB, much of the initial procedure is as described above for the reoperative fundoplication. There are a total of five ports, with two of them being 10 mm to accommodate the linear cutting stapler. The previous fundoplication was always taken down and the mesh removed as described. The gastric pouch was created from the proximal stomach and was ∼30 mL. The lesser curve neurovascular bundle was preserved. The left gastric artery was always identified and the pouch was created below this artery to ensure an adequate blood supply. In select cases with multiple previous fundoplication attempts undergoing conversion to RYGB, indocyanine green was administered intravenously to help identify the left gastric artery and branches and to ensure that the proposed pouch had adequate perfusion before firing the laparoscopic stapler. A laparoscopic camera system with near infrared imaging capabilities (Stryker 1588 Advanced Imaging Modalities Platform; Stryker, Kalamazoo, MI) is used for this purpose. Resection of a portion of proximal fundus was undertaken if intraoperative perforation occurred or if the blood supply to this portion of the fundus was questionable following creation of the proximal gastric pouch.

New mesh at the hiatus was avoided in these reoperative cases undergoing conversion to RYGB. There was also significantly less clinical concern for tension at the hiatus leading to hiatal failure given the nature of the RYGB procedure. A 50-cm biliopancreatic limb, antecolic and antegastric 100-cm Roux limb (150 cm for those with a BMI >50), and a linear stapled gastrojejunostomy were constructed. The gastrojejunostomy was completed with a 32F bougie in place, and the entire stapled anastomosis is oversewn with absorbable Vicryl sutures. The gastric pouch was sutured to the hiatal closure posteriorly and to the anterior left and right crural pillar as described for the reoperative fundoplication procedures. As in the reoperative fundoplication procedures, endoscopy is performed liberally and often to help guide the creation of the pouch, to identify the anatomy, and to ensure the gastrojejunostomy is hemostatic, without leak, and easy to traverse with an endoscope.

Baseline symptom severity was assessed before surgery using the Gastroesophageal Reflux Disease Health Related Quality of Life Index (GERD-HRQL). The GERD-HRQL is a validated 10-item survey, with a composite score range of 0 to 50. The best possible score is 0, indicating no/resolved symptoms, while the worst possible score is 50, indicating incapacitating symptoms.⁵ Our institutional review board approved this study. Statistical analysis was performed using Fisher's exact tests for categorical and unpaired t-tests for continuous variables. Statistical significance was defined as P < .05. All calculations were performed using GraphPad Software, Inc., 2015.

Results

Among the 104 patients within the database, 14 (13.5%) were identified that had undergone a previous fundoplication attempt where mesh was placed at the hiatus in a procedure before the most recent revision. GERD-HRQL scores were available before surgery in 9 of 14 (64%) in the prior mesh cohort and 10/18 (56%, P = NS) in the no mesh control group. Preoperative GERD-HRQL scores did not differ between the study groups (24.7 ± 10.4 in the prior mesh cohort and 27.7 ± 15 in the control group, P = .6). Preoperative dysphagia was rated as a three or higher (symptoms of difficulty swallowing bothersome every day) on a 6-point Likert scale in 5/9 prior mesh (56%) and 10/18 (56%) control group patients (P = NS).

The indication for surgery at the time of the first operation in patients who had mesh in place at the time of revision was medically refractory GERD in 13 of 14 patients. In one patient, the primary indication for surgery was symptoms other than GERD related to a type III paraesophageal hernia. In the control group of patients, indications for the index operation included symptoms other than GERD related to a type III paraesophageal hernia in three patients and GERD in the remaining 24. Procedures performed in the previous mesh cohort were reoperative Nissen fundoplication (n = 7), conversion of Nissen to Toupet (n = 3), and conversion to RYGB (n = 4).

Mesh was implanted at the time of the current reoperative procedure in nine patients who previously had mesh placed (64%) and in seven (26%, P = .02) patients in the no previous mesh group at the time of revisional surgery (7 × 10 cm U-shaped absorbable synthetic mesh in all patients). All nine patients who had previous mesh and underwent implantation of another piece of mesh at the time of the most recent revision had the original piece of mesh removed at the time of revision. Demographics and types of mesh previously placed in the study group of 14 patients are listed in Table 1. Among patients with prior mesh, there were five patients with biologic and nine with synthetic mesh. Outcomes and operative time did not differ for synthetic versus biologic mesh.

Table 1.

Demographics and Hiatal Mesh Type in Previous Mesh Cohort

Pt	Age (years)	Sex	Body mass index (kg/m ² )	No. of prior foregut surgeries	Type (s) of prior failed foregut surgery	Type of prior mesh	Type of laparoscopic revisional surgery performed
1	36	F	40.5	1	Lap Nissen fundoplication	Biologic (Strattice)	RYGB
2	59	M	36.6	1	Open Nissen fundoplication	Synthetic (Parietex)	RYGB
3	56	F	33.8	2	Lap Nissen × 2	Biologic (XenMatrix)	RYGB
4	46	F	29	3	Lap Nissen × 3	Synthetic (Parietex)	RYGB
5	60	F	28.1	1	Lap Nissen	Biologic (AlloDerm)	Nissen fundoplication
6	42	M	23.6	1	Robotic Nissen	Synthetic (Bio-A)	Nissen fundoplication
7	61	F	36.6	1	Robotic Nissen	Synthetic (Bio-A)	Nissen fundoplication
8	58	M	25.7	2	Lap Nissen × 2	Biologic (Unknown)	Nissen fundoplication
9	46	F	20.3	2	Lap Nissen × 2	Synthetic (CruraSoft)	Nissen fundoplication
10	69	F	36.2	2	Lap Toupet × 2	Synthetic (Proceed)	Nissen fundoplication
11	67	M	31	2	Lap Nissen × 2	Synthetic (CruraSoft)	Nissen fundoplication
12	77	F	32.6	1	Lap Dor	Synthetic (Bio-A)	Toupet fundoplication
13	61	F	18.1	1	Lap Nissen	Synthetic (Bio-A)	Toupet fundoplication
14	64	F	32.3	2	Robotic Nissen × 2	Biologic (Strattice)	Toupet fundoplication

RYGB, Roux-en-Y Gastric Bypass.

There were no statistically significant differences in perioperative morbidity in patients with or without prior hiatal mesh in revisional surgery for failed fundoplication (Table 2). The morbidities that occurred in patients with prior mesh included one reoperation for a diagnostic laparoscopy, pneumonia, and readmissions for dehydration. The morbidities for those without prior mesh included two patients with pneumonia, one pulmonary embolism, and one empyema. There were no mortalities. There was also no statistically significant difference in 30-day readmissions, operative time, or length of stay. There was one patient (7.1%) with prior mesh who had an intraoperative perforation, defined as a full thickness injury, and no patients without prior mesh (P = .34). This perforation was recognized immediately in the operating room. This patient had two prior fundoplication procedures and was undergoing conversion to RYGB. The site of the perforation was the lateral fundus, and this portion of the stomach was safely resected and removed following creation of the gastric pouch. This patient did not experience any adverse postoperative issues related to this intraoperative event.

Table 2.

Perioperative Outcomes in Patients with and Without Prior Mesh at the Hiatus

	Prior mesh (n = 14)	Without mesh (n = 27)	P
30-day readmission	4 (28.6%)	7 (25.9%)	1.00
30-day morbidity	4 (28.6%)	10 (37%)	.73
Intraoperative perforation	1 (7.1%)	0	.34
Mesh placed at revisional operation	9 (64.3%)	7 (25.9%)	.02
Operative time	227.5 minutes	247.7 minutes	.59
Length of stay	2.92 days	4.96 days	0.23

Discussion

Our study evaluated the difference in perioperative morbidity in patients undergoing revisional surgery for failed fundoplication with and without the presence of mesh at the hiatus at the time of revisional surgery. We found no statistically significant difference among 30-day readmission, morbidity, intraoperative perforation, operative time, or length of stay between these two groups.

We did not operate on these patients during their index operation(s), which often took place many years before the most recent revisional procedure evaluated in the current study. For these reasons, we cannot comment objectively on the rationale regarding the use of mesh in these procedures that ultimately failed and required a revision at our center. After hiatal hernia repair or fundoplication for GERD, one of the major outcomes that can ultimately lead to reoperation is recurrence of the hernia or recurrent GERD (also most commonly secondary to herniation of the fundoplication through the hiatus). Recurrence after hiatal hernia repair occurs in 4%–42% of patients.^6,7 The goal of using mesh at the hiatus is to decrease the rate of subsequent hiatal hernia. The use of hiatal mesh for patients undergoing hiatal hernia repair and fundoplication is a controversial practice, with varying opinions among surgeons. A survey of the Society of American Gastrointestinal and Endoscopic Surgery (SAGES) members demonstrated that a majority of surgeons have used mesh for hiatal hernia repair, although only a minority uses it routinely.⁸ Another study by Obeid and Velanovich⁹ describes a decision analysis model of choosing between primary and mesh-based repair in hiatal hernia and demonstrates that the difference with regard to recurrence or reoperation is minimal. Oeschlager et al. conducted a multicenter randomized-controlled trial of patients undergoing primary versus biologic mesh repair during laparoscopic paraesophageal hernia repair. After a median follow-up of 58 months, radiologic hernia recurrence was reported in 59% of patients with primary repair and 54% with mesh repair (P = .70).¹⁰

Many surgeons have reported concerns about the use of mesh at the hiatus, including the potential for erosion, stricture, and dysphagia.¹¹ A study by Stadlhuber et al¹² identified 26 cases with hiatal mesh-related complications, including mesh erosion, stenosis, and dense fibrosis. Mesh erosion was more common among those with synthetic mesh, but esophageal stenosis was more common with biologic mesh. There are limited studies specifically focusing on perioperative morbidity in patients with prior hiatal mesh, especially compared to those without mesh, as we do in this study. Parker et al. conducted a retrospective review of 78 patients who underwent revisional foregut surgery, 10 of which had prior hiatal mesh and 68 with no mesh.³ They found no difference in length of stay. They did observe that patients with prior hiatal mesh had a 6.8-fold increased risk of major resection (esophagogastrectomy) compared to those without mesh (P = .05). Nandipati et al⁴ described a case series of patients undergoing revisional foregut surgery with prior hiatal mesh. In this study, which consisted of 26 patients, just over half were completed laparoscopically. Eight of their patients required major esophageal resection and their perioperative morbidity rate was 22%. In addition, they performed a subset analysis of patients with synthetic or biologic mesh and found no difference in outcomes. Even when mesh was not used at the hiatus in a previous operation, revisional hiatal hernia repair and reoperative antireflux surgery is a complex intervention associated with an increased risk of morbidity. Furnee et al² conducted a systematic review of 81 studies evaluating the outcomes following reoperative antireflux surgery, including 4584 procedures. They identified that intraoperative complications occurred at a rate of 21.4%, with the most common being laceration or perforation of the esophagus and/or stomach.

A limitation of the current study is its small size and retrospective nature. We also did not follow patients' long term to evaluate for recurrence. The point of this review is not to make an assessment of the impact of mesh on long-term recurrence however. Our primary objective is to determine if mesh that has already been placed at the hiatus in a patient who ultimately requires revision impacts perioperative morbidity. Because of these objectives, we feel that 30-day follow-up is sufficient. The small number of patients with previous mesh precludes us from making conclusions about the differences in morbidity following revisional surgery in patients with previous synthetic or biologic mesh. The case–control nature of the current study overcomes several of the limitations of previous studies that examined the impact of mesh at the hiatus on morbidity.

In conclusion, our study suggests that in the hands of experienced surgeons, mesh at the hiatus does not seem to have a major impact on morbidity or operative time in revisional surgery for failed fundoplication.

Footnotes

Disclosure Statement

No competing financial interests exist. Jon Gould is a consultant for Torax Medical.

References

Symons

, Purkayastha

, Dillemans

, Athanasiou

, Hanna

, Darzi

, Zacharakis

. Laparoscopic revision of failed antireflux surgery: A systematic review. Am J Surg, 2011; 202:336–343.

Furnee

, Draaisma

, Broeders

, Gooszen

. Surgical reintervention after failed antireflux surgery: A systematic review of the literature. J Gastrointest Surg, 2009; 13:1539–1549.

Parker

, Bowers

, Bray

, Harris

, Belli

, Pfluke

, Preissler

, Asbun

, Smith

. Hiatal mesh is associated with major resection at revisional operation. Surg Endosc, 2010; 24:3095–3101.

Nandipati

, Bye

, Yamamoto

, Pallati

, Lee

, Mittal

. Reoperative intervention in patients with mesh at the hiatus is associated with high incidence of esophageal resection–a single-center experience. J Gastrointest Surg, 2013; 17:2039–2044.

Velanovich

. The development of the GERD-HRQL symptom severity instrument. Dis Esophagus, 2007; 20:130–134.

Mattar

, Bowers

, Galloway

, Huter

, Smith

. Long-term outcome of laparoscopic repair of paraesophageal hernia. Surg Endosc, 2002; 16:745–749.

Hashemi

, Peters

, DeMeester

, Huprich

, Quek

, Hagen

, Crookes

, Theisen

, DeMeester

, Sillin

, Bremner

. Laparoscopic repair of large type III hiatal hernia: Objective follow-up reveals high recurrence rate. J Am Coll Surg, 2000; 190:553–560.

Pfluke

, Parker

, Bowers

, Asbun

, Smith

. Use of mesh for hiatal hernia repair: A survey of SAGES members. Surg Endosc, 2012; 26:1843–1848.

Obeid

, Velanovich

. The choice of primary repair or mesh repair for paraesophageal hernia: A decision analysis based on utility scores. Ann Surg, 2013; 257:655–664.

10.

Oelschlager

, Peterson

, Brunt

, Soper

, Sheppard

, Mitsumori

, Rohrmann

, Swanstrom

, Pellegrini

. Laparoscopic paraesophageal hernia repair: Defining long-term clinical and anatomic outcomes. J Gastrointest Surg, 2012; 16:453–459.

11.

Granderath

, Schweiger

, Kamolz

, Asche

, Pointer

. Laparoscopic Nissen fundoplication with prosthetic hiatal closure reduces postoperative intrathoracic wrap herniation. Arch Surg, 2005; 140:40–48.

12.

Stadlhuber

, Sherif

, Mittal

, Fitzbibbons

, Brunt

, Hunter

, DeMeester

, Swanstrom

, Smith

, Filipi

. Mesh complications after prosthetic reinforcement of hiatal closure: A 28-case series. Surg Endosc, 2009; 23:1219–1226.