The laparoscopic fundoplication has emerged as the gold standard for the treatment of refractory reflux disease and symptoms of gastric herniation. Surgical success rates as measured by patient’s satisfaction and symptom relief have been reported between 85 and 95%. [1,2] However, as with any operation we should continue to critically examine clinical failures in an effort to identify and correct potential sources of technical weaknesses.

There are several important technical features of the standard hiatal hernia repair/ fundoplication including hernia sac excision, mobilization of the esophagus, return of the stomach and distal esophagus to the intra-abdominal position, crural repair and wrap. Each has been recognized as important features of successful clinical outcome. [3] It has also been recognized that breakdown of the hiatal hernia repair is a common cause for clinical failure often within the first two years leading to wrap herniation, paraesophageal herniation and contributing to the slipped Nissen phenomenon. [4,5,6]

During the last 15 years our focus has been on the hiatal hernia repair as a source of technical failure and a prosthetic patch of the esophageal defect or reinforcement of the crural closure has been used on all patients. With experience the technique has evolved from a pure non-tension patch of the hiatal defect to partial crural closure with sutures reinforced with a PTFE/e PTFE prosthetic stapled and sutured into position to prevent motion, migration or erosion. In short, the crural closure is treated like a hernia repair anywhere else on the abdominal wall.

Materials and methods;

Records of 217 patients undergoing laparoscopic Nissen fundoplication are examined. Each patient had a prosthetic closure of the hiatal hernia or reinforcement of the crural repair. The first 112 patients (Group A) had a non-tension patch using a PTFE mesh stapled into position with 3 cm of mesh overlapped beyond crural bundles onto the diaphragm [7]. The next 105 patients (Group B) underwent hiatal hernia repair using an ePTFE/PTFE prosthesis[CruraSoft, Davol/Bard]. Fixation of this prosthetic as part of the crural closure evolved quickly to partial crural closure with one to 3 sutures and fixation of the prosthetic with staples[EMS 20; Ethicon] and 2-0 silk sutures (Figure 1). The prostheses was positioned onto the crural closure so that none of the material was abutting the esophagus. This required thorough esophageal mobilization and complete dissection of the crural columns posteriorly.

The focus of review was on short-term and long-term safety as well as durability of prosthetic crural repair. Post-operative dysphasia, stricture, integrity of the crural repair, reherniation and development of long-term complications were examined. Continuing analysis and retrospective review then permitted evaluation of the crural repair and fixation techniques that provided the best clinical outcome.

Results;

217 patients underwent laparoscopic Nissen fundoplication beginning in November 1993 with a non-tension prosthetic closure of the hiatal defect (Group A=112 patients) or prosthetic reinforcement of the hiatal hernia repair (Group B=105 patients).

Patient demographics are listed in Table 1. Of note is the large number of type II and type III hiatal hernias with 75 patients having “large” hiatal defects > 6 cm. Twenty-two patients (10%) underwent concomitant cholecystectomy and 11 patients were referred for recurrent hiatal hernias (two to 26 years postop).

Table 1.

Patient Demographics;

217 pts, 120 F, ages 19-87 (m=57.5 yrs.)

Mean Operative time: 155 min.

Pure paraesophageal hernias: 22 (10%)

Large Hiatal Defects > 6 cm: 75 pts (34.5%)

Mean age large defects = 69.7 yrs.

Mean Op. time(large defects): 205 min.

Safety short-term; There were no deaths, no transfusions and no intra-abdominal or mediastinal infections. There was no tearing or disruption of crural tissues. Placement of the prosthetic patch or reinforcement took 3 -5 minutes when staples alone were used. Adding two sutures to the fixation process added roughly 8 to 10 minutes to the procedure.

While most operations were Class I, totally clean procedures without violation of GI tract, 10% of patients had a cholecystectomy at the same time. One gastrotomy occurred during a redo operation that was closed immediately without contamination. One esophageal lengthening, Collis-Nissen was performed that officially made these cases Class II, clean-contaminated operations. Three urinary tract infections, two IV site cellulites, 3 pneumonias (in large paraesophageal hernias), a port site infection and sinusitis were treated aggressively. 16 postoperative fevers were related to atelectasis. There were no complications related to PTFE mesh fixation with staples or sutures.

Dysphagia to solid foods occurred routinely in the first month after surgery and patients were kept on a liquid or very soft diet. Dysphagia requiring endoscopic dilation after 6 weeks occurred in 36 patients (12%). 19 of these 36 patients had had pre-operative dilatation for stricture disease. Five patients in Group A have required two to four dilations. One patient developed a fibrous stricture of the esophageal hiatus four years after initial surgery and required reoperation. This patient had no intra luminal pathology or erosion. Five of seventeen primary postop dysphagia patients were from Group B each requiring only one dilation.

Safety long-term: There were 165 patients on whom data was available since November 1993. Data was available on all Group B patients. The mean follow-up is 71 months (6 to 173 months). Barium studies on 90 patients one day to 8 years postop (mean 38 months) found no impingement of the prosthesis onto the esophagus or stomach. 116 endoscopies on 86 patients (8 weeks to 8 years; mean 42 months) postop found no erosion or impingement. In the years after their index surgery 6 patients have undergone cholecystectomy, two for acute disease, two patients had appendicitis and 5 patients have been treated for diverticulitis. Nine others have had intra-abdominal surgery and two thoracic surgery. One patient had an esophagectomy for progressing Barrett’s where the pure mesh prosthesis was found to be thoroughly ingrown to the crus and required excision with scissors to allow room for the gastric tube. [7] Two patients had full-term pregnancies with one C-section. There have been no mesh infections and no mesh erosions.

Durability: Failure of the hiatal hernia repair as determined by symptoms and/or x-ray findings of wrap herniation occurred in 3 patients (1.4%). Two required reoperation. Hiatal hernia disruption occurred at 5 days, two weeks and two months after surgery. The earliest failure at 5 days was a totally non-tension PTFE repair (Group A) and occurred after heavy lifting. The next two were technical failures where the preshaped PTFE/ePTFE mesh was merely stapled to the left and right crus without crural closure and without sutures. In all subsequent patients (101 patients) the crus were closed partially or tight to a 58 F bougie and the prosthesis secured in position with staples and sutures.

All 217 patients underwent laparoscopic Nissen fundoplication after crural repair with prosthesis. In total, there have been three reoperations. One for stricture developing four years after initial surgery and 2 for disruption of the crural repair with wrap migration.

The l late occurring stricture and one crural disruption occurred in Group A the totally non-tension group. Two disruptions of the crural repair occurred in Group B before we learned to close the crus as a backing for the prosthesis and to suture as well as staple the prosthetic in place. Since the development of crural closure and suture fixation along with staples, there have been no crural disruptions or reherniation in the last 101 patients.

(followup 6 – 55 months).

Discussion;

It is recognized that disruption of the crural closure leads to wrap herniation and clinical failure. It should be no surprise to us that a primary tension closure of the crural pillars can fail. The crural pillars are pure muscle. There is no substantial fascia, no tendinous tissue and no ligamentous fibers. The crura are pure muscle and nowhere else in surgery do we close muscle to muscle and expect it to hold under tension. The myriad of studies [8,9,10,11] that show improved integrity of the crural repair after prosthetic reinforcement force us to examine the next step: How to do it safely. Granderath et al. define the ideal mesh as “easy to handle at laparoscopy, able to adhere to the diaphragmatic surface on one side and be benign to visceral surface on the other”. [12]. The next challenge is to position the prosthetic properly and secured it properly.

Several authors have shown that disruption of crural closure occurs frequently in the early postoperative months to years. [13,14,15,16] Filipi has shown that various “stressors” or intra-abdominal pressures can predispose to crural disruption.[17] Meshed prostheses do become ingrown into the crural muscle with time [7] . Our clinical failures even with a prosthesis occurred within the first two months after a non-tension repair and therefore are considered failure of fixation.

In subsequent patients then, the crus was closed with one to 3 sutures depending on the size of the defect to provide backing to the mesh, to provide substrate for staples and sutures were added to the fixation to provide extra strength and to prevent prosthetic movement until tissue ingrowth could occur.

The problem of erosion was not seen in this series over 15 years and yet there are scattered published reports of such and anecdotal reports (un-published) to this author. Erosion could result from improper placement of prosthetic i.e. placing the prosthesis too anterior and abutting the esophagus that may not be mobilized fully from its posterior position. Erosion could also result from migration and motion of a free-floating prosthesis (eg. Angelchik). For these reasons the prosthetic is now fixed to a closed crural defect with sutures and staples and positioned so that the esophagus is not touching any part of the prosthetic. The posterior wrap then further pulls the esophagus anteriorly and away from the crural repair.

Short term safety, avoidance of infection related to prosthetic at the hiatus is really a function of careful surgical technique and the positioning of the prosthesis on well vascularized crural muscle. No intra-abdominal or mediastinal infections were encountered despite 24/217 clean contaminated procedures and 10 peripheral postoperative infections.

Dysphasia occurring after 6 weeks occurred not surprisingly in 19 patients who had required dilation preoperatively. In 17 (17/217) patients primary postoperative dysphasia was treated with dilation and seemed more a consequence of the wrap and postoperative scar retraction or motility issues than narrowing of the esophageal hiatus. Among Group B patients only 5/105 had late postop dysphagia consistant with primary crural closure or mesh at the hiatus.5] In only one patient was there clinically significant hiatal scarring resulting in late dysphasia at 4 years. This rare phenomenon is known to occur historically with primary closure alone and is poorly understood.

Long term safety after now 15 years is quite satisfactory. Despite intra-abdominal infections from subsequent disease and other elective clean contaminated operations the prosthetic meshes placed at the hiatus are not susceptible to bacterial seeding. Just as prostheses on the anterior abdominal wall are covered with a protective peritoneal covering so too is the prosthetic at the hiatus.[18 ] Proper positioning and proper fixation of the prosthetic has prevented migration, prevented impingement and prevented erosion into the esophagus. The smooth ePTFE surface of the patch used on the most recent 105 patients prevents tissue ingrowth and thereby prevents fistulization to the stomach.

See Figure 1

In a primary crural closure one would not expect to see the top of the closure or the suture impinging on the esophagus at barium study or endoscopy. Similarly with a properly placed prosthetic patch reinforcing the crural closure no impingement was seen on barium study (90 patients) or at endoscopy (116 endoscopies/86 patients). This too is a consequence of proper placement and proper fixation of the prosthesis as well as thorough mobilization of the posterior seated intrathoracic esophagus. This mobilization is necessary to return the distal two to 3 cm of the esophagus to the intra-abdominal position and helps lift the esophagus anteriorly away from the crural closure and any prosthetic reinforcement. The bulky posterior wrap further holds the esophagus anteriorly away from the top of the closure. Most often the prosthetic is placed below the top of the crural closure. The approximated hiatal closure provides a substrate behind the mesh for both near term fixation with staples and long-term tissue ingrowth. The fixation with 15 to 20 staples (not tacks) and the addition of two sutures to fix the mesh to the crus in a full-thickness fashion. (Figure 1) further secures the prosthetic and prevents long-term migration and erosion. Suturing of the stomach, the wrap or the esophagus to the prosthesis is to be a voided.

Durability of the hiatal hernia repair using prosthesis is also quite satisfactory in this experience. In 3 patients breakdown occurred at 5 days, two weeks and two months. Two underwent subsequent repair with prosthetic. The very early failure was a consequence of extreme Valsalva at postoperative day 5 on a non-tension repair. The next two break downs were a failure of proper fixation where the PTFE/e PTFE prosthesis was merely stapled to the crural columns without closure( also non-tension). It became quickly apparent that crural backing was necessary for secure fixation as well as for long-term tissue ingrowth. The technique evolved quickly and incorporated fixation sutures on either side of the prosthesis incorporating the full-thickness crural column and the inner aspect of the prosthesis (Figure 1). Now after 15 years, 214 of 217 patients have had successful long-term treatment of their hiatal hernia. The hiatal hernia repairs remain intact after initial experience using a non-tension prosthetic repair (111 pts) or subsequent experience with 101 patients in whom a primary crural repair was reinforced with a PTFE/ePTFE prosthesis. The PTFE/ePTFE prosthesis was thought to be safer with the ePTFE surface facing the viscera and thereby preventing tissue ingrowth and erosion.

Seventy-five patients in this series had hiatal defects measuring 6 cm or greater. This large percentage of type II and type III hernias (34%) reflects an aging population (mean age of this subset was 69), increasing use of endoluminal therapies for small hiatal hernias and a continuing reluctance of gastroenterologists to surrender pure reflux disease that can be medically managed. This 15 year experience hopes to show that the hiatal hernia repair can be made durable with the use of prosthetic reinforcement and can be done safely using lessons learned from other prosthetic hernia repairs.

Table 2.

Review Articles

Summaries of Studies comparing Primary crural closure vs. Prosthetic Hiatal Hernia Repair.

Targarona et al [19] Mesh at the Hiatus (2004) 480 pts

Granderath et al [12] Prosthetic closure of the esophageal hiatus in the large hiatal

hernia repair and laparoscopic anti reflux surgery (2006) 766 pts

Johnson et al [20] Laparoscopic mesh hiatoplasty for paraesophageal hernias and fundoplications: a critical analysis of the available literature. (2006) 639 pts

The superiority of the prosthetic repair has been demonstrated in numerous studies over the last decade and includes 3 review articles(Table 2). With thousand of patients in hundreds of studies ethically the time has passed for randomization of primary repair versus prosthetic repair of the hiatal hernia. While there are surgeons who are capable of tailoring the crural repair to the individual patient [21] it may be safer for the general surgeon at large if we as a community can further define the best practice. Prosthetic reinforcement of the crural repair should now be considered as the standard of care. Remaining as a challenge is the ideal prosthetic, proper positioning and proper fixation.