Randomized trial of total fundoplication and fundal mobilization with or without division of short gastric vessels . A short-term clinical evaluation 1

Purpose: Evaluate short-term results after fundoplication procedure, concerning the division of short gastric vessels. Methods: A prospective randomization of 90 patients with indication for hiatoplasty and total fundoplication with fundus mobilization was performed. They were divided into two groups: no SGV division (group A, n= 46) and with SGV division (Group B, n=44), although in both groups the gastric fundus was mobilized to perform a fl oppy valve. Early outcome with clinical follow up (1 year) was observed. Results: Both groups were similar regarding preoperative parameters and severity of gastroesophageal refl ux disease (GERD). No difference in morbidity was observed during hospital stay. Nevertheless, the median operating time was 80,2 minutes in group A and 94,1 minutes (p=0,021) in Group B. Transitory dysphagia during the fi rst year was signifi cantly lower in group B (46,6% versus 23,2%, p=0,012). However, in 12 months clinical outcome was similar in both groups (clinical symptoms of GERD, persistent dysphagia and reoperations). Conclusion: There was no improvement in routine division of SGV in total fundoplication procedure when the gastric fundus was mobilized.


Introduction
Recently, total fundoplication followed by esophageal hiatoplasty under laparoscopic technique has been the most common surgical treatment for Gastroesophageal Refl ux Disease (GERD) 1,2 .The surgical technique was described by Nissen in 1956 and, since then, some modifi cations of the original method were reported, in order to decrease side effects as dysphagia, gas bloating, and diffi culty in vomiting and so on.The most important modifi cations were described in the pre-laparoscopy era.Donahue et al. 3 proposed the fl oppy fundoplication and DeMeester et al. 4 observed that a two centimeters valve was enough for the success of the operation (short fundoplication).This last group suggested a routine short gastric vessels (SGV) division, including gastric adhesions as well as the pancreatic-gastric vessels division, so as to achieve a short and fl oppy fundoplication.Randomized trial during laparotomy era showed similar results with this technique but a higher hiatal hernia long-term incidence after SGV division 5,6 .After the development of laparoscopic access described by Geagea 7 and Dallemagne et al. 8 , the fi rst reports suggested the routine SGV division to reduce postoperative dysphagia 9,10,11 .However, other researches did not had the same results 12,13,14,15 .The randomized clinical trials (RCT) during laparoscopic era present now different standards of technique: the mobilization extension, number of divided vessels and the technique to make those divisions.Watson et al. 16 , used metallic staples; other techniques included monopolar cauterization and harmonic scalpel 17,18 .When the SGV were not divided, the anterior wall of gastric fundus was used for the fundoplication, and not the posterior wall 16,17,18 .A metanalysis published by Catarci et al. 19 concluded that the routine division of SGV had no advantage.However, criticism was made to this study, concerning specifi c aspects related to the gastric fundus mobilization such as: instruments, experienced surgical team and the number of different surgeons performing the procedure 20 .The aim of this study was to compare the results of total fundoplication and gastric mobilization with and without SGV division, after a one-year follow-up

Methods
We included 122 adult patients with indication for surgical treatment of GERD in the Gastrointestinal Division of the Federal University of Sao Paulo (UNIFESP) and the General Surgery Division of the State Server Public Hospital (HSPE).The study was approved by the Ethics Committee in Research of the Federal University of Sao Paulo.Thirty two patients were excluded from the study due to the following reasons: esophageal body dysfunction characterized by low pressure on manometric study (<30 mmHg) and the presence of more than 20% of nonperistaltic waves; hard stenosis, unable to dilate; Barrett esophagus with complications as dysplasia, ulcers and stenosis; short esophagus; paraesophageal hernia; previous surgery in the gastroesophageal area; and patient denial to participate in the research.As a result, this studied was carried out with 90 patients, mean age of 46,3 years, composed by 53,3% male and 46,7% female patients.Ten patients (11,1%) underwent cholecystectomy at the same procedure time.The patients were randomized in two groups: Group A (without SGV division) and Group B (with SGV division).The study was a clinical trial with intention to treat method; thus the patients remained in the same group as previously allocated, independently on the real treatment received.

Pre-operative evaluation
GERD was diagnosed based on typical clinical symptoms relieved after proton pump inhibitor use, as well as endoscopic fi ndings of erosive esophagitis.When different presentations were noted, a 24 hour pHmetric exam was performed.Symptoms were graded regarding severity and occurrence: grade 0 -asymptomatic; grade 1 -occasional complaint without specifi c medical therapy; grade 2frequent complaints, with continuous medical therapy to relieve; grade 3 -no relieve even after medical therapy and recommendations.Dysphagia symptom was also graded21: grade 0 -no dysphagia; grade I -mild dysphagia; grade II -slight obstruction, with liquid intake necessary to relieve; grade III -progressive dysphagia for solid seeking medical care, hospital stay or total obstruction.All patients underwent endoscopy and esophageal manometric study.Endoscopic fi ndings were graded concerning the degree of esophagitis and its complications, following Savary-Miller modifi ed classifi cation.

Surgical technique
All procedures were performed by the same surgeon, with similar dissection and hiatoplasty time in both groups.The abdominal esophagus was dissected and 5,0 cm extension was isolated together with vagus nerve.The hiatoplasty was done with "X" 2.0 non-absorbable stitches, tightening the hiatus so as to allow a 10 mm instrument throughout the hiatus.

Gastric fundus mobilization -Group A -no SGV division.
The gastric fundus mobilization started with dissection of the left arm of the right hiatal arm, from the right side of the esophagus isolated and pulled to the left, in order to fi nd a low irrigated area.This presentation was achieved with pulling the hiatus and contra-traction with exposition of the posterior wall of gastric fundus.The vessels that communicate the gastric fundus with the inferior phrenic vessels were divided routinely (Figure 1).Most of the time, the mobilization was completed by the left side, with incision of phrenic-esophageal tissue and the adhesions near the superior area of the spleen.Gastric fundus mobilization -Group B -with SGV division SGV were divided by the use of harmonic scalpel from the inferior splenic area following the gastro-splenic omentum in cranial direction until the esophago-gastric transition (EGT).The retroperitoneal cavity was opened after division of the initial gastric vessels and following the great gastric curvature, in order to proceed division of all SGV, with complete liberation of the gastric wall from the splenic vessels and retroperitoneum, through the pancreatic-gastric vessels division (Figure 2).The anti-refl ux valve was performed with the tension-free gastric fundus placed on the right side of the esophagus (no need to use instrument for this traction -Figure 3) and without torsion of the gastric fundus (shoe shine maneuver).The fundoplication had 2-3 centimeters, with three 2.0 non-absorbable stitches.The proximal and the distal stitches were applied to the gastric wall and the middle one was applied to the gastric wall together with the esophageal wall, to avoid migration of the valve.The procedure was performed with calibration of a 10 mm instrument, to confi rm the fl oppy valve (Figure 4).Postoperative care Patients were kept with nasogastric tube for eight to 12 hours.Liquid diet was given on the fi rst postoperative day and the progression to solid food after three to four weeks after the procedure.Excessive physical exercises were prohibited for three months.Patients were evaluated by the surgeon on postoperative days: seven, 30, 90, 180 and after one year of the surgical procedure.

Evaluation of intra-operative results
The following issues were evaluated: intra-operative complications; signifi cant bleeding (need to remove blood and clots with suction); visceral perforations or lesions; operative time; conversion to open technique; change in the beginning planned technique.

Evaluation of immediate post-operative results (until 30º P.O.)
Evaluated concerning morbidity (general and specifi c complications); hospital stay length; and mortality.

Results evaluation after one year
The patients were evaluated during the fi rst year (three months, six months and one year) regarding the presence of dysphagia and graded according to Gotley et al. 21from 0 to 3. Dysphagia was considered transitory until 90 days and continuous if lasting longer than 12 months postoperatively.The recurrence of symptoms of GERD and the need for medical therapy was considered in the postoperative evaluation.

Statistical analysis
The categorical variables were compared between the groups using non-parametric Chi-square test and Fisher´s exact test.Regarding numeric variables, the t-Student test was applied when normality supposition was satisfi ed or the Mann-Whitney non-parametrical when this supposition was not satisfi ed.The normality of data was verifi ed using the Kolmogorov-Smirnov test.The signifi cance level of the tests were α=0,05.

Results
The 90 patients enrolled in this trial, 46 were allocated in Group A (no SGV division) and 44 patients in group B (with division).Tables 1 and 2 describe demographic, clinical, endoscopic and manometrical fi ndings.No signifi cant difference was found between both groups concerning the presence of hiatal hernia, complications from esophagitis, Barrett esophagus and manometric data of lower esophageal sphincter pressure (LES).As regards to intestinal metaplasia, eight patients had long Barrett esophagus and seven patients presented short Barrett esophagus.

Intra-operative results
The only signifi cantly different data found was the mean surgical procedure duration (Table 2), being 80,2 minutes in group A against 94,1 minutes in group B (p=0,0045).Seven (7,7%) from the 90 patients developed intra-operative complications, three (6,5%) in group A and four (9,1%) in group B. No conversion to open surgery was necessary.However, in one patient from group A, it was not possible to perform the fundoplication without division of SGV (technique failure), that as carried out with no complications.

Immediate postoperative results
Seven patients presented low risk postoperative complications, with no difference between groups, with same hospital stay length (Table 3).There was no mortality rate and no re-intervention procedure was necessary.

One-year follow-up results
Dysphagia was related until three months postoperatively in 39 (43,3%) patients (Table 4), more frequent in group A (56,5%) than in group B (29,5%), p=0,012.After 6 months, this difference turned out to not to be statistically different between groups (six patients -three from group A and three from group B -complaint of dysphagia grade I or II).Severe dysphagia (grade III) was referred by three patients (two from group A and one from group B), with the need of endoscopic dilation to relieve symptoms.After one-year follow-up, one patient from each group had symptoms recurrence (Table 4).

Discussion
Routine division of SGV has been proposed since laparotomy era.Nevertheless, after the publications of laparoscopic technique to perform the fundoplication for the treatment of GERD 7,8 , the controversies regarding this issue remained.Furthermore, new technical details such as new instruments for a new surgery, learning curve on this new surgical practice, technical diffi culties for the division of the SGV have been discussed.Some publications relating important and long-lasting dysphagia with the non-liberation of the gastric fundus through SGV division 9,22,23 .New techniques were described, as the fundoplication without SGV division to prevent such complication 24 .The fi rst randomized clinical trial in the laparoscopic era, concerning the routine division of SGV, showed that no manometric, endoscopic and clinical improvement was detected after one-year follow-up.However, in that study, seven different surgeons performed the procedures and different techniques (metallic staples) were used for the SGV division 16 .Another randomized trial evaluated 90 patients and found no difference comparing clinical outcome one-year after the surgical procedure 17 .Moreover, another study with 56 patients compared phmetric, manometric, endoscopic and contrast-enhanced radiography data and revealed no signifi cant difference between groups.After 12 months, dysphagia and recovery from GERD were similar in both groups, although patients that underwent SGV division presented more gas bloating complaint 18 .However, none of the reported studies had a uniform technique of SGV division with harmonic scalpel (group B) and gastric fundus mobilization using posterior and anterior wall (group A) for the fundoplication, one of the most important differences of our study.Indeed in both groups we probably performed more extensive mobilization than others previous RCTs.The use of harmonic scalpel for SGV division has been advocated to be better than with the use of staples 25,26 .Another study compared the harmonic scalpel with the bipolar electrocautery and found no signifi cant difference comparing intra and immediate postoperative aspects 27 .Despite the fact that we used the harmonic scalpel for SGV division in every procedure, this group had higher procedure duration.All three randomized clinical trials (RCT) also revealed longer procedure time in the SGV division groups 16,17,18 .We must emphasize that the extended gastric mobilization including division of all SGV and adhesions of the posterior wall as proposed by DeMeester et al. 4 was reposted only in our study.The others described three to four SGV division, usually the most superior and no care was taken with adhesions and posterior vessels 16,17,18 .In our study, three patients developed no-signifi cant bleeding during SGV division (no hemodynamic alteration, conversions and no transfusion required).Previous RCT reported important bleeding in seven patients, all after the use of metallic staples, requiring conversion in two cases and re-intervention in one16 and bleeding and re-operation 17,18 .These complications should have been avoided by the use of the harmonic scalpel.In one of our cases, we were unable to perform the fundoplication without SGV division (initial proposal), due to technical diffi culties.Blomqvist et al. 17 had the same situation and Watson et al. 16 gave up dividing the SGV to shorten operation time due to clinical problems.The performance of the fundoplication with gastric body without SGV division may lead to anatomical and functional dysfunction.Some anatomical issue may be related to the necessity of SGV division to create a tension-free anti-refl ux valve.The variations of gastric fundus determined the need to divide SGV28.Anyhow, the access to gastric fundus on the right side of the esophagus without SGV division had shorter operation time and was accomplished safely in almost every procedure.After oneyear follow-up, the results show that dysphagia was the main complaint.Transitory dysphagia was more common in the SGV division group, different from the previous RCT´s 16,17,18 .The exact reason for dysphagia remains unclear and may be related to patients past medical history or surgical technique details.DeMeester et al.4 proposed the use of large tube (48-50FR) inside the esophagus for calibration of the fundoplication to prevent transitory dysphagia.These fi ndings were similar to the results of Patterson et al. 29 .Some authors advocate no advantages by this routine technique 30 , emphasizing the risk for perforation and severe complications related to these tubes 29,31 .We do not use this tube, but we double-check the gastric fundus fl oppy anti-refl ux valve as described previously.Dysphagia defi nition, intensity and classifi cation characterization may present variations.However, it may happen in the early postoperative period in more than 90% of patients, and this shall be considered to be normal and sometimes even desired 32 .As regards to permanent dysphagia , we observed that this complaint was less common in both groups after six months, and the difference turned out to be not signifi cant, as it was in early postoperative time.Six patients (6,2%) presented with any dysphagia degree after one year (grade I and II, with no need for endoscopic dilation or medical therapy).No difference was found when comparing the severity of dysphagia, as dysphagia grade III was similar in both groups, as it was the need for endoscopic dilation and reoperation for this purpose.This observation matches with the literature.Permanent dysphagia was reported from 0% to 5% in analyzed series 2,33,34,35,36 .Finally, two patients presented GERD symptoms recurrence after one year.Still, these evaluations need long-term results.

Conclusion
Routine SGV division for gastric fundus mobilization and fundoplication did not prove to have any defi nitive advantage concerning clinical outcome one year after surgery.

FIGURE 1 -
FIGURE 1 -Gastric fundus dissection close to the diaphragm (a) and vessels division (b) with mobilization of the posterior wall of gastric fundus (c) on the right side of esophagealgastric transition

FIGURE 2 -
FIGURE 2 -SGC division, from the inferior splenic area (A and B) and the gastric-pancreatic vessels (C)

FIGURE 3 -FIGURE 4 -
FIGURE 3 -Gastric fundus posterior wall mobilized (A and B) and the positioning of the gastric fundus without instrument fi xation (C and D)

TABLE 1 -
Characteristics of patients enrolled in the trial Q: Chi-Square test; F: Fisher´s exact test ; t-S: t-Student test; M-W: Mann-Whitney test; *: signifi cant difference.

TABLE 4 -
First year follow-up Q: Chi-Square test; F: Fisher´s exact test ; t-S: t-Student test; M-W: Mann-Whitney test; *: signifi cant difference.# unable to perform