Five-Year Follow-up of a Multicenter, Double-Blind Randomized Clinical Trial of Laparoscopic Nissen vs Anterior 90° Partial Fundoplication FREE

Laparoscopic Nissen fundoplication is generally considered to be the surgical procedure of choice for the treatment of gastroesophageal reflux disease. It provides durable control of reflux but at the expense of a small but significant rate of dysphagia and gas-related adverse effects.¹ To minimize the risk of such adverse effects, partial fundoplications, anterior and posterior, have been adapted from the original Nissen fundoplication. In the recently reported long-term follow-up of a randomized trial comparing laparoscopic Nissen fundoplication with laparoscopic anterior 180° partial fundoplication, it has been shown that both procedures produce equivalent and durable control of reflux at up to 10 years' follow-up, with trends toward less dysphagia and better preservation of belching following anterior 180° partial fundoplication.²

With the aim of developing an effective antireflux procedure, but with minimal adverse effects, we modified the anterior 180° fundoplication further to an anterior 90° partial fundoplication.^3- 4 We first piloted this procedure in an experimental animal reflux model, and postfundoplication manometry in the model showed that the resting lower esophageal pressures and reflux were corrected similarly by the anterior 90° fundoplication when compared with a standard Nissen fundoplication.³ This technique was then evaluated clinically against laparoscopic Nissen fundoplication in a multicenter, double-blind randomized control trial—this study.⁵We have previously reported 6 months' follow-up outcomes from this trial, and these early results suggested that patients were more likely to be able to belch and relieve gas bloat symptoms after anterior 90° partial fundoplication, offset by more heartburn-like symptoms, compared with patients who underwent laparoscopic Nissen fundoplication.⁵ Overall satisfaction with the outcome of antireflux surgery was very high (98%) at 6 months following anterior 90° partial fundoplication, and the early data lent support to the hypothesis that anterior 90° fundoplication is an appropriate surgical procedure for the treatment for gastroesophageal reflux disease, although longer-term follow-up was needed to support this early outcome data. Five years have passed since the early outcomes were evaluated, and in this article, we are now able to report longer-term (5 years' follow-up) clinical outcomes from this randomized trial.

The trial was undertaken in 9 university teaching hospitals in 6 major cities in Australia and New Zealand. Fifteen surgeons performed the surgical procedures, and all had significant experience with laparoscopic antireflux surgery before the trial commenced. The full details of the trial protocol have been described previously.⁵ In brief, symptomatic patients with objective evidence of gastroesophageal reflux at upper gastrointestinal endoscopy or 24-hour pH monitoring were considered for entry. Patients were excluded if they had an esophageal motility disorder that was deemed to preclude performing a Nissen fundoplication or if they were undergoing a concurrent procedure or had undergone previous antireflux surgery.

Randomization was performed by opening a sealed envelope that specified the type of fundoplication to be performed. The envelopes were prepared independently and were only opened after the operation had commenced and it was determined that both operative approaches were feasible. The operative techniques have been described previously.^4- 5Common to both procedures was a laparoscopic approach and routine posterior hiatal repair. A loose 1- to 2-cm Nissen fundoplication was fashioned following division of the short gastric vessels and full mobilization of the gastric fundus. Key steps for the anterior 90° partial fundoplication included a posterior esophagopexy to the right hiatal pillar, fixation of a length of esophagus within the abdomen, recreation of the angle of His, and construction of a fundoplication that covered the left anterolateral intra-abdominal esophagus. A bougie was not used, and short gastric blood vessels were not divided. If a laparoscopic procedure was converted to an open approach, the selected fundoplication was still performed, and patients were still included in the data analysis on an intention-to-treat basis. Operative techniques were standardized by agreement between participating surgeons. Additional quality control was achieved by presentation and distribution of videotape demonstrations at meetings in December 1998 and December 2000.

The type of procedure performed was concealed from the patient as well as the research assistant performing subsequent clinical follow-up. Preoperative information was collected pro forma by the operating surgeon. Postoperatively, a research assistant collected information using a structured telephone interview at 1, 3, and 6 months and then annually thereafter. Details of the interview have been published previously.⁵ Patients were asked about the presence or absence of certain symptoms and asked to grade heartburn and dysphagia (separately for liquids and solids) using analog scores from 0 to 10 (0 = no symptoms, 10 = severe symptoms). Dysphagia was also scored using a previously validated composite dysphagia score graded from 0 to 45.⁶ Patients were also asked whether they were able to belch and eat a normal diet.

The overall outcome was graded using an analog score for overall satisfaction (0 = unsatisfied, 10 = totally satisfied) and a previously described 4-point outcome scale (excellent, good, fair, and poor).⁷ Patients were considered to be satisfied if the outcome was rated as good or excellent. A further yes/no question asked whether patients thought they had made the correct decision to undergo antireflux surgery.

Data were analyzed on an intention-to-treat basis, with all patients remaining in their initially allocated groups. Data were stored in a database (Filemaker Pro version 8.5; Filemaker Corp, Santa Clara, California) and statistical analysis was performed with GraphPad Prism version 5a (GraphPad Software Inc, La Jolla, California). The human research ethics committee in each participating hospital approved the study protocol. The study was conducted in accordance with the guidance of the World Medical Association of Helsinki (revised 1989) and the National Health and Medical Research Council of Australia guidelines on human experimentation.

From February 11, 2000, to February 26, 2003, 112 patients were enrolled in this trial, and all underwent surgery. Sixty were randomized to undergo laparoscopic anterior 90° partial fundoplication and 52 to laparoscopic Nissen fundoplication. All patients underwent the specific fundoplication type that was allocated by the randomization process. There was 1 conversion to open operation during anterior 90° partial fundoplication. This was due to intra-abdominal adhesions. There were 2 conversions during Nissen fundoplication due to intra-abdominal obesity and bleeding.

Demographic, preoperative, and early postoperative (first 6 months) data for the 2 groups have been reported earlier.⁵ There were no significant differences with respect to any preoperative variable except that a higher proportion of patients in the Nissen fundoplication group had undergone previous upper abdominal surgery.

At 1-year follow-up, data were collected from 105 patients (57 anterior, 48 Nissen; 94% follow-up). Three patients died between 1 and 5 years' follow-up from unrelated causes. At 5 years, follow-up data were not able to be obtained from a further 12 patients. Hence, a clinical outcome was available for 89% at 5 years. In the anterior fundoplication group, 5 patients were lost to follow-up, 1 refused follow-up, and 1 died, whereas in the Nissen fundoplication group, 3 were lost to follow-up, 3 refused follow-up, and 2 died. No revision operations were undertaken in either group during the 6 months' to 5 years' postoperative follow-up period.

Clinical follow-up data at 12 months and 5 years are summarized in Tables 1, 2, 3, and 4. When patients were asked using a yes/no question whether they had heartburn at 12 months' follow-up, significantly more reported heartburn after anterior 90° partial fundoplication (Table 1), although at 5 years there was no statistically significant difference (Table 2). However, when heartburn was assessed using the 0-10 analog score, patients reported significantly more heartburn after anterior 90° partial fundoplication at both 12 months' and 5 years' follow-up (Table 1 and Table 2).

At 12 months after surgery, patients were less likely to report dysphagia for solids following anterior 90° partial fundoplication, although there were no significant differences between the 2 groups for the dysphagia scores for both liquids and solids at 1 year (Table 1), and at 5 years' follow-up, the incidence and severity of dysphagia were not significantly different for the 2 groups for all measures of dysphagia. The ability to belch was better preserved 1 and 5 years after anterior fundoplication (Table 1 and Table 2). At 12 months' and 5 years' follow-up, there were no significant differences for any measures of satisfaction with the overall outcome (Table 3 and Table 4).

Antireflux surgery has an important role in the treatment of patients with refractory gastroesophageal reflux disease and in patients who do not wish to continue taking lifelong antisecretory medication. Traditionally, the Nissen 360° fundoplication has been widely applied, and it provides good control of reflux for most patients, with success rates approaching 90% at 10 years' follow-up.^8- 9 However, in some patients, achieving excellent reflux control can be followed by significant long-term adverse effects, most commonly dysphagia, inability to belch, gas bloat, and flatulence. To minimize the risk of adverse effects, many surgeons construct a partial fundoplication.^2,4,10 This strategy is reported to reduce the risk of dysphagia and wind-related adverse effects, although there has been concern that the control of gastroesophageal reflux may not be adequate following a partial fundoplication.¹¹

When constructing a partial fundoplication, there are 2 broad approaches to consider: anterior vs posterior. Posterior fundoplications have been more widely performed, particularly in North America and Europe, and studies of case series generally report good outcomes. This type of partial fundoplication has been compared with Nissen fundoplication in randomized controlled trials.^10,12- 13 In general, the results from these trials demonstrate similar outcomes for reflux control and dysphagia, but with less wind-related adverse effects following posterior partial fundoplication. To our knowledge, only 1 randomized trial has demonstrated less dysphagia following posterior partial fundoplication, and this was only at 4 months' follow-up.¹³ Hence, we have hypothesized that anterior partial fundoplication may be associated with fewer adverse effects but still achieve adequate reflux control.

Four randomized trials, including our current trial, have compared an anterior partial fundoplication variant with Nissen fundoplication.^2,5,14- 15 Two of these evaluated an anterior 180° partial fundoplication.^2,14 Outcomes at 6 months and 5 years from the trial undertaken in Adelaide, Australia, demonstrated similar reflux control, but less dysphagia and other adverse effects, following anterior 180° partial fundoplication.^16- 17 However, at the 10-year follow-up, the outcome following anterior 180° partial fundoplication was not significantly different from that of Nissen fundoplication.² The second trial from Baigrie et al¹⁴ also demonstrated fewer adverse effects offset by more heartburn symptoms at 2 years' follow-up following anterior 180° partial vs Nissen fundoplication. In both of these trials, high rates of patient-reported satisfaction were observed following anterior 180° partial fundoplication at up to 10 years' follow-up. Anterior 90° partial fundoplication has also been compared with Nissen fundoplication (without division of the short gastric blood vessels) in a further trial from Adelaide.¹⁵ The 6 months' follow-up outcomes were similar to the early outcomes from our current trial, ie, fewer adverse effects following anterior 90° partial fundoplication, good reflux control, and high rates of satisfaction.¹⁵ The only trial to report a poor result following an anterior partial fundoplication variant was reported by Hagedorn et al.¹¹ They compared anterior 120° partial fundoplication with posterior fundoplication. In their trial, there was a high rate of recurrent reflux following anterior partial fundoplication at early follow-up, and this was confirmed by pH monitoring. However, these outcomes were very different from those reported in the other trials. A possible explanation for this is that the construction of the anterior 120° partial fundoplication applied in the trial reported by Hagedorn et al was different from that tested in other trials, with fewer sutures used to attach the gastric fundus to the right hiatal pillar and therefore less anchorage of the esophagus and fundoplication within the abdomen.

In our current trial, the initial 6-month follow-up outcomes were excellent following anterior 90° partial fundoplication.⁵ Reflux control, as measured by symptom scores and pH monitoring, was very good, adverse effects were less frequent, and 98% of patients were highly satisfied with the overall outcome following anterior 90° partial fundoplication. However, when treating reflux surgically, long-term outcomes are very important, and the 5-year follow-up reported in the current article suggests patients were more likely to experience reflux symptoms at later follow-up following anterior 90° partial fundoplication. Furthermore, the previously reported advantages of better satisfaction and fewer adverse effects were not seen at 5 years' follow-up. However, while there were no statistically significant differences between the dysphagia scores for the 2 study groups, there was a trend toward less dysphagia following anterior 90° partial fundoplication, and this trend is consistent with the results of other trials.

In considering heartburn and other clinical indicators of recurrent reflux, we use 3 clinical questions. While it is unlikely that any one of these correlates perfectly with reflux, it is likely that they do provide an overall indication of the relative risk of further reflux following each operation. For instance, heartburn was assessed with a yes/no question and an analog score. Affirmative answers to these questions do not always indicate reflux, as other symptoms could be confused with actual reflux. Similarly, not all individuals who consume proton pump inhibitors following fundoplication have reflux.¹⁸ For these reasons, it is unlikely that 27% and 44% of patients had recurrent reflux after Nissen and anterior fundoplication, respectively, particularly because at 5 years' follow-up satisfaction with surgery was approximately 90% following both procedures. Nevertheless, the reported clinical outcomes are likely to be informative, to the extent that they suggest reflux is a greater problem at 5 years' follow-up following anterior 90° partial fundoplication compared with the Nissen procedure.

Unlike other randomized trials that have evaluated antireflux procedures within a single unit, our current trial is a multicenter study from 9 sites spread across Australia and New Zealand. While all the surgeons involved in the trial work in university teaching hospitals, they also work in both teaching hospital and private practice environments, and approximately 50% of the patients in this trial were managed in the private sector. Furthermore, the majority of laparoscopic antireflux surgery in our countries is performed by specialist upper gastrointestinal surgeons who work in these settings, and this reflects the range of patients recruited into the trial. For these reasons, we believe that the results better reflect the outcomes for this type of surgery across Australia and New Zealand than previously reported studies from single centers.

The follow-up rate of approximately 90% at 5 years in our trial was not as high as the 96% to 99% follow-up rates reported in other randomized controlled trials from Adelaide.^2,7,15,17 This was because the multicenter recruitment process increased the difficulties associated with tracking patients across multiple Australian states and New Zealand. Nevertheless, we were still able to obtain a high rate of follow-up, and few other randomized trials have been able to obtain this level of long-term follow-up. The outcomes reported in this trial were clinical symptom scores. Even though at late follow-up we did not undertake further investigation with studies such as pH monitoring, we are confident that the symptom scores reflect the clinical situation and provide reliable information about the surgical outcome. Esophageal manometry, pH monitoring, and endoscopy were undertaken at early follow-up, but in our community, it is not possible to obtain adequate compliance with pH studies and other investigations at multiple points during clinical trials. When patients undergo surgery for gastroesophageal reflux, arguably the most important outcome is a patient who is satisfied with the overall result following surgery. When patients consider this outcome, they balance reflux control against new postsurgical adverse effects. Some patients with excellent postsurgical reflux control are unsatisfied with the results of antireflux surgery because they developed unwanted adverse effects. There were no statistically significant differences between the satisfaction outcomes at 1 or 5 years in this trial.

Given the outcome at 5 years, how have the results of our trial influenced the practice of antireflux surgery in our hospitals? Following completion of the enrolment phase, the majority of participating surgeons reverted to their standard clinical practice. For most surgeons, this meant undertaking a Nissen fundoplication, although some surgeons adopted posterior or anterior 180° partial fundoplication procedures, with 1 surgeon continuing to apply the anterior 90° partial fundoplication for selected patients undergoing surgery for reflux. In general, however, the participants chose to await the longer-term outcomes from the trial before considering routine clinical application of the anterior 90° partial fundoplication procedure.

In conclusion, at 5 years' follow-up, patients who underwent a laparoscopic Nissen fundoplication were less likely to report heartburn-type symptoms compared with those undergoing anterior 90° partial fundoplication. For the other outcomes (adverse effects and satisfaction), the results were not significantly different at 5 years' follow-up. This suggests a better overall outcome at 5 years following laparoscopic Nissen fundoplication compared with anterior 90° partial fundoplication.

Correspondence: David I. Watson, MD, FRACS, Flinders University Department of Surgery, Room 3D211, Flinders Medical Centre, Bedford Park, South Australia 5042, Australia (david.watson@flinders.edu.au).

Accepted for Publication: June 17, 2009.

Author Contributions:Study concept and design: Watson, Devitt, Fletcher, Martin, and Nathanson. Acquisition of data: Nijjar, Watson, Archer, Bessell, Booth, Cade, Cullingford, Fletcher, Hurley, Kiroff, Nathanson, and Windsor. Analysis and interpretation of data: Nijjar, Watson, Jamieson, Archer, Cade, Fletcher, and Windsor. Drafting of the manuscript: Watson, Devitt, and Nathanson. Critical revision of the manuscript for important intellectual content: Archer, Bessell, Booth, Cade, Cullingford, Fletcher, Hurley, Kiroff, Martin, and Windsor. Statistical analysis: Nijjar and Watson. Obtained funding: Watson. Administrative, technical, and material support: Nijjar, Watson, Archer, Booth, Cade, Fletcher, Martin, and Windsor. Study supervision: Watson, Jamieson, Bessell, Devitt, Hurley, and Windsor.

International Society for the Diseases of the Esophagus–Australasian Section: Rajwinder S. Nijjar, MBChB, MSc, FRCS; David I. Watson, MD, FRACS; Glyn G. Jamieson, MS, FRACS; Stephen Archer, PhD, FRACS; Justin R. Bessell, MD, FRACS; Michael Booth, MBBS, FRACS; Richard Cade, MBBS, FRACS; Graham L. Cullingford, MBBS, FRACS; Peter G. Devitt, MS, FRCS, FRACS; David R. Fletcher, MD, FRACS; James Hurley, MBBS, FRACS; George Kiroff, MS, FRACS; Ian J. G. Martin, MBBS, FRACS; Leslie K. Nathanson, MBChB, FRACS; John A. Windsor, MD, FRACS.

Financial Disclosure: None reported.

Funding/Support: This trial was supported by project grants from the National Health and Medical Research Council of Australia.

Additional Contributions: Carolyn Lally, RN, RM, Lorelle Smith, BNurs, Janet Pinno, and Nicky Ascott conducted the follow-up in this trial and managed the trial database. The late Christopher Martin, MD, FRACS, Nepean Hospital, Sydney, Australia, supported this trial and contributed patients to the study. We continue to appreciate his contribution and support.