An up-to-date review of the literature has revealed that many of the 2005 CER's original conclusions remain valid. In addition to the reconfirmation of these findings, further data were identified and included in the present report, expanding on these previous results. Notably, we added a section on the treatment of extraesophageal manifestations of GERD, which was not covered in the 2005 review. Furthermore, the present update also reviewed two new PPIs and one new endoscopic procedure. Table 46 summarizes the strength of evidence on treatments for GERD.

Table 46

Strength of evidence.

With regard to comparisons between surgery and medical therapy, we found that laparoscopic fundoplication in patients whose GERD symptoms were already well-controlled by medical treatments was at least as effective as continued medical treatment (and in some cases superior) in controlling GERD-related symptoms for the first 1 to 3 years following surgery, provided that the procedure was performed by experienced surgeons in high volume centers.

Bearing these findings in mind, the choice of laparoscopic fundoplication would be mainly targeted at those who wished to avoid the potential burden of lifelong medical treatments. Therefore, it is important to know how well the laparoscopic fundoplication actually succeeds in doing so. Of the three trials on laparoscopic fundoplication versus medical treatment reviewed, one reported that no patients treated with surgery were on medications at 1 year followup;²⁰ one reported 13 percent of the surgically treated patients were on medications at 1 year;¹³ and one (with a 3 year followup) did not report medication use.¹⁶ It appears clear that laparoscopic fundoplication is efficacious in helping patients to decrease the use of antireflux medications in the short term (≤1 year), but the longer term effect is uncertain. Of note, a long-term trial on open fundoplication versus medical treatments found that one-third of the surgically treated patients had received some form of antireflux medication by 12 years.¹⁷

Adverse events from surgery must also be considered. Fundoplication is associated with procedural complications like postoperative infections and incisional hernia, and morbidities like dysphagia and postprandial bloating, some of which may require surgical revisions. It would be helpful if one can predict preoperatively who would be at a higher risk of some of these postoperative complications. However, our review did not identify reliable patient or operative predictors of clinical outcome; age, morbid obesity, female sex, baseline symptoms, esophagitis, and hiatal hernia were all inconsistently associated with worse surgical outcomes.

However, medical therapy has also been associated with potentially serious complications. As in our previous review, serious complications reported with the use of PPIs include an increased risk of enteric infections (including Campylobacter and C. difficile) and pneumonia. An observation made since the 2005 review is a possible association between the use of PPIs and an increased risk of fractures.

For patients with GERD symptoms that cannot be adequately managed by standard medical treatments, published evidence to guide the choice of further therapy is not particularly helpful, as the available data are restricted to cohort studies lacking a proper control group. Of note, the two studies reviewed that explicitly included patients with an unsatisfactory response to medical treatments found that GERD symptoms had significantly improved after laparoscopic fundoplication in more than 5 years of followup.^80,81

Another important consideration is whether medical therapy or surgery is more effective in preventing long-term complications of GERD such as the development of Barrett's esophagus or esophageal adenocarcinoma. We did not identify sufficient evidence to conclude whether one or the other approach was more effective in preventing these adverse outcomes.

In addition to comparing medical and surgical therapies, our review also evaluated several new studies comparing specific medications, including two new PPIs. No consistent comparative difference in symptom relief was observed between esomeprazole (20 to 40 mg), lansoprazole (15 to 30 mg), pantoprazole (20 to 40 mg), or rabeprazole (10 to 20 mg). However, there is some evidence that rabeprazole 10 mg may provide better symptom relief than esomeprazole 40 mg at 4 weeks, and pantoprazole 40 mg better than esomeprazole 40 mg over 24 weeks.

With respect to dosing intervals, continuous dosing with PPIs was found to be more effective than on-demand dosing. For example, continuous daily intake of esomeprazole 20 mg or rabeprazole 20 mg appeared to provide better symptom control and quality of life relative to on-demand dosing over a period of 6 months. As for comparisons of different PPIs with over-the-counter dosages of omeprazole (20 mg), pantoprazole 40 mg and rabeprazole 20 mg provided significantly better symptom relief and healing of esophagitis at 8 weeks, and esomeprazole 20 mg provided better endoscopic remission rates as compared to over-the-counter dosages of lansoprazole (15 mg) at 6 months. While significant, the observed magnitude of these differences was generally small and the clinical relevance remains uncertain. It is possible that the variations in effectiveness may have been due to the specific doses examined.

As for the three available endoscopic procedures (EndoCinch^TM, Stretta^TM, EsophyX^TM) for the long-term management of GERD, effectiveness remains substantially uncertain. EndoCinch (suturing) and Stretta (radiofrequency ablation) had been previously examined in the 2005 CER; EsophyX (endoscopic fundoplication) is a new introduction. While some clinical benefits were observed in patients who had these procedures, the studies were generally small, of variable quality, and of short duration. In addition, all of these procedures have been associated with complications including dysphagia, infection/fever, and bloating.

For the treatment of patients with extraesophageal manifestations of GERD symptoms, no consistent benefit could be attributed to either medication or surgery. A small RCT found that patients' asthma symptoms improved after antireflux surgery compared to antireflux medical treatments, but these improvements could not be substantiated by objective testing.¹²¹ Similarly, some observational studies reported that antireflux surgery could be beneficial for those with asthma, chronic cough, or laryngeal symptoms. Despite the focus in our report on only those patients with asthma, chronic cough, or laryngeal symptoms, we surmise that the considerable clinical heterogeneity within these subgroups precluded the detection of a reliable effect, if one exists. It is recognized that the quality of primary studies in this field is limited. Within these studies, the impact of GERD treatment may be limited, as GERD may not be the cause for symptoms in all of the subjects in the study. The treatment population will include both responders (participants with reflux triggered symptoms) and non-responders (participants whose symptoms are not reflux triggered). This will dilute the overall treatment effect. Without any tests and biomarkers to identify GERD-related symptoms, it is not possible to accurately estimate the potential effects from anti-acid treatments.

We acknowledge the limitations of largely relying on existing systematic reviews to review the evidence for the treatment of patients with extraesophageal manifestations of GERD symptoms; it would not have been feasible to review all the primary literature on this topic within the time period of this report. Even though we have assessed the reporting quality of these systematic reviews, we did not re-analyze the primary studies.

While we have made every attempt to address the Key Questions set out in the present review, the available evidence had several important limitations:

• Studies directly comparing surgery to medical therapy generally had high dropout-rates in long-term followup (e.g., 58 percent of patients were lost to followup at 12 years in a study comparing medical treatment and open fundoplication).
• There was a great deal of variability in the rigor of how the outcomes were evaluated across studies, particularly in subjective endpoints (e.g., some used a validated measure of quality of life, while others used symptom scales whose measurement properties have not been well characterized).
• Most studies were non-randomized or lacked a suitable control group.
• The majority of the included studies had a relatively short followup (typically no longer than 1 year), particularly those concerned with medical treatments.
• Pharmacologically equivalent doses of various PPIs have not been well established (or universally agreed upon), thus clouding interpretation of existing comparative PPI studies.
• Reporting of adverse events was often incomplete and inconsistent across studies; some studies did not report specific adverse events and the definitions of adverse events differed across studies.