Pii: s1542-3565(04)00288-5

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2004;2:656 – 664 Effectiveness of Proton Pump Inhibitors in Nonerosive Reflux BONNIE B. DEAN,* ANACLETO D. GANO, JR.,* KEVIN KNIGHT,* JOSHUA J. OFMAN,*,‡ andRONNIE FASS§*Cerner Health Insights, Beverly Hills, California; ‡Cedars-Sinai Medical Center, Department of Medicine, Los Angeles, California;and §Southern Arizona VA Health Care System, Tucson, Arizona Background & Aims: Little information is available injury.2–5 Attention recently has shifted to patient symp- about the efficacy of proton pump inhibitors (PPIs) in toms rather than endoscopically verified esophageal mu- patients with nonerosive reflux disease (NERD). We aimedto synthesize available data anddetermine the In practice, decisions about diagnosis and clinical in- effectiveness of PPIs on symptom resolution in patients terventions for GERD often are based on symptomology, with NERD. Methods: A systematic review of the litera- with heartburn and acid regurgitation as 2 of the defin- ture identified studies reporting the effects of PPIs inpatients with NERD. Heartburn resolution data were ing characteristics. Diagnostic clinical confirmation with pooledacross studies. The effectiveness of PPI therapy upper endoscopy or 24-hour esophageal pH monitoring in inducing complete heartburn resolution was com- is invasive, costly, and may not be readily available for paredin patients with NERD vs. erosive esophagitis (EE).
Results: Seven trials evaluating heartburn resolution in The use of proton pump inhibitors (PPIs) is becoming NERD were identified. Higher proportions of patients more common because mounting evidence indicates their reportedachieving sufficient heartburn resolution com- superior efficacy in both NERD and erosive esophagitis paredwith complete heartburn resolution. The effect of (EE).6 Despite available information pertaining to the PPIs on sufficient heartburn resolution was observed efficacy of PPI therapy in NERD, little has been done to sooner than was complete heartburn resolution. Thera- synthesize existing data with respect to different PPI peutic gain of PPI therapy over placebo rangedfrom30% to 35% for sufficient heartburn control andfrom therapeutic regimens.7 Clinical efficacy measures for EE 25% to 30% for complete heartburn control. Pooled therapies commonly focus on acute healing and mainte- response rates at 4 weeks were significantly higher for nance of healing using upper-gastrointestinal endoscopy.
patients with EE comparedwith NERD (56% vs. 37%, Much less information pertaining to symptomatic reso- P < 0.0001). Conclusions: PPIs provide a more modest
lution in EE patients can be found. Most clinical end therapeutic gain in patients with NERD as compared points in NERD are based on less-verifiable clinical with those with EE. A trendin increasedtherapeutic gain outcomes such as symptom severity, frequency, and in- for NERD patients was shown throughout the 4 weeks, tensity. Furthermore, outcome assessment in NERD is suggesting that 4 weeks of follow-up evaluation may be complicated by the lack of clear disease definition and insufficient to show full therapeutic gain in this patient inconsistent definition of symptom relief.8 The objective of this study was to perform a systematic review of the literature and to synthesize all available Gastroesophagealrefluxdisease(GERD)isacommon data on the effectiveness of PPI therapy in resolving disorder characterized by heartburn and acid regur- symptoms in NERD patients. The secondary objective gitation, with or without the presence of esophagealmucosal damage.1 Patients with nonerosive reflux disease Abbreviations used in this paper: CI, confidence interval; EE, erosive (NERD) may experience severe and even debilitating esophagitis; GERD, gastroesophageal reflux disease; NERD, nonero- heartburn episodes. In fact, studies have shown that sive reflux disease; PPI, proton pump inhibitor.
2004 by the American Gastroenterological Association GERD symptom severity, frequency, and intensity are correlated poorly with degree of esophageal mucosal EFFECT OF PPI USE IN PATIENTS WITH NERD 657 was to compare the effectiveness of PPI therapy between Table 2. Classification of Esophageal Mucosal Involvement patients with NERD and those with EE.
a) Endoscopically negative and erythema and friabilityb) Erythema and friability A review of the medical literature was conducted to b) Endoscopically negative and erythema and friability identify trials evaluating the effect of PPIs in patients with classic symptoms of GERD and negative upper endoscopy.
Medline and HealthStar databases were searched for Englishlanguage articles published between 1980 and 2002. Medicalsubject headings for the search included: proton pump inhib- tation, daytime heartburn, nighttime heartburn, and measures itor (PPI), esophagitis, gastroesophageal reflux, rabeprazole, of symptom severity. Information on available outcome mea- omeprazole, lansoprazole, esomeprazole, and pantoprazole. We sures was abstracted for all available periods.
searched the Food and Drug Administration web site for Applying the definitions of complete and sufficient heart- pertinent reports concerning PPI treatment indications for burn resolution as described later (see the Complete Versus NERD and abstracts presented during the American College Sufficient Resolution Over Time section), we abstracted the of Gastroenterology and the Digestive Disease Week confer- proportions of patients meeting these definitions from the ences between 1999 and 2001. Additional articles were iden- selected articles based on intent-to-treat results. Data were tified through a hand search of references from relevant articles available on the following combinations: complete heartburn and from physicians and researchers in the field.
resolution, sufficient heartburn resolution, complete heartburn Titles, abstracts, and articles were reviewed serially against maintenance, and sufficient heartburn maintenance.
explicit exclusion criteria. Two reviewers conducted indepen- dent appraisals and resolved disagreements by consensus. In-terrater reliability was tested on a 10% sample, with ␬ val- In accepted manuscripts, the study population was identified as having NERD or mild nonerosive GERD based Articles were rejected at the title stage if they did not on endoscopic results. A majority of investigators classified pertain to PPI monotherapy or a condition of interest (i.e., patients based on the grade of EE. A few investigators did not esophagitis, acid reflux, heartburn, GERD, NERD, gastric or use a standard esophageal grading scheme but provided esophageal acid), or were classified as reviews, case reports, enough information to determine if patients were endoscopy editorials, letters, or meta-analyses. Similar criteria were ap- negative, had erythema or friability, and/or had erosions. The plied to abstracts, with the additional requirement of data on most common classification schemes were the Savary–Miller, NERD, on esophagitis grade 0 and/or 1, and on outcomes of the Hetzel–Dent, and the Los Angeles, along with their mod- interest. The full text of selected articles were reviewed and rejected if they did not report primary results from a random- Criteria varied across the classification systems (Table 1). For ized trial of symptomatic resolution with acute PPI treatment, example, although patients without erythema or friability or failed to include adults age 18 or older with esophagitis would receive no grade using the Savary–Miller or Los Angeles systems, they would be assigned a grade 0 using the Hetzel–Dent system. Likewise, patients with erythema or friability are assigned a grade I using the Savary–Miller, a grade 1 using the Abstracted study data included: location, design, de- Hetzel–Dent, and no grade using the Los Angeles system.
scription, multi- or single-center setting, endoscopic grade and Another issue in comparing patient populations was that some definition of included patients, classification system, treatment studies included only patients with negative endoscopy, others goal, and symptom severity. Data pertaining to reported with erythema or friability only, and still others included symptom occurrence, frequency, and severity also were col- patients with negative endoscopy, erythema, or friability, but lected. Symptom measures included heartburn, acid regurgi- Because of the different definitions of endoscopic findings and study populations, we developed a system for classifying Table 1. Classification of Esophageal Mucosal Involvement studies based on reported grades from various classification systems (Table 2). Treatment comparisons were separated into3 groups. Group 1 included only endoscopically negative patients. Group 2 was more flexible, allowing for erythema and friability but without erosions; this group included mixed patient populations, some with no endoscopic findings and others with erythema and friability. Group 3 combined all CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 8 studies of patients with negative endoscopy, negative endos- indications for symptomatic GERD or NERD. Our copy or erythema and friability, or erythema and friability search of the American College of Gastroenterology and Digestive Disease Week databases resulted in 59 ab- stracts for review. None presented data that were not also available in the published literature or Food and DrugAdministration reports. Seven trials10–16 were included Acute heartburn resolution was defined as either com- plete or sufficient. Complete resolution was defined as no Of the 7 clinical trials evaluating heartburn resolu- heartburn during the preceding 7 days. Studies also measured tion, there were 2 rabeprazole studies,15,16 2 esomepra- sufficient (or satisfactory) heartburn resolution, defined as less zole studies,13,14 and 3 omeprazole studies.10–12 All 7 than 1 day of moderate heartburn during the preceding 7 days studies evaluated PPI therapy vs. placebo in patients with NERD. A total of 1854 patients were evaluated, with the smallest study including 123 patients16 and theremaining studies including over 200 patients each. Al- We focused our analysis on placebo-controlled trials.
though all studies allowed for a run-in period without Estimates of the pooled treatment effect were calculated for therisk difference (the proportion with heartburn resolution in the PPI or histamine-2 receptor antagonists, only one treated group minus the proportion with heartburn resolution study10 excluded patients with any previous use.
in the placebo group) using Bayesian modeling. The risk Five10,12–15 of the 7 studies presented data with multiple difference captures the therapeutic gain or effect of treatment treatment arms. A total of 12 treatment arms and 7 placebo arms were identified. The most frequent drug- We stratified available treatment arms by the following dose combination examined was omeprazole 20 mg (3 parameters: endoscopic grade of study population (Table 2), treatment arms). Two treatment arms were available for duration of PPI administration, and complete or sufficient omeprazole 10 mg, esomeprazole 20 mg, esomeprazole control of heartburn. Differences across grades, time periods, 40 mg, and rabeprazole 20 mg, with only one treatment and definitions of heartburn control were evaluated.
Table 4 shows all sufficient and complete heartburn resolution outcomes at all time periods across all 7 We compared the effectiveness of PPI treatment on studies. The majority of studies reported heartburn res- complete heartburn resolution in patients with NERD vs. EE.
olution rates for 2 time periods. Thus, our analysis We conducted a systematic review of the published literature explored the risk difference (therapeutic gain) for the and Food and Drug Administration reports pertaining to theeffectiveness of PPI therapy in patients with endoscopically confirmed EE (Hetzel–Dent, grades 2– 4).9 Strict criteria were Complete Versus Sufficient Resolution Over used to identify relevant placebo-controlled articles of PPI therapy in EE. Results were pooled in a similar manner to theNERD estimates.
Group 1: endoscopically negative subjects. Two studies10,11 provided data on heartburn resolution for endoscopically negative patients (Figure 2). The thera-peutic gain of PPI treatment over placebo for sufficient resolution at the 2-week assessment was 0.29 (95% The search identified 1169 references published confidence interval [CI ], 0.21– 0.37). At 4 weeks, the between January 1980 and January 2002. Figure 1 shows therapeutic gain increased to 0.34 (95% CI, 0.28 – 0.41).
the inclusion and exclusion of articles at different points No significant difference was observed between the ther- during the review process. We accepted 363 titles for apeutic gains at 2 and 4 weeks (P ϭ 0.42). No measure- further screening and reviewed their abstracts, and re- ments of complete resolution were available at 2 weeks.
viewed the full text of 107 articles.
At the 4-week assessment, the therapeutic gain of PPI Three published articles met inclusion criteria, pre- treatment over placebo was 0.25 (95% CI, 0.18 – 0.31).
senting results from randomized placebo-controlled trials In comparing therapeutic gains at 4 weeks, a higher with sufficient information for estimating the effects of proportion of treated vs. placebo patients achieved suffi- acute treatment on symptomatic GERD or NERD. Our cient resolution compared with achieving complete res- search of the Food and Drug Administration web site olution. This difference between therapeutic gains was yielded 4 additional reports concerning PPI treatment not statistically significant (P ϭ 0.14).
EFFECT OF PPI USE IN PATIENTS WITH NERD 659 Figure 1. Flowchart of the sys-tematic review for NERD andsymptomatic GERD.
Group 2: patients with negative endoscopy and the therapeutic gain of PPI treatment was 0.26 (95% CI, erythema andfriability, or erythema andfriability alone.
0.23– 0.30). Although therapeutic gains increased over Five studies12–16 contributed data on heartburn resolu- time, no significant difference was observed between the tion for our group 2 patient population (Figure 3). For assessments (1-week vs. 2-week, P ϭ 0.58; 2-week vs.
sufficient resolution, the therapeutic gain of PPI treat- ment over placebo at the 2-week measurement was 0.35 When therapeutic gains were compared within time (95% CI, 0.26 – 0.44). At the 4-week measurement, the periods, a higher proportion of group 2 patients reported therapeutic gain decreased to 0.29 (95% CI, 0.19 – 0.39).
achieving sufficient resolution compared with achieving The difference between the therapeutic gains at the 2- complete resolution. The therapeutic gain for sufficient and 4-week measurements was nonsignificant (P ϭ resolution was significantly different from the therapeu- tic gain for complete resolution at 2 weeks (P ϭ 0.04).
For complete resolution, therapeutic gains were calcu- At 4 weeks, the gap between the therapeutic gains lated for the 1-, 2-, and 4-week time periods. The decreased, resulting in a nonsignificant difference be- therapeutic gain of PPI treatment over placebo at 1 week tween the 2 measurements (P ϭ 0.69).
was 0.14 (95% CI, 0.10 – 0.17). An increase in therapeu- Group 3: patients with negative endoscopy, neg- tic gain was observed between the 1- and 2-week assess- oscopy anderythema andfriability, or ery- ments, with the therapeutic gain at 2 weeks calculated as thema andfriability alone. Results are consistent when 0.22 (95% CI, 0.19 – 0.25). For the 4-week assessment, data for patients in both groups 1 and 2 are combined CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 8 Table 3. Identified Placebo-Controlled Trials of NERD Patients OME10, omeprazole 10 mg; OME20, omeprazole 20 mg; ESO20, esomeprazole 10 mg; ESO40, esomeprazole 20 mg; FDA, Food and DrugAdministration; RAB10, rabeprazole 10 mg; RAB20, rabeprazole 20 mg.
aThe JADAD score is a way of assessing the quality of reports involving randomized controlled clinical trials.30 (Figure 4). For sufficient resolution, the therapeutic gain Again, no significant difference was observed between of PPI treatment over placebo at 2 weeks was 0.32 (95% the therapeutic gains at 2 and 4 weeks (P ϭ 0.25).
CI, 0.26 – 0.38). At the 4-week measurement, the ther- Within time periods, higher proportions of patients apeutic gain was 0.33 (95% CI, 0.27– 0.38). No signif- achieved sufficient resolution as opposed to complete icant difference was detected between the therapeutic resolution. At the 2-week measurements, the difference between therapeutic gains was significant (P Ͻ 0.05).
Results for complete resolution mirrored the results The difference between therapeutic gains decreased at the for group 2 patients at the 1-week (therapeutic gain ϭ 4-week measurement, and was not significant (P ϭ 0.14; 95% CI, 0.10 – 0.17) and 2-week (therapeutic gain ϭ 0.22; 95% CI, 0.19 – 0.25) assessments. The Summary. In general, the therapeutic gain of PPI therapeutic gain of PPI treatment for complete heartburn treatment over placebo ranged from 30%–35% for suf- resolution at 4 weeks was 0.26 (95% CI, 0.23– 0.29).
ficient heartburn control and 25%–30% for complete Table 4. Heartburn Measures Available for All Studies OME20, omeprazole 20 mg; OME10, omeprazole 10 mg; ESO40, esomeprazole 20 mg; ESO20, esomeprazole 10 mg; RAB20, rabeprazole 20mg; RAB10, rabeprazole 10 mg.
EFFECT OF PPI USE IN PATIENTS WITH NERD 661 Figure 2. Treatment comparison with placebo for group 1 (endoscopynegative) patients: pooled difference (treatment Ϫ placebo) in propor- Figure 4. Treatment comparison with placebo for all studies (group tion with heartburn resolution by time. Œ, Sufficient; ■, complete.
3): pooled difference (treatment Ϫ placebo) in proportion with heart-burn resolution by time. Œ, Sufficient; ■, complete.
heartburn control. In other words, PPIs successfullytreated between one quarter and one third of the popu- therapeutic gain for complete resolution for both group lation. This held true for pooled estimates across all 2 patients (endoscopically negative and erythema and treatments. Among all grades and time periods, PPIs friability) and for all patients combined (group 3) (P ϭ resulted in higher proportions of patients achieving suf- 0.04 and P ϭ 0.05, respectively). Over time, the thera- ficient resolution compared with complete resolution.
peutic gain for complete resolution increased. However, The effects of PPIs for sufficient resolution appeared the difference between sufficient and complete resolution sooner (approximately 2 weeks) compared with complete at 4 weeks was not significant (P ϭ 0.14, P ϭ 0.69, and resolution (4 weeks). At 2 weeks, the therapeutic gain for P ϭ 0.12 for groups 1, 2, and 3, respectively).
sufficient resolution was significantly different from the Comparison of Heartburn Resolution inNERD andEE We compared the symptomatic response rates of patients with NERD with those diagnosed with EE.
Although endoscopic healing rates are reported com-monly in EE studies, symptomatic response to PPI ther-apy is not reported commonly. When reported, 4- and8-week data usually are presented. Thus, across theNERD and EE literature, only 4-week symptomaticresponse rates were reported for each population. As wasperformed for NERD, the combined effect of PPI ther-apy on the symptomatic response rate at 4 weeks waspooled across treatment strategies and dosages for EEpatients.
A systematic literature review of symptomatic healing rates associated with PPI therapy in patients with EEidentified 2 randomized controlled trials.17,18 Both wereplacebo-controlled and contributed multiple treatment Figure 3. Treatment comparison with placebo for group 2 (negative arms of rabeprazole and pantoprazole, respectively. Com- endoscopy and erythema and friability) patients: pooled difference plete resolution of heartburn was the outcome measure (treatment Ϫ placebo) in proportion with heartburn resolution by time.
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 8 Table 5. Symptomatic Response Rates and Therapeutic Gains at 4 Weeks Among Patients With NERD and EE Table 5 displays the pooled 4-week symptomatic re- The lower symptom response rate to PPI treatment in sponse rates and therapeutic gains for patients with NERD patients as compared with patients with EE NERD and EE. The 4-week symptomatic response rate likely is owing to the different subgroups of patients was significantly higher for patients with EE compared identified as having NERD. A recent study by Martinez with NERD (56% vs. 37%, respectively; P Ͻ 0.0001), et al.28 showed that 50% of the patients with NERD whereas the placebo symptomatic response rates were have normal esophageal acid exposure. Of these patients similar between the 2 groups (9.5% vs. 7.5%, respec- with functional heartburn, 37% showed a close correla- tively; P Ͼ 0.05). Thus, the therapeutic gain of PPI tion between GERD symptoms and acid reflux events, therapy over placebo was more than 75% higher among and the rest (63%) were likely to experience heartburn EE patients (48% for EE vs. 27% for NERD).
caused by non–acid-related stimuli. These findings areconsistent with the view that patients with NERD are a heterogeneous group with different symptom causes and NERD is the most common type of GERD that thus different therapeutic responses to PPIs.
community-based physicians encounter. Because of the The analysis also suggests that owing to the signifi- lack of esophageal mucosal injury, treatment of NERD is cantly lower symptom response rates to PPIs in patients based commonly on a step-up approach. However, ther- with NERD compared with those with EE, the step-up apeutic studies in NERD have shown that PPIs are approach has little merit in NERD patients. Adding the superior to histamine-2 receptor antagonists or promo- studies that showed a very limited (30%) symptom tility compounds.19–21 In this study, we compared the response rate of NERD patients to histamine-2 receptor effectiveness of PPI therapy with that of placebo for antagonists, a step-in approach with a PPI should be the symptom resolution at different time periods in NERD appropriate therapeutic intervention for NERD.19,23 patients, using published literature. Because of the lack An important finding of our study is that the pooled of esophageal mucosal injury, almost all NERD studies therapeutic gain of PPI treatment for complete heartburn limited the treatment period to 4 weeks.12,22–25 Last, resolution increased from the 1-week to 2-week assess- symptom response to PPI was compared between pa- ment and again at the 4-week time period (0.14 3 0.22 3 0.26). Our study showed the need to reconsider Several important points stand out. First, higher pro- the 4-week study design in NERD patients undergoing portions of NERD patients reported attaining sufficient therapeutic intervention because some patients with heartburn resolution as compared with attaining com- NERD may have a longer lag time to symptom response.
plete heartburn resolution. This was apparent across all This study systematically assessed and synthesized the grades of NERD and across all time periods. In addition, literature on the effectiveness of PPI therapy in the PPI treatment led to sufficient heartburn resolution, NERD population and compared symptomatic response with the effects observed at the 2-week time period; the rates on PPI therapy in both NERD and EE populations effects of complete heartburn resolution were observed at in the same time frame. Therapeutic gain (response to the 4-week period. Finally, symptomatic response rate at treatment minus response to placebo) allows for the 4 weeks was significantly higher for patients with EE combination of data across a variety of studies and pro- vides clinicians with a clinically meaningful range of Our analysis suggests that to improve the outcome of responses they should expect with therapy.
therapeutic studies in NERD, more modest clinical end A few limitations were inherent to this study. One was points should be considered. For many NERD patients, the lack of a standard definition for NERD and the use sufficient heartburn control may be a satisfactory thera- of different classifications for GERD. Another limitation peutic outcome. Because NERD patients rarely develop was the lack of data points included in this analysis. This EE, physicians can accept sufficient heartburn control as may have resulted from the selected parameters used in the systematic review or the definitions of NERD used EFFECT OF PPI USE IN PATIENTS WITH NERD 663 by the researchers. A third limitation was the lack of 9. Hetzel DJ, Dent J, Reed WD, Narielvala FM, Mackinnon M, Mc- comparable end points among included studies. Some Carthy JH, Mitchell B, Beveridge BR, Laurence BH, Gibson GG.
Healing and relapse of severe peptic esophagitis after treatment studies reported complete daytime and/or nighttime with omeprazole. Gastroenterology 1988;95:903–912.
heartburn resolution, whereas others reported complete 10. Lind T, Havelund T, Carlsson R, Anker-Hansen O, Glise H, Hern- resolution of global symptoms. A hierarchy was created a qvist H, Junghard O, Lauritsen K, Lundell L, Pedersen SA, Stub-berod A. Heartburn without oesophagitis: efficacy of omeprazole priori to account for the choice of end points used in the therapy and features determining therapeutic response. Scand J analysis when a study reported more than one end point.
Last, we found low placebo symptom response rates for 11. Hatlebakk JG. Heartburn treatment in primary care: randomised, double blind study for 8 weeks. BMJ 1999;319:550 –553.
PPI therapy in patients with NERD and with EE. Low 12. Richter JE, Peura D, Benjamin SB, Joelsson B, Whipple J. Efficacy placebo response rates may be the result of assessing the of omeprazole for the treatment of symptomatic acid reflux dis- rigorous outcome of complete symptom resolution. In ease without esophagitis. Arch Intern Med 2000;160:1810 – addition, the low placebo response rates could be the 13. AstraZeneca. A comparative efficacy and safety study of H result of the time period assessed because patient expec- 199/18 (20mg), H 199/18 (40mg) vs. placebo in study subjects tations of treatment response may vary by time.
with symptomatic GERD. FDA Report 21-153-225. 1999.
In summary, as expected, NERD patients are likely to 14. AstraZeneca. A comparative efficacy and safety study of H 199/18 (20mg), H 199/18 (40mg) vs placebo in study subjects report sufficient heartburn resolution more often than with symptomatic GERD. FDA Report 21-153-226. 1999.
complete heartburn resolution and achieve sufficient res- 15. Janssen. A placebo-controlled trial of rabeprazole tablets, 10mg olution sooner than complete resolution. Symptom re- QD or 20mg QD, in the treatment of subjects with symptoms ofchronic Gastroesophageal Reflux Disease (GERD). FDA Report sponse rate was significantly lower in NERD patients compared with those with EE receiving standard-dose 16. Janssen. A double-blind, placebo-controlled trial of rabeprazole PPI. Most importantly, during the first 4 weeks of tablets, 20 mg once daily, in the treatment of subjects without treatment, patients with NERD showed increased ther- erosive esophagitis and who have symptoms of chronic gastro-esophageal reflux disease (GERD). FDA Report 20-973-RAB- apeutic gain for complete heartburn resolution (from 0.14 to 0.26), suggesting that the trend of symptom 17. Cloud ML, Enas N, Humphries TJ, Bassion S. Rabeprazole in improvement may well continue into the second month treatment of acid peptic diseases: results of three placebo-controlled dose-response clinical trials in duodenal ulcer, gastric of therapy. Future study designs in NERD patients ulcer, and gastroesophageal reflux disease (GERD). The Rabe- should not be limited to 4 weeks, and randomized pla- prazole Study Group. Dig Dis Sci 1998;43:993–1000.
cebo-controlled studies comparing various PPI therapies 18. Richter JE, Bochenek W. Oral pantoprazole for erosive esoph- for sufficient and complete heartburn resolution in pa- agitis: a placebo-controlled, randomized clinical trial. Panto-prazole US GERD Study Group. Am J Gastroenterol 2000;95: tients with NERD should be conducted.
19. Sabesin SM, Berlin RG, Humphries TJ, Bradstreet DC, Walton- Bowen KL, Zaidi S. Famotidine relieves symptoms of gastro- esophageal reflux disease and heals erosions and ulcerations.
1. Kahrilas PJ. Diagnosis of symptomatic gastroesophageal reflux Results of a multicenter, placebo-controlled, dose-ranging study.
disease. Am J Gastroenterol 2003;98:S15–S23.
USA Merck Gastroesophageal Reflux Disease Study Group [see 2. Green JRB. Is there such an entity as mild oesophagitis? Eur comments]. Arch Intern Med 1991;151:2394 –2400.
20. Kahrilas PJ, Fennerty MB, Joelsson B. High- versus standard- 3. Laursen LS, Havelund T, Bondesen S, Hansen J, Sanchez G, dose ranitidine for control of heartburn in poorly responsive acid Sebelin E, Fenger C, Lauritsen K. Omeprazole in the long-term reflux disease: a prospective, controlled trial. Am J Gastroenterol treatment of gastro-oesophageal reflux disease. A double-blind randomized dose-finding study. Scand J Gastroenterol 1995;30: 21. Hatlebakk JG, Johnsson F, Vilien M, Carling L, Wetterhus S, Thogersen T. The effect of cisapride in maintaining symptomatic 4. Kuster E, Ros E, Toledo-Pimentel V, Pujol A, Bordas JM, Grande L, remission in patients with gastro-oesophageal reflux disease.
Pera C. Predictive factors of the long term outcome in gastro- Scand J Gastroenterol 1997;32:1100 –1106.
oesophageal reflux disease: six year follow up of 107 patients.
22. Bate CM, Griffin SM, Keeling PW, Axon AT, Dronfield MW, Chap- man RW, O’Donoghue D, Calam J, Crowe J, Mountfords RA, Watts 5. Trimble KC, Douglas S, Pryde A, Heading RC. Clinical character- DA, Taylor MD, Richardson PD. Reflux symptom relief with ome- istics and natural history of symptomatic but not excess gastro- prazole in patients without unequivocal oesophagitis. Aliment esophageal reflux. Dig Dis Sci 1995;40:1098 –1104.
6. Chiba N, De Gara CJ, Wilkinson JM, Hunt RH. Speed of healing 23. Venables TL, Newland RD, Patel AC, Hole J, Wilcock C, Turbitt and symptom relief in grade II to IV gastroesophageal reflux disease: ML. Omeprazole 10 milligrams once daily, omeprazole 20 milli- a meta-analysis. Gastroenterology 1997;112:1798 –1810.
grams once daily, or ranitidine 150 milligrams twice daily, eval- 7. Fass R, Fennerty MB, Vakil N. Nonerosive reflux disease— current uated as initial therapy for the relief of symptoms of gastro- concepts and dilemmas. Am J Gastroenterol 2001;96:303–314.
oesophageal reflux disease in general practice. Scand J 8. Fass R. Epidemiology and pathophysiology of symptomatic gastroesophageal reflux disease. Am J Gastroenterol 2003; 24. Richter JE, Campbell DR, Kahrilas PJ, Huang B, Fludas C. Lanso- prazole compared with ranitidine for the treatment of nonerosive CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 8 gastroesophageal reflux disease. Arch Intern Med 2000;160: 29. Harding SM, Richter JE, Guzzo MR, Schan CA, Alexander RW, Bradley LA. Asthma and gastroesophageal reflux: acid suppres- 25. Galmiche JP, Barthelemy P, Hamelin B. Treating the symptoms of sive therapy improves asthma outcome. Am J Med 1996;100: gastro-oesophageal reflux disease: a double-blind comparison of omeprazole and cisapride. Aliment Pharmacol Ther 1997;11: 30. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Ga- vaghan DJ, McQuay HJ. Assessing the quality of reports of ran- 26. Isolauri J, Luostarinen M, Isolauri E, Reinikainen P, Viljakka M, domized clinical trials: is blinding necessary? Control Clin Trials Keyrilainen O. Natural course of gastroesophageal reflux dis- ease: 17-22 year follow-up of 60 patients. Am J Gastroenterol1997;92:37– 41.
27. Fass R, Ofman JJ. Gastroesophageal reflux disease—should we adopt a new conceptual framework? Am J Gastroenterol 2002; Address requests for reprints to: Bonnie B. Dean, Ph.D., M.P.H., 9100 Wilshire Boulevard, East Tower, Suite 655, Beverly Hills, Califor- 28. Martinez SD, Malagon IB, Garewal HS, Cui H, Fass R. Non-erosive nia 90212. e-mail: bdean@cerner.com.
reflux disease (NERD)—acid reflux and symptom patterns. Ali- Supportedin part by an unrestrictedresearch grant from Janssen ment Pharmacol Ther 2003;17:537–545.
Pharmaceutica Inc. (Titusville, NJ) andEisai Inc. (Teaneck, NJ).

Source: http://www.sandhillsci.com/pdf/Effectiveness-of-PPI-1.pdf