Treatment Options for Gastroesophageal Reflux Disease: Scientific Review UC Davis Center for Health Services Research in Primary Care June 2004 Treatment Options for Gastroesophageal Reflux Disease: Scientific Review Prepared for: CALIFORNIA HEALTHCARE FOUNDATION Prepared by: UC Davis Center for Health Services Research in Primary Care Authors: John K. Siepler, Pharm.D., B.C.N.S.P., F.C.S.H.P. Nutrition Support Pharmacist Department of Pharmaceutical Services University of California, Davis, Medical Center Clinical Professor Department of Clinical Pharmacy University of California, San Francisco, School of Pharmacy Walter L. Trudeau, M.D. Professor Emeritus of Medicine Division of Gastroenterology University of California, Davis, School of Medicine June 2004 Acknowledgments About the Authors The authors would like to thank the following individuals for reviewing these materials: Rosemary R. Berardi, Pharm.D., F.C.C.P., F.A.S.H.P Professor of Pharmacy, College of ., Pharmacy, University of Michigan; and Dennis McCarthy M.D., Professor of Medicine, Chief of Gastroenterology, Veterans Administration Medical Center. The authors also thank Diane Romac, Pharm.D. and Jeff King, Pharm.D., for scientific editing. About the Foundation The California HealthCare Foundation, based in Oakland, is an independent philanthropy committed to improving California’s health care delivery and financing systems. Additional copies of this report and other publica- tions can be obtained by visiting www.chcf.org. ISBN 1-932064-72-9 Copyright © 2004 California HealthCare Foundation Contents 4 Overview of the Prescription Drug Information Project 6 I. Introduction to Gastroesophageal Reflux Disease 7 II. Therapeutic Management Nondrug Treatments Pharmacotherapy Acid Suppression Symptom Control Healing of Esophagitis Prevention of Recurrence of Esophagitis Prevention of Complications Adverse Effects and Drug Interactions Endoscopic and Surgical Treatment Cost Considerations 16 III. Clinical Conclusions 17 IV. Appendices Table 1: Proton Pump Inhibitors and Their Indicators Table 2: Cost Comparisions of Acid-Reducing Agents 19 Endnotes Overview of the Prescription Drug Information Project AS PRESCRIPTION DRUG BENEFITS SHRINK EITHER because of annual spending caps, increased copayments, limited formularies, or combinations of these scenarios, out-of-pocket expenses for prescription drugs are increasing. Consumers are faced with choices and potential clinical tradeoffs concerning their medications. The Prescription Drug Information Project (PDIP) is a collab- orative venture between the University of California (UC), with UC Davis as the lead site, and the California HealthCare Foundation. The goal of the PDIP is to support clinicians and patients in California in their day-to-day decisions about which drugs to prescribe or take. The governing principle is that accurate, understandable information on effectiveness, side effects, and costs will help clinicians and patients to select the best drug or treatment at the best price. The PDIP focuses on drugs that are currently promoted directly to the public. The United States has witnessed a steep increase in direct-to-consumer (DTC) advertising of prescrip- tion drugs, from $791 million in 1997 to nearly $2.5 billion in 2001.1 Fewer than 40 DTC drugs account for most of these expenditures. Many physicians, managed care organizations, and health policy analysts believe that DTC advertisements provide incomplete or inaccurate information to consumers, which nonetheless contribute to increased consumer demand for DTC drugs. Furthermore, there is widespread concern that consumer demand translates into inappropriate and unneces- sary prescribing of DTC drugs. On the other hand, propo- nents of DTC advertising argue that advertisements raise awareness of undertreated conditions and encourage more dialog between patients and physicians. Still, proponents and critics agree that making accurate information accessible to both consumers and clinicians is never a bad thing. To this end, the PDIP has two components: information retrieval and dissemination. The information retrieval process is led by UC Davis, whereas the dissemination campaign is coordinated by the California HealthCare Foundation (CHCF). In the first round of information retrieval, the project focuses on treatments for gastroesophageal reflux disease (GERD), hypercholesterolemia, and osteoarthritis (OA). In a subsequent phase, treatments for depression, allergic rhinitis, and asthma will be undertaken. 4 | CALIFORNIA HEALTHCARE FOUNDATION To develop an appropriate evidence base for the The UC Davis reports were reviewed by two planned informational campaign, two approaches outside national experts and then reviewed by are utilized. The first uses publicly available drug a formal expert panel (Scientific Advisory class reviews produced by the Drug Effectiveness Committee) comprised of seven distinguished Review Project (DERP). These systematic litera- faculty from five UC health sciences campuses. ture reviews are conducted by Evidence-based All experts submitted conflict-of-interest disclo- Practice Centers (EPC) with oversight and coor- sure statements that are on file at UC Davis. The dination from the Oregon EPC. To date, the reports were revised based on input from the DERP has prepared nine such reports on topics Scientific Advisory Committee. Formal delibera- ranging from skeletal muscle relaxants to estrogen tions by the Committee, including their ratings preparations. The three reports that are most rel- of the validity of key findings, are available else- evant to the first round of information retrieval where (CHCF Interim Report, 12/12/2003). are entitled: s Drug Class Report on Proton Pump Inhibitors (updated April 2003), s Drug Class Report on HMG-CoA Reductase Inhibitors (Statins) (updated July 2003), and s Drug Class Report on Cyclooxygenase-2 (COX-2) Inhibitors and Nonsteroidal Anti- inflammatory Drugs (NSAIDs) (updated May 2003). To address treatment alternatives not covered by the DERP reports, teams of pharmacists and physicians at UC Davis prepared supplemental reports on treatment options for each of the three target conditions: GERD, hypercholesterolemia, and osteoarthritis. Each UC Davis writing team consisted of one academic pharmacist and one physician-specialist. The teams were instructed to compose a concise, evidence-based synopsis of the treatment options for a particular condition, including pharmacologic and nonpharmacologic options. They were also asked to provide infor- mation on the costs of drug therapy, focusing on average wholesale and Medicare cash prices. The teams consisted of John Siepler, Pharm.D., and Walter Trudeau, M.D. (GERD); Robert Mowers, Pharm.D., and Thomas Balsbaugh, M.D. (hypercholesterolemia); and Robert Mowers, Pharm.D., and Gurtej Cheema, M.D. (OA). Treatment Options for GERD: Scientific Review | 5 I. Introduction to Gastroesophageal Reflux Disease ABOUT 20% OF THE U.S. POPULATION OVER the age of 18 years experiences heartburn or regurgitation at least once per week.2-4 When these symptoms occur for at least 3 months, reflux of gastric contents into the esophagus can be considered a disease. This review discusses the current treat- ment status of gastroesophageal reflux disease (GERD). Heartburn is a burning retrosternal or epigastric pain that radi- ates up into the chest, often occurring after meals or while lying down.4 If there is no relief from pain with oral antacids, the condition is usually not true heartburn and will typically not respond to any kind of acid-suppressive therapy. Functional heartburn, which is not relieved by eliminating acidity, is common. Reflux of gastric contents back into the esophagus is a reversal of their “usual” distal passage, and gas- troesophageal reflux is increased when gastric emptying is impaired. Regurgitation occurs when gastric contents are returned into the mouth, regardless of their acidity. About 1 in 10 patients with GERD suffers from regurgitation in the absence of heartburn. Other less common esophageal symp- toms include atypical chest pain and, rarely, a globus sensation (a “lump” in the throat that is unrelated to eating or swallow- ing). Other symptoms, such as bloating, dyspepsia, or vomit- ing, are usually predictive of a poor response to proton pump inhibitor (PPI) therapy. Morning nausea may occur in about 75% of patients with nocturnal GERD symptoms. There are also a considerable number of symptoms associated with effects of reflux on the pharynx, larynx, respiratory tract, middle ear, and teeth. Symptoms tend to be worse at night or in the recumbent position, and sleep disturbance, with consequently impaired quality of life, occurs in about one third of patients. Nocturnal symptoms include both classic heartburn and regur- gitation, as well as additional extraesophageal symptoms such as cough, hoarseness, dysphonia, sore throat, coughing or choking episodes that awaken patients from sleep, aspiration pneumonia, and exacerbations of wheezing or breathlessness in those with asthma, chronic bronchitis, or dyspnea from other causes (e.g., cystic or pulmonary fibrosis). Sleep disturbances include nightmares, snoring, sweating, apnea, or sleep depriva- tion resulting in sleepiness, headache, or impaired productivity the next day. There is a strong association between obstructive sleep apnea and episodic reflux, but there is little evidence to suggest that GERD plays a causative role. All of these prob- 6 | CALIFORNIA HEALTHCARE FOUNDATION lems may also occur during waking hours. The numerous factors that predispose to reflux, realization that atypical chest pain and many including a delay in stomach emptying,8 which is other extraesophageal symptoms are in fact most commonly due to hyperglycemia in other- attributable to GERD accounts for the rapid wise uncomplicated diabetes or to external influ- increase in PPI use by otorhinolaryngologists, ences on intraabdominal pressure, such as consti- pulmonologists, internists, and other specialists. pation, obesity, or pregnancy. At present, there are no medications that are effective in strength- Although most cases of GERD can be ening the LES to prevent gastroesophageal reflux. approached deliberately, certain “alarm features” Unless the patient has a history of peptic ulcer, should trigger an expedited invasive approach, testing and treatment for Helicobacter pylori is not such as early endoscopy. Alarm symptoms helpful. In some patients with fundic gastritis, include dysphagia, odynophagia, weight loss, treatment may exacerbate GERD, and an emesis, or a laboratory finding of anemia or increase in the dose of PPI is then needed to blood in the stool. Dysphagia and weight loss control symptoms. may be symptoms of benign esophageal stricture or esophageal adenocarcinoma.5 The presence of Complications of GERD include esophageal any of these features contraindicates the use of ulcers, stricture, bleeding, perforation, and, rarely, empirical therapy.6 esophageal adenocarcinoma.5, 9 Avoiding compli- cations is an important goal in the management Esophagitis is an inflammation of the esophageal of GERD. The occurrence of stricture is uncom- mucosa. The condition is usually diagnosed by mon and is usually associated with long-term endoscopy, with the severity of the disease docu- esophagitis. Esophageal adenocarcinoma, mented by a formal classification system.4,7 although rare, is the most serious complication of However, endoscopic appearances do not corre- GERD and is more strongly associated with the late with the severity of symptoms, ease of heal- duration or frequency of heartburn, particularly ing, duration of therapy, or likelihood of relapse. nocturnal heartburn, than with Barrett’s esopha- More than half of patients with heartburn or gus. The incidence of esophageal cancer among regurgitation demonstrate no esophagitis on patients with Barrett’s esophagus is less than endoscopy, regardless of whether their symptoms 1 case per 100 patient-years.10 While patients are mild or severe. The current gold standard for with GERD and Barrett’s esophagus are at the diagnosis of reflux is 24-hour monitoring of increased risk of cancer, prospective studies have esophageal pH. Endoscopy measures the extent repeatedly failed to document the development and severity of tissue injury. The procedure may of Barrett’s esophagus as a complication of also show the presence of Barrett’s esophagus, GERD in patients who do not have Barrett’s which is associated with prolonged tissue expo- esophagus on initial endoscopy.9 Whether or not sure to gastric contents. treatment of Barrett’s esophagus improves the While patients with GERD usually have normal condition or reduces the incidence of gastric acid secretion, they also generally have esophageal adenocarcinoma, and whether more total time of esophageal exposure to acid. patients with Barrett’s esophagus who do not In addition to reduced lower esophageal sphinc- show symptoms should be treated with PPIs, ter (LES) pressure, frequent transient LES relax- are issues that remain controversial.3,9 ation, or the presence of a hiatus hernia, there are Treatment Options for GERD: Scientific Review | 7 In clinical practice, anyone with reflux symptoms for more than twice a week for 3 months, and whose quality of life has been adversely affected, is considered to have reflux disease that warrents formal treatment.3 The diagnosis is largely clini- cal, and proper initial treatment is empirical. However, for patients with alarm features or symptoms that do not resolve after appropriate treatment, referral for upper endoscopy is indi- cated. Although symptoms correlate poorly with the presence or absence of esophagitis, the need for treatment is based on symptoms. A policy requiring endoscopy prior to antireflux therapy is expensive, does not justify the withholding of PPIs, is not cost-effective, and has already been abandoned by the governments of Australia and Canada, as well as by many health maintenance organizations.11,12 8 | CALIFORNIA HEALTHCARE FOUNDATION II. Therapeutic Management THE GOALS OF TREATMENT INCLUDE RELIEF OF symptoms, healing of esophagitis and esophageal ulcers, and prevention of recurrence and complications of GERD. Symptom relief is most important to the patient.13,14 Symptoms are the primary reason the patient seeks physician help, and failure to relieve them leads to dissatisfaction. Patients over 50 years of age who have had symptoms for more than 5 to 7 years should have a once-in-a-lifetime endoscopy to screen for Barrett’s esophagus.9 If absent, no further endoscopy is needed unless a complication is suspected. Repeated endoscopy does not influence the management of the patient’s symptoms or therapy. Once a clinical diagnosis of GERD is made, treatment can begin.14 A number of therapeutic measures may be taken to reduce reflux, including changes in lifestyle or the use of proki- netic medications. These interventions may be very useful adjuncts to acid-suppressive therapy. However, the cornerstone of treatment is reduction of gastric acid output, which reduces esophageal acid reflux both by decreasing the volume of gastric secretion and by increasing esophageal pH.13-15 Nondrug Treatment Lifestyle changes attempt to reduce acid reflux by altering the risk factors that contribute to the occurrence, severity, and chronicity of GERD. Avoidance of obesity, excess alcohol intake, smoking, caffeine consumption, certain foods such as peppermint or chocolate, tight-fitting clothes, weight lifting, and recumbency are helpful in some cases. Elevating the head of the bed by 8 inches during sleep and avoiding large, fatty meals or lying down after meals helps most patients. Although weight loss is often advocated, one small randomized trial found no benefit.16 Most patients continue to have symptoms unless changes in lifestyle are combined with acid-suppressive therapy.3,15,17 There is insufficient evidence to support the use of alternative or herbal medications in the treatment of GERD. Treatment Options for GERD: Scientific Review | 9 Pharmacotherapy Metoclopramide is also associated with diarrhea Antacids and constipation, and its long-term use, in low or moderate doses, remains a treatment option Antacids are effective in increasing the intragas- mainly in patients with diabetic gastroparesis, tric pH. They are highly effective in relieving primarily as an adjunct to PPI therapy. Tegaserod heartburn rapidly, but these effects are usually (Zelnorm) is a 5-HT4 receptor partial agonist very short, lasting only 1 to 2 hours. Antacid use that is used for the management of irritable is limited by palatability, inconvenience, and side bowel syndrome. It also has prokinetic effects in effects. Antacids contain aluminum, magnesium, the upper gastrointestinal tract and its role in the or calcium salts, which may cause diarrhea (as treatment of GERD is currently being evaluated with magnesium), constipation (aluminum), or in clinical trials. milk-alkali syndrome (calcium carbonate). Phosphorous depletion and increased absorption Prescription H 2RAs of aluminum are of particular concern in certain groups of patients, such as those who are on dial- Detailed reviews of prescription H2RA use in ysis. Most experts feel that antacids have little GERD can be found elsewhere.22,23 (The results role in the primary treatment of esophagitis.17,18 of studies of low, once-daily, or nighttime-only These agents, however, are effective for the occa- dosing should not be used to guide clinical prac- sional relief of heartburn and may be used in tice as such dosing regimens are incorrect.) conjunction with PPIs to relieve occasional H2RAs reduce acid reflux by reducing gastric acid nighttime symptoms. production, and numerous studies have docu- mented their effectiveness in patients with mild- Over-the-counter Histamine H2-receptor to-moderate heartburn or esophagitis. Symptoms Antagonists are relieved in 70% to 80% of patients after 12 Over-the-counter (OTC) doses of histamine H2- weeks, and mild-to-moderate esophagitis heals in receptor antagonists (H2RAs) are available at 60% to 70% of cases. In patients with severe ero- about half of the prescription strength. Although sive esophagitis, efficacy is less impressive, with inexpensive and effective for the short-term relief healing in less than 50% of patients.22-25 Since of occasional heartburn, they are ineffective as GERD is often a chronic disease, long-term long-term therapy for more severe disease. OTC treatment is necessary in some patients to prevent H2RAs are not recommended in esophagitis heal- recurrence. Recurrence rates range from 75% to ing and their use should be limited to patients 92% if the drugs are stopped acutely. Treatment with occasional heartburn.13,15-19 should be tapered over 2 to 4 weeks (depending on the dose of H2RA, duration of therapy, and Prokinetics needs of the patient) and re-instituted only in Metoclopramide, at a dose of 10 mg four times those in whom symptoms recur. H2RAs can con- daily, is as effective as H2RAs in the treatment of trol symptoms in about 50% of patients with mild-to-moderate heartburn or esophagitis. recurrent symptoms. In randomized double-blind About 50% to 80% of patients with mild trials of maintenance drug therapy, 20% to 40% esophagitis heal after 12 weeks on metoclo- of those on placebo showed no recurrence of pramide. Patients with more severe esophagitis symptoms during 1 to 2 years of follow-up. usually have less satisfactory results.20,21 The efficacy, safety, and low cost of H2RAs make Metoclopramide cannot be advocated for long- them an appropriate option for patients with term use because of the high incidence of neu- mild-to-moderate GERD, especially when ropsychiatric side effects, including depression, resources are limited, thus reserving PPIs for parkinsonism, and tardive dyskinesia. those who do not respond. Still, the limited effi- 10 | CALIFORNIA HEALTHCARE FOUNDATION cacy of H2RAs in more severe disease makes trials involving patients with GERD have used them less attractive than PPIs in patients with the following daily doses: omeprazole 20 mg, troublesome symptoms or previous complica- lansoprazole 30 mg, rabeprazole 20 mg, panto- tions. Further, the use of some H2RAs, especially prazole 40 mg, or esomeprazole 40 mg. The cimetidine, can affect the mechanisms of actions practice of comparing unequal doses creates of other medications. almost insurmountable difficulties in evaluating the relative effects of this class of drugs. This PPIs review compares the effectiveness of PPIs in the PPIs are prodrugs that prevent acid secretion by treatment of GERD, focusing on available pre- binding to the enzyme H+/K+-ATPase (proton scription doses that have been used in trials. In pump) of the acid-generating cells of the stom- reaching conclusions about relative safety and ach (parietal cells). Inhibition of this enzyme effectiveness, we used the systematic review pro- hinders transport of the H+ ion across the cell duced by the DERP27 supplemented by addition- , membrane. In the absence of a stimulus to acid al articles from the authors’ own files. Outcomes secretion 30 to 45 minutes after the ingestion of of prime interest include the degree of acid sup- a PPI, parietal cells remain largely inactive. For pression, control of symptoms, and healing of this reason, PPIs are likely to be less effective esophagitis. There are insufficient data to com- when used together with drugs that suppress acid pare the effects of the various PPIs on prevent- secretion, such as H2RAs, prostaglandins, anti- ing symptomatic recurrences and complications. cholinergics, tricyclic antidepressants, and somatostatin analogs. There are currently five PPIs approved for use in the United States (Table Acid Suppression 1). The pharmacology of PPIs has been reviewed elsewhere.26 In brief, PPIs have short half-lives All PPIs suppress the secretion of gastric acid bet- and a variably prolonged duration of action (acid ter than do H2RAs. Acid secretion is measured as suppression for 13 to 46 hours), depending on the average pH in 24 hours or the percentage of the dose. This prolonged duration of acid sup- time that the intraesophageal pH is >4. However, pression is the result of slowly reversible binding most comparisons of PPIs have found their to the proton pump. Acid suppression with PPIs, effects on acid suppression to be similar. As there which also varies with the method of administra- are pharmacologic differences among the PPIs, tion, is consistently better than that associated one might expect differences in their abilities to with H2RAs. suppress acid. Indeed, some studies have found such differences, but variability in study design Several PPIs (e.g., omeprazole, lansoprazole, and (including the methods of measuring acid) and esomeprazole) are available in various dosage for- subject selection make comparisons difficult. mulations, whereas pantoprazole and rabeprazole are available only in one strength. Despite the DERP did not address the topic of acid suppres- availability of different strengths, most clinical sion directly. However, we found four relevant studies that did. Blum et al. found acid suppres- sion with omeprazole and lansoprazole to be bet- ter than with ranitidine.28 Pantoflickova and asso- ciates compared the acid suppression seen during the first 24 hours of therapy with omeprazole, Treatment Options for GERD: Scientific Review | 11 lansoprazole, rabeprazole, and pantoprazole, and Symptom Control found that intragastric pH and the duration that Most patients with GERD have good symptom the pH was >4.0 was the greatest with rabepra- control within the first few days of starting PPI zole.29 However, there were no differences among therapy; 80% experience relief of heartburn by the four PPIs by the fifth day. (Rapid onset of 8 weeks.13 Heartburn relief with PPIs is consis- action might be advantageous if “on-demand” tently superior to that obtained with H2RAs. therapy becomes an accepted method of treat- Several studies have compared symptom control ment.) Frazzoni and associates reported that lan- with the different PPIs. Castell et al.32 found that soprazole was more effective at “normalizing” heartburn relief in the first week of treatment esophageal acid exposure than was pantopra- was achieved by 11% more of patients who zole.30 The fact that all patients had good symp- were taking lansoprazole 30 mg than those tom control challenges the assumption that the taking omeprazole 20 mg (P <0.05). However, degree of acid suppression is a clinically relevant by 8 weeks the difference had decreased to 3% outcome. In the review by Hatlebakk,31 the con- (P <0.05). A review by Kromer et al.33 concluded clusion was that esomeprazole (40 mg QD) pro- that all PPIs are about equal on a per milligram duces greater acid suppression than does 30 mg basis, and all trials of unmatched dose compar- of lansoprazole, 40 mg of pantoprazole, and 20 isons are pharmacologically spurious. As asserted mg of rabeprazole. However, all of the studies by DERP,27 there are few studies comparing equal that examined acid suppression from esomepra- doses of PPIs, making conclusions about relative zole in that review were in abstract form. efficacy quite difficult. Nevertheless, the prepon- A major problem with most of the comparative derance of evidence suggests that equal doses of studies to date is the heterogeneity of patients different PPIs provide similar levels of sympto- included, of whom most had mild or moderate matic relief. disease and were thus more likely to have responded to H2RA therapy. If there are mean- Healing of Esophagitis ingful differences among PPIs, they would more likely be observed in trials involving patients with Healing is the primary endpoint in many GERD pre-existing severe or complicated disease. Also, studies. Esophagitis heals in 70% to 95% of long-term follow-up data beyond 3 years of PPI patients within 4 to 8 weeks of treatment. More use have not been obtained. severe forms of erosive esophagitis may take longer to heal. Higher PPI doses, often given In summary, PPIs suppress acid secretion more twice daily, may be helpful in some of these effectively than do H2RAs, but differences in patients. PPIs consistently have healing rates that efficacy among the PPIs are small. There is little are superior to those of H2RAs,34 meaning that evidence to support a clear relation between acid 10% to 20% more of patients who are taking suppression and symptom control, healing of PPIs will experience healing of their esophagitis esophagitis, maintenance of remission, or preven- compared with those taking H2RAs. This differ- tion of complications. ence is more pronounced with erosive esophagi- tis. There are no studies that directly compare all of the available PPIs. One paper that evaluated studies comparing omeprazole 20 mg with esomeprazole 40 mg determined that esomepra- 12 | CALIFORNIA HEALTHCARE FOUNDATION zole was superior.35 However, these studies used a low-dose PPIs (e.g., pantoprazole 20 mg, a dose higher dose (40 mg) of esomeprazole as com- that is not available in the United States) have pared with omeprazole (20 mg). In two meta- yielded variable results. Thus, there is insufficient analyses that compared the various PPIs,36,37 one evidence to conclude that one PPI is superior to concluded that omeprazole and lansoprazole have another for maintenance therapy. Furthermore, similar efficacy,36 whereas the other found that there have been no trials involving patients with esomeprazole 40 mg was superior to omeprazole GERD who have major problems, such as noc- 20 mg (relative risk = 1.18; 95% confidence turnal heartburn, uncontrolled reflux despite a interval: 1.14 to 1.23).37 A very large, well- pH of >4.0, recurrent strictures, or intractable designed randomized controlled trial of esophagi- extraesophageal disease, despite therapy. tis in 5241 patients with GERD found higher healing rates at 8 weeks with esomeprazole 40 Prevention of Complications mg/d than with lansoprazole 30 mg/d (92.6% vs. 88.8%).38 Furthermore, the superiority of Esophageal strictures, ulcers, and hemorrhages esomeprazole over lansoprazole increased with are serious complications of GERD. These com- the baseline severity of disease, and sustained plications, and the presence of Barrett’s esopha- resolution of symptoms (including nocturnal gus, are generally believed to be associated with heartburn) was also superior in the esomeprazole- prolonged exposure of the esophagus to acid. treated group. Indeed, a 40-mg dose of drug may While the inflammatory complications of GERD be effective in patients with severe erosive may be reduced with PPI therapy, there is still no esophagitis.39-41 clear evidence of this beneficial effect. For instance, there are no sound data showing that In summary, all PPIs are very effective at healing Barrett’s esophagus or the risk of cancer is esophagitis, and the extent of healing with PPI reduced solely as a result of PPI therapy. therapy is greater than with H2RA therapy. Maintenance PPI therapy following esophageal Patients with more severe esophagitis may require dilatation has been shown to result in fewer a longer course of treatment or higher doses to recurrences of stricture as compared with placebo achieve or maintain healing. Reported differences therapy. However, there have been no studies between PPIs may not be clinically meaningful. comparing the efficacy of different PPIs in this application. Prevention of Recurrence of Esophagitis Adverse Effects and All five of the PPIs have been evaluated for the Drug Interactions prevention of the recurrence of esophagitis. In The most common adverse events associated general, about 80% to 90% of patients remain in with PPIs are nausea, diarrhea, constipation, and remission after 1 year of treatment. Most studies headache.43 The frequency of these events is usu- find recurrence rates with PPIs to be about 50% ally less than 5%. The sole exception is the fre- lower than with H2RAs. Studies comparing the quent association of diarrhea with lansoprazole, efficacy of different PPIs in preventing the recur- which may approach 7% or higher if patients do rence of esophagitis are rare. In one trial, patients not take the drug at least 30 to 45 minutes taking omeprazole 20 mg or lansoprazole 30 mg before meals. The rates of other adverse events had similar rates of recurrence of esophagitis.2 are similar across PPIs. There is some concern Similar comparisons between “normal-dose” and that chronic PPI use might place patients at an Treatment Options for GERD: Scientific Review | 13 increased risk of developing a carcinoid tumor or Endoscopic and Surgical Treatment gastric carcinoma (as a result of accelerating the Not all patients with GERD respond to medical progression of H. pylori–induced atrophic gastri- treatment. A small percentage continue to have tis). These concerns have probably been overstat- persistent symptoms or tissue injury (often in the ed, and there is a growing consensus that use of bronchopulmonary tree, larynx, pharynx, or PPIs for 3 years or more does not increase the esophagus) despite high doses of PPIs. risk of gastric tumors.44 Overall, PPIs are associat- Endoscopic therapy or surgery may be options ed with a low rate of adverse reactions, similar to for these patients. Endoscopic treatment is largely H2RAs. There is also little evidence that the rates experimental at present; it has not been validated of adverse effects are lower with any one PPI. in randomized controlled trials and should be In general, drug interactions have not been a seri- used under protocol in special centers. Surgery ous problem in younger patients taking PPIs. (usually a Nissen fundoplication) may occasional- However, this may not apply in patients taking ly be helpful in the short-term in patients who large numbers of medications, such as elderly fail medical treatment, but at least one study has patients and those in a chronic care or polyphar- concluded that two thirds of patients who under- macy setting. For instance, in patients taking go surgery still need PPI therapy for control of medicines (e.g., carbamazepine) for epilepsy symptoms.45 The success of the procedure (seizures), omeprazole may be the most likely to depends highly on the expertise of the surgeon. interact and thus should be used with caution. Results in a small number of centers have ranged All PPIs impair the bioavailability of drugs that from good to very good, but the long-term suc- are dependent on the presence of gastric acid for cess rate is not known. Increasingly, the patients absorption, for example, iron salts, digoxin, and selected for surgery are those with severe regurgi- ketoconazole. While PPIs also reduce the secre- tation or extraesophageal symptoms who often tion of gastric intrinsic factor in humans, to date exhibit satisfactory pH control with PPI therapy no clinical cases of vitamin B12 deficiency have and who are at risk of aspiration. been reported. In animal studies, the PPIs exhibiting the lowest potential for pharmacoki- Cost Considerations netic interactions are rabeprazole, pantoprazole, and lansoprazole. Package inserts for omeprazole PPIs account for about $12 billion yearly in U.S. and esomeprazole now warn about the risks of sales. Both PPIs and H2RAs are very effective in drug interactions with warfarin, phenytoin, and most patients with GERD who are seen in pri- benzodiazepines. Concerns about drug-drug mary care settings (Tables 2 and 3). In fact, many interactions are the subject of much ongoing patients are prescribed PPIs for vague dyspeptic research. symptoms or functional heartburn, which com- pounds the problem of overusage. Current rec- ommendations to initiate empirical treatment in patients who are suspected of GERD on the basis of symptoms also contribute to the high use of acid-suppressing drugs. Several strategies to con- trol symptoms and lower the cost of GERD treatment have been proposed and are justified largely based on modeling studies. Among these are the “step” treatment plans (“step up” and “step down”) and “on-demand therapy”. 14 | CALIFORNIA HEALTHCARE FOUNDATION Step-up Regimens demand” approach regarding use of their drug The step-up plan initiates treatment involving over the long-term, especially those with income lifestyle management and H2RAs (in adequate constraints. Such an approach may be cost-sav- doses) at the first visit, progressing to use of PPIs ing, but results in fewer symptom-free days than only in patients whose symptoms persist after with continuous PPI treatment.46 4 to 8 weeks of H2RA therapy. This approach is PPIs generally cost two to five times more than cost-effective but delays relief of symptoms in as H2RAs, if only pharmacy costs are considered many as 10% to 15% of patients.46 (Table 2). However, several studies have com- pared the total health care costs associated with Step-down Regimens the use of these two classes of medications and In the step-down treatment plan, patients start found similar overall costs when physician and with a PPI and, when symptoms are controlled, other costs were added. It should be noted that “step down” to H2RA therapy. If symptoms all of these models assumed that patients had remain controlled with H2RA therapy, the doses esophagitis (which was incorrect in most cases) are reduced and the patient is eventually taken and included the cost of initial and, sometimes, off medication. Drug treatment is re-initiated if follow-up endoscopy. symptoms return, with the choice of therapy depending on the frequency and severity of symptoms as well as other patient factors.46 Howden and associates compared GERD symp- tom control and health care costs in patients using either of the “step” plans and in patients solely using PPIs.46 After 1 year, costs were lower with both “step” plans, but symptom control was poorer than was observed with continuous PPI therapy. On-demand Therapy Since many patients experience good symptom control with PPIs, it has been suggested that after symptoms are controlled for several months, patients may wish to stop medication and resume PPI treatment intermittently if symptoms recur. This can be done without a physician visit. If symptoms recur very frequently, or if they persist for unacceptably long periods after resuming use of acid-suppressing drugs, re-evaluation by a physician is necessary. The clinical effectiveness of this on-demand strategy has not been validated in carefully controlled trials, even though it has received favorable anecdotal endorsements by several opinion leaders. Many doctors believe that most patients in practice adopt an “on- Treatment Options for GERD: Scientific Review | 15 III. Clinical Conclusions GERD IS NOT ONLY A BOTHERSOME CONDITION, but it is also associated with decreased quality of life and can sometimes lead to serious complications. Medical treatment can relieve symptoms, heal esophagitis, and prevent recurrences of the disease, thus reducing the risks of complications and the need for expensive and often ineffective surgery. Patients should adopt healthy lifestyles and avoid activities or foods that cause heartburn or reflux. Still, most patients will require drug therapy at least some of the time. Antacids and OTC H2RAs are effective in relieving occasional symptoms for short periods, but are less effective in controlling more serious or acute symp- toms. H2RAs, in adequate dosage, reduce symptoms and heal esophagitis in many patients, but they do so less rapidly and effectively than do PPIs. Currently, PPIs are the most effective treatment for all grades of GERD. They are far superior to H2RAs in patient accept- ability, as well as in reducing the burden to physicians and health care systems taking care of the entire disease burden. There is no evidence at present to document the superiority of one PPI over another, but 40 mg of any PPI is more effective than 20 mg of the same or any other PPI. Recently, omepra- zole became available as an OTC drug. Its role in the treat- ment of GERD is currently evolving. Nonmedical treatments other than surgery are unvalidated and should not be support- ed outside of a research setting. Nissen fundoplication (and its variants) is reserved for patients who fail expertly supervised medical management, many of whom have serious extrae- sophageal problems. Currently available surgical techniques are not effective enough to be regarded as curative in all but a small number of cases. 16 | CALIFORNIA HEALTHCARE FOUNDATION IV. Appendices Table 1. Proton Pump Inhibitors and Their Indications Indications Drug Dosage Form Daily Dose GERD Acute Maintenance of Esophagitis Esophagitis Omeprazole* A 20 mg X X X Omeprazole† A 20 mg X X X Lansoprazole A 15 mg X X Lansoprazole A 30 mg X Rabeprazole B 20 mg X X X Pantoprazole B 40 mg X X X Esomeprazole A 20 mg X X X Esomeprazole A 40 mg X X X A: Capsule filled with enteric-coated beads. B: Enteric slow-release tablet. * Prilosec (AstraZeneca). † Generic GERD = gastroesophageal reflux disease. Treatment Options for GERD: Scientific Review | 17 Table 2. Cost Comparisons of Acid-Reducing Agents Drug Category Dose Range per Day AWP Cost Range AWP Cost for for 30-Day Supply 30-Day Average Drug Name Brand Name Daily Dosage PPIs Brand* Generic* Brand Generic & & Esomeprazole (Nexium®) 20 mg – 40 mg $134 - $134 N/A $$$$ N/A & & Lansoprazole (Prevacid®) 15 mg – 30 mg $133 - $136 N/A $$$$ N/A & s Omeprazole (Prilosec®) 10 mg – 40 mg $118 - $189 $111 - $190 $$$$ $$$$ & & Pantoprazole (Protonix®) 20 mg – 40 mg $108 - $109 N/A $$$$ N/A & & Rabeprazole (Aciphex®) 20 mg – 20 mg $141 -$141 N/A $$$$ N/A H2RAs & s Cimetidine (Tagamet®) 800 mg - 1.6 g $93 - $187 $55 - $109 $$$ $$$ & s Famotidine (Pepcid®) 20 mg – 40 mg $84 - $112 $8 - $14 $$$ $ & s Nizatidine (Axid®) 150 mg – 300 mg $87 - $171 $30 - $100 $$$ $$ & s Ranitidine (Zantac®) 150 mg – 300 mg $62 - $112 $6 - $13 $$$ $ Over the Counter ** ** q q Cimetidine (Tagamet HB®) 200 mg – 400 mg $8 - $18 $6 - $11 $ $ q q Famotidine (Pepcid AC®) 10 mg – 20 mg $8 - $16 $6 - $12 $ $ q q Nizatidine (Axid AR®) 75 mg – 150 mg $9 -$18 N/A $ N/A q q Ranitidine (Zantac 75®) 75 mg – 150 mg $9 - $17 $6 - $13 $ $ q q Omeprazole Prilosec® 20 mg – 40 mg $20 - $40 N/A $ $ * Discounted AWP cost reflects brand discount of AWP - 5% and generic AWP - 30%. 30-day drug supply based on minimum and maximum dosage ranges. Pricing reference dated 12/31/03, pricing results may vary due to use of different strengths/cost to achieve dosage. ** OTC Pricing is retail-based q = Drug is non-prescription item and may not be covered by health plans. s = Available as brand and generic. Generic will have a lower copay; brand copay is dependent on plan formulary. & = Brand, copay varies by plan. $ = Less than $40.00 $$ = Between $40.00 and $79.99 $$$ = Between $80.00 and $119.99 $$$$ = More than $120.00 18 | CALIFORNIA HEALTHCARE FOUNDATION Endnotes 1. Wilkes M.S., Bell R.A., Kravitz R.L. Direct-to- 14. DeVault K.R., Castell D.O. Updated guidelines consumer prescription drug advertising: trends, for the diagnosis and treatment of gastroe- impact, and implications. Health Aff (Millwood). sophageal reflux disease. The practice parameters 2000;19:110-128. committee of the American College of 2. Heartburn Across America: A Gallup Organization Gastroenterology. Am J Gastroenterol. National Survey. Princeton, New Jersey: The 1999;94:1434-1442. Gallup Organization, Inc; 1988. 15. Meining A., Classen M. The role of lifestyle 3. Dent J., on behalf of the Genval Workshop measures in the pathogenesis and treatment of Group. An evidence based appraisal of reflux gastroesophageal reflux disease. Am J disease management—The Genval Workshop Gastroenterol. 2000;95:2692-2697. Report. Gut. 1999;44(suppl 2):S1-S16. 16. Kjellin A., Ramel S., Rossner S., Thor K. 4. Dent J., Jones R., Kahrilas P., Talley N.J. Gastroesophageal reflux in obese patients is not Management of gastro-esophageal reflux disease reduced by weight reduction. Scand J in general practice. BMJ. 2001;322:344-347. Gastroenterol. 1996;31:1047-1051. 5. Richter J.E. Peptic strictures of the esophagus. 17. Holloway R.H., Dent J. Medical management of Gastroenterol Clin North Am. 1999;28:875-891. gastroesophageal reflux disease—beyond the pro- ton pump inhibitors. Dig Dis. 2000;18:7-13. 6. Lagergren J., Bergstrom R., Lindgren A., et al. Symptomatic gastroesophageal reflux as a risk 18. Castell D.O. Long-term management of GERD: factor for esophageal adenocarcinoma. N Engl J the pill, the knife, or endoscope? Gastrointest Med. 1999;340:825-831. Endosc. 1994;40:252-253. 7. Kasano M., Ino K., Yamada T., et al. 19. Graham D.Y., Patterson D.J. Double-blind com- Interobserver and intraobserver variation in parison of liquid antacid and placebo in the endoscopic assessment of GERD using the “Los treatment of symptomatic reflux esophagus. Dig Angeles” classification. Gastrointest Endosc. Dis Sci. 1983;28:559-563. 1999;49:700-704. 20. McCallum R.W. Cisapride: a new class of proki- 8. Demeester T.R., Johnson L.F., Joseph G.L., et al. netic agents. Am J Gastroenterol. 1991;86:135-149. Patterns of gastroesophageal reflux in health and 21. McCallum R.W., Ippoliti A.F., Cooney C., disease. Ann Surg. 1976;184:459-470. Sturdevant R.A. A controlled trial of metoclo- 9. Reynolds J.C., Waronker M., Pacquing M.S., pramide in symptomatic gastroesophageal reflux. Kyassin R.R. Gastroesophageal reflux disease: N Engl J Med. 1977;296:354-357. Barrett’s esophagus—reducing the risk of pro- 22. Garnett W.R. Lansoprazole: a proton pump gression to adenocarcinoma. Gastroenterol Clin. inhibitor. Ann Pharmacother. 1996;30:1425-1436. 1999;28:917-945. 23. Richardson P., Hawkey C.J., Stack W. Proton 10. Murray L., Watson P., Johnston B., Sloan J., pump inhibitors: pharmacology and rationale for Mainie I.M., Gavin A. Risk of adenocarcinoma use in gastrointestinal disorders. Drugs. in Barrett's oesophagus: population based study. 1998;56:307-335. BMJ. 2003;327:534-535. 24. Bright-Asare P., Behar J., Brand D.L., et al. 11. Dent J. Controversies in long-term management Effects of long-term maintenance cimetidine on of reflux disease. Ballieres Best Pract Res Clin gastroesophageal reflux disease [abstract]. Gastroenterol. 2000;14:811-826. Gastroenterology. 1985;82(pt 2):1025. 12. Lundell L.R., Dent J., Bennett J.R., et al. 25. Sontag S., Vlachevic R., Orr W., et al. Ranitidine Endoscopic assessment of esophagitis: clinical vs. placebo in long term treatment of gastroe- and functional correlates and further validation sophageal reflux [abstract]. Gastroenterology. of the Los Angeles Classification. Gut. 1985;88(pt 2):1595. 1999;45:172-180. 26. Klinkenberg-Knol E.C., Jansen J.B.M.J., Lamers 13. Siepler J.K. Gastroesophageal reflux disease. In: C.B.H.W., et al. Use of omeprazole in the man- Meuller B.A., Bertch K.E., Dunsworth T.S., et agement of reflux esophagitis. Scand J al, eds. Pharmacotherapy Self-Assessment Program. Gastroenterol. 1989;24 (suppl 166):88-93. 4th ed. Kansas City, Missouri: American College 27. Proton Pump Inhibitors: Subcommittee Report. of Clinical Pharamcy; 2002. Oregon Health Resources Commission; June 2002. Treatment Options for GERD: Scientific Review | 19 28. Blum R.A., Shi H., Karol M.D., et al. The com- 40. Jones R., Bytzer P. Acid suppression in the man- parative effects of lansoprazole, omeprazole, and agement of gastro-oesophageal reflux disease—an ranitidine in suppressing gastric acid secretion. appraisal of treatment options in primary care. Clin Ther. 1997;19:1013-1023. Aliment Pharmacol Ther. 2001;15:765-772. 29. Pantoflickova D., Dorta G., Ravic M., et al. Acid 41. Sharma V.K., Leontiadis G.I., Howden C.W. inhibition on the first day of dosing: comparison Meta-analysis of randomized controlled trials of four proton pump inhibitors. Aliment comparing standard clinical doses of omeprazole Pharmacol Ther. 2003;17:1507-1514. and lansoprazole in erosive esophagitis. Aliment 30. Frazzoni M., De Micheli E., Grisendi A., Pharmacol Ther. 2001;15:227-231. Savarino V. Effective intra-oesophageal acid sup- 42. Carling L., Axelsson C.K., Forsell H., et al. pression in patients with gastro-oesophageal Lansoprazole and omeprazole in the prevention reflux disease: lansoprazole vs. omeprazole. of relapse of reflux esophagitis: a long-term Aliment Pharmacol Ther. 2003;17:235-241. study. Aliment Pharmacol Ther. 1998;12:985-990. 31. Hatlebakk J.G. Review article: gastric acidity— 43. Berardi R.R., Welage L.S. Proton pump comparison of esomeprazole with other proton inhibitors in acid-related disorders. Am J Health pump inhibitors. Aliment Pharmacol Ther. Syst Pharm. 1998;55:2289-2298. 2003;17(suppl 1):10-15. 44. Laine L., Ahnen D., McClain C., Solcia E., 32. Castell D.O., Richter J.E., Robinson M., Sontag Walsh J.H. Review article: potential gastrointesti- S.J., Haber M.M. Efficacy and safety of lanso- nal effects of long-term acid suppression with prazole in the treatment of reflux esophagitis. proton pump inhibitors. Aliment Pharmacol Ther. Am J Gastroenterol. 1996;91:1749-1757. 2000;14:651- 668. 33. Kromer W., Horbach S., Luhmann R. Relative 45. Spechler S.J., Lee E., Ahnen D., et al. Long-term efficacies of gastric proton pump inhibitors: their outcome of medical and surgical therapies for clinical and pharmacological basis. Pharmacology. gastroesophageal reflux disease. JAMA. 1999;59:57-77. 2001;285:2331-2338. 34. Richter J.E., Campbell D.R., Kahrilas P.J., et al. 46. Howden C.W., Henning J.M., Huang B., et al. Lansoprazole compared with ranitidine for the Management of heartburn in a large, randomized treatment of nonerosive gastroesophageal reflux community-based study: comparison of four disease. Arch Intern Med. 2000;160:1803-1809. therapeutic strategies. Am J Gastroenterol. 35. Edwards S.J., Lind T., Lundell L. Systematic 2001;96:1679-1681. review of proton pump inhibitors for acute treat- ment of reflux esophagitis. Aliment Pharmacol Ther. 2001;15:1729-1736. 36. Sharma V.K., Leontiadis G.I., Howden C.W. Meta-analysis of randomized controlled trials comparing standard clinical doses of omeprazole and lansoprazole in erosive esophagitis. Aliment Pharmacol Ther. 2001;15:227-231. 37. Klok R.M., Postma M.J., Van Hout B.A., Brouwers J.R. Meta analysis: comparing the effi- cacy of proton pump inhibitors in short term use. Aliment Pharmacol Ther. 2003;17:1237-1245. 38. Vakil N. Esomeprazole, 40 mg once daily, com- pared with lansoprazole 30 mg once daily, in healing and symptom resolution of erosive esophagitis. Aliment Pharmacol Ther. 2003;14(suppl 1):21-23. 39. Caro J.J., Salas M., Ward A. Healing and relapse rates in gastroesophageal reflux disease treated with the newer proton pump inhibitors lansopra- zole, rabeprazole, and pantoprazole compared with omeprazole, ranitidine and placebo. Evidence from randomized clinical trials. Clin Ther. 2001;23:998-1017. 20 | CALIFORNIA HEALTHCARE FOUNDATION