SUMMARY INTRODUCTION. Aliment Pharmacol Ther 2000; 14: 1595±1603. Accepted for publication 14 August 2000

Transcription

1 Aliment Pharmacol Ther 2000; 14: 1595±1603. Omeprazole 40 mg once a day is equally effective as lansoprazole 30 mg twice a day in symptom control of patients with gastro-oesophageal re ux disease (GERD) who are resistant to conventional-dose lansoprazole therapyða prospective, randomized, multi-centre study R. FASS*, U. MURTHYà, C.W.HAYDEN*,I.B.MALAGON*,G.PULLIAM*,C.WENDEL & T. O. G. KOVACS *Section of Gastroenterology, Department of Medicine, and Health Services Research Center, Tucson VA Medical Center, Tucson, Arizona, àdepartment of Medicine, Section of Gastroenterology, Syracuse VA Medical Center, Syracuse, New York, and Center of Ulcer Research and Education (CURE), VAGLAHS, Los Angeles, California Accepted for publication 14 August 2000 SUMMARY Background: Comparative studies of omeprazole and lansoprazole are scarce and even scarcer are comparisons of higher doses. Most of the comparative studies have assessed the effect of the two proton pump inhibitors (PPIs) on gastric acid secretion or gastric ph. Few studies have compared clinical end-points such as oesophageal healing and symptom control. Aim: To determine the clinical ef cacy of omeprazole 40 mg daily as compared to lansoprazole 30 mg twice a day in symptom control of patients with severe symptomatic GERD. Methods: Ninety-six patients who failed a standard dose of lansoprazole (30 mg once daily), were enrolled in a prospective fashion from three VA medical centres and were randomized to receive 6 weeks of either omeprazole 40 mg daily or lansoprazole 30 mg twice daily. Patients reported daily on symptom severity and frequency, antacid consumption and side-effects. Results: Forty-six patients received omeprazole and 44 lansoprazole. Although not statistically signi cant, there was a consistent trend of better symptom control in the omeprazole group for daytime and night-time heartburn and acid regurgitation. There was no statistical difference between the two groups in mean antacid consumption overall and at the end of each of the 6 weeks of the study. In addition, there was no signi cant difference in the overall frequency of sideeffects between the two groups nor for each individual side-effect. Conclusion: Omeprazole 40 mg once daily is equally effective and tolerated as lansoprazole 30 mg twice daily in symptom control of patients with GERD. INTRODUCTION Correspondence: Dr R. Fass, Southern Arizona VA Health Care System 111G-1, 3601 S 6th Avenue, Tucson, AZ 85723, USA. Ronnie.Fass@Med.VA.gov The failure of standard doses of proton pump inhibitor (lansoprazole 30 mg and omeprazole 20 mg) to control patient's symptoms of GERD may indicate a more severe form of the disease. 1 Evaluation by 24-h oesophageal ph Ó 2000 Blackwell Science Ltd 1595

2 1596 R. FASS et al. monitoring on proton pump inhibitor therapy to assess adequate acid suppression has been recommended in patients with GERD who continued to be symptomatic on standard doses of proton pump inhibitor. 2 However, due to the limited availability and discomfort of ph testing, many patients will be placed on double dose proton pump inhibitor without further evaluation. The currently available proton pump inhibitors are omeprazole, lansoprazole, rabeprazole and pantoprazole. Comparative studies between omeprazole and lansoprazole using standard doses are scarce and even scarcer are comparisons of higher doses. Most of the comparative studies have assessed the effect of the two PPIs on gastric acid secretion or indirectly on gastric ph. 3±8 The results of these studies are very variable. Several studies have suggested that lansoprazole 15 mg daily has an equal antisecretory potency as omeprazole 20 mg daily in controlling gastric ph. 4, 7, 8 In contrast, other studies found no difference between lansoprazole 30 mg daily and omeprazole 20 mg daily in raising intragastric ph, regardless of the Helicobacter pylori status of the patient. 3, 5, 6 However, these comparative studies were mainly done in normal subjects and their clinical relevance in GERD patients remained unclear. Clinical studies in patients with GERD that have compared the two PPIs using standard doses also demonstrated con icting results. 9±13 Oesophageal mucosal healing and symptom relief were the main clinical end-points assessed. In general, these studies have suggested that omeprazole 20 mg daily was equally effective as lansoprazole 30 mg daily in providing oesophageal mucosal healing and overall symptom relief. However, lansoprazole appeared to be more ef cacious in inducing symptom relief during the rst week of treatment. 9±11 Thus far, there have been almost no studies that compared the two proton pump inhibitors in patients with severe GERD that required double-dosing. 13 The current preferred treatment approach of patients that require double dose proton pump inhibitor is to split the dose and administer one capsule in the morning and the other in the evening (half and hour before a meal). Dividing the dosing of omeprazole (20 mg b.d.) has been shown to provide superior gastric acid suppression as compared to a once-a-day regimen. 14 Translation of this effect into studies assessing clinical end-points in GERD patients has yet to be carried out. The aim of this study was to determine the clinical ef cacy of omeprazole 40 mg daily as compared to lansoprazole 30 mg twice a day in symptom control of patients with severe symptomatic GERD. Our hypothesis was that omeprazole 40 mg daily is at least as effective as lansoprazole 30 mg twice a day. MATERIALS AND METHODS This was a prospective, randomized, parallel-group, comparative, multicentre VA study, involving three VA medical centres. Patients were enrolled at Tucson VA, Syracuse VA and VAGLAHS. The human subject committee of each participating medical centre approved this study. Patients Patients who remained symptomatic on a standard dose of lansoprazole (30 mg) after a trial of therapy for at least 3 months were recruited into the study. Therapeutic failure of standard dose lansoprazole was considered if patients continued to experience heartburn episodes more than once a week. Patients who ful lled the inclusion criteria were referred to the study from primary care and out-patient clinics. Patients were excluded if they were allergic to one of the proton pump inhibitors, were unable or unwilling to sign an inform consent, unable to complete all symptom diaries or all stages of the study. All participating subjects signed an informed consent before beginning the study. Study protocol Initially, patients recruited from primary care and gastroenterology clinics underwent a detailed interview about their demographics, past medical and surgical history, current usage of medication and habitsðsmoking and alcohol. Subsequently, patients lled out a previously validated GERD assessment questionnaire, which included questions about heartburn and acid regurgitation frequency and severity. 15 Questionnaire responses were used to construct baseline symptom scores for heartburn and acid regurgitation. By using a randomization list, patients were then assigned to either omeprazole 40 mg daily or lansoprazole 30 mg twice a day (morning and evening, 30 min before a meal). Thereafter, patients received their treatment for 6 weeks. During this period, patients were asked to complete a daily symptom assessment diary and document any side-effects that might have developed from

3 OMEPRAZOLE VS. LANSOPRAZOLE TREATMENT IN GERD 1597 proton pump inhibitor usage. Mylanta tablets were provided for treatment rescue (when heartburn symptoms recur) and their usage was documented as well in patient's diaries. The diaries were collected after 2, 4 and 6 weeks. Full completion of all recorded information was ensured and compliance with therapy was assessed by pill count during these visits. At the end of the study, all daily symptom records and remaining pills were collected. Patients were able to terminate their participation at any time during the study. Furthermore, medical personnel were available 24-h-a-day to answer questions from patients and to receive reports about adverse reactions from treatment. Symptom assessment Patients kept a daily record of the frequency and severity of each symptom they experienced. Symptoms such as daytime heartburn, night-time heartburn and acid regurgitation were evaluated. The following scale was used to determine severity of each symptom: none, mildðsymptom easily tolerated and did not last long; moderateðsymptom caused some discomfort but did not interfere with usual activities; severeðsymptom caused much discomfort and interfered with usual activities; and disablingðsymptoms unbearable and interfered considerably with usual activities. Symptom score was calculated by adding the reported daily severity (none 0, mild 1, moderate 2, severe 3 and disabling 4) multiplied by the reported daily frequency values as obtained during each day of symptom 16, 17 recording. Statistical analysis Results are presented as mean S.E.M. Patients' interview responses, prior to beginning the study, were used to construct baseline symptom scores for heartburn and acid regurgitation. In order to display an average improvement over time, change in symptom severity score was calculated by subtracting severity scores for each week from the baseline severity score. For symptom severity change in the rst 7 days of treatment, the daily score was subtracted from the average daily baseline score. Severity scores were analysed with a cross-sectional time-series linear model (baseline to week 6 for weekly symptoms, baseline to day 7 for rst week) using generalized estimating equations (GEE), to account for autocorrelation across time periods within subjects. 18 The GEE model allows the characterization of patterns of individual change over time and the effect of covariates on these patterns. Advantages of this method over repeated measures ANOVA include (i) the focus on individual subject changes rather than group means, and (ii) one does not lose cases due to missing observations. The model used in this analysis assumed a Gaussian distribution and an exchangeable withingroup correlation structure (similar to a random-effects model) and assessed the effect of drug assignment on severity scores over time, adjusting for age, sex and medical centre. Indicator variables compared enrolment at the Syracuse VA or the Los Angeles VA to enrolment at the Tucson VA as a reference. The coef cients from the model indicated a difference across time in symptom score resulting from a unit change in the variable, adjusting simultaneously for the effects of other variables in the model. In the case of dichotomous variables such as drugs, the coef cient represents the difference in symptom score between one drug and the other. There were six patients who dropped out from this study. However, in two subjects symptom diaries were available for 1 and 3 weeks, respectively. These diaries were included in the nal statistical analysis. The absolute level of antacid use between the two drug groups was analysed by t-tests for each week, as well as v 2 tests on the proportions taking any antacids for each week. Comparison of the proportion of subjects between the two drug groups reporting speci c side-effects any time in the course of follow-up was assessed by v 2 tests. All statistical tests utilized an a of As the study hypothesis is that omeprazole daily is at least as effective as lansoprazole twice daily, the power calculation involved a one-sided alpha of Based on a t-test of severity change scores by week, the study had 80% power to detect a difference in severity change score of approximately 4 units. RESULTS Patients Of the 96 patients who were enrolled in the study and subsequently randomized into one of the treatment arms, six (6.3%) patients dropped out. Four dropped out from the omeprazole group and two from the

4 1598 R. FASS et al. Table 1. Patient demographics Omeprazole (40 mg once daily) n =46 Lansoprazole (30 mg twice daily) n =44 Age Mean Range 32±84 35±77 Gender Male 44 (95.7%) 40 (90.9%) Female 2 (4.3%) 4 (9.1%) Ethnicity Caucasian 34 (73.9%) 32 (72.7%) African American 7 (15.2%) 5 (11.4%) Hispanic 2 (4.4%) 1 (2.3%) Asian 1 (2.2%) Ð Unknown* 2 (4.4%) 6 (13.6%) * Patients who elected not to report ethnicity. lansoprazole group. Two patients were unable to complete all diaries, two patients requested to be removed from the study, another was lost to followup, and one developed angina pectoris and withdrew himself from the study. Forty-four patients were randomized to receive lansoprazole 30 mg twice daily and 46 omeprazole 40 mg once daily. There were no demographic differences between the two groups except the inclusion of four (9.1%) female patients in the lansoprazole group as compared to two (4.3%) in the omeprazole group, which was not statistically signi cant (Table 1). There was no signi cant difference in cigarette consumption between the two groups. Only 14 (31.8%) patients in the lansoprazole group and 10 (21.7%) patients in the omeprazole group were smoking cigarettes (mean cigarettes per day: 10 and 8, respectively). Similarly, there was no statistical difference in alcohol consumption between the lansoprazole and omeprazole group. Ten (22.7%) patients were actively drinking alcohol in the former group and eight (17.4%) patients in the latter group. Symptom control There was no signi cant difference in the effect of both drugs on daytime and night-time heartburn and acid regurgitation symptom score. The mean improvement from baseline of daytime, night-time heartburn and acid regurgitation symptom score between the two drug groups over 6 weeks of heartburn is shown in Figure 1. Improvement in the omeprazole group was consistently Figure 1. Mean improvement in symptom score for daytime heartburn (A), night-time heartburn (B) and acid regurgitation (C) for the 6 weeks of the study. Improvement in symptom score was calculated for each patient by subtracting severity scores for each week from the baseline severity score. Using time-series regression analysis revealed a consistently higher symptom score improvement in the omeprazole group as compared to the lansoprazole group, which did not reach statistical signi cance (P > 0.05).

5 OMEPRAZOLE VS. LANSOPRAZOLE TREATMENT IN GERD 1599 Figure 2. Mean improvement in symptom score for daytime heartburn (A), night-time heartburn (B) and acid regurgitation (C) for the rst 7 days of treatment. For symptom severity change in these rst 7 days, the daily score was subtracted from the average daily baseline score. Using time-series regression analysis revealed a consistently higher symptom score improvement in the omeprazole group as compared to the lansoprazole group, which did not reach statistical signi cance (P > 0.05). higher than in the lansoprazole group, but did not reach statistical signi cance. Figure 2 shows a similar trend in the mean improvement from baseline of daytime and night-time heartburn and acid regurgitation symptom score over the rst 7 days. The preponderance of negative improvement in the lansoprazole group probably re ects bias in our estimates of baseline symptoms, particularly for acid regurgitation, rather than a deterioration of symptoms in this group. Because this bias would be equal between the two drug groups, however, it does not in uence the relative comparison of the two groups. Time-series models indicated no signi cant drug effect on daytime and night-time heartburn and acid regurgitation symptom scores, controlling for age, sex, and medical centre (Tables 2 and 3). These models produced beta coef cients and P-values that are interpreted in a similar fashion to least-squares regression, i.e. the change in symptom severity per unit change in the predictor variable, with a signi cant P-value (< 0.05) indicating a signi cant association. The time-series models involved raw symptom scores, rather than improvement, so that a negative coef cient indicated a decrease in symptom scores. Although modest (less that 1 unit on average in symptom score) and not statistically signi cant, the coef cient for drug was consistently negative across the three symptoms evaluated, in both the rst week and over the entire study. Because omeprazole was coded as 1 and lansoprazole as 0, a negative coef cient denotes that omeprazole subjects experienced greater decreases in symptoms. Over the 6 weeks of study, subjects at the Syracuse VA had a 3.5 unit decrease in acid regurgitation symptom score, relative to Tucson VA subjects (P < 0.05). The Los Angeles VA subjects showed a similar difference, which was not signi cant. These ndings re ect the fact that the Tucson subjects had higher average severity scores at baseline and throughout treatment, but not that the relative bene t from treatment was different from other centres. Over the rst 7 days, the Syracuse VA subjects had a 0.7 unit increase in night-time heartburn scores relative to the Tucson VA subjects (P < 0.05). Current smoking status, when included in the models, was not associated with symptom severity and did not appreciably alter the coef cients of other variables. The v 2 statistic for the overall model is a test that the predictor variable coef cients are all jointly zero. Taken jointly, the coef cients were not signi cant in any of the six models tested.

6 1600 R. FASS et al. Table 2. Time-series regression on weekly symptom severity Daytime heartburn Night-time heartburn Acid regurgitation b coef. 95% CI b coef. 95% CI b coef. 95% CI Drug )0.698 ()3.29, 1.89) )0.111 ()3.36, 3.13) )0.528 ()3.28, 2.22) Syracuse )0.706 ()3.94, 2.53) ()3.74, 4.36) )3.33* ()6.77, 0.097) Los Angeles )2.39 (5.60, 0.82) )1.60 ()5.63, 2.42) )3.52à ()6.94, )0.118) Age )0.020 ()0.11, 0.07) )0.049 ()0.16, 0.06) )0.046 ()0.143, 0.049) Gender ()4.76, 5.72) ()6.45, 6.68) )0.206 ()5.76, 5.35) Overall model for daytime heartburn v 2 (5) = 2.69, P = 0.75, for night-time heartburn v 2 (5) = 1.63, P = 0.9 and acid regurgitation v 2 (5) = 5.67, P = à P < * P = Table 3. Time-series regression on rst 7-day symptom severity Daytime heartburn Night-time heartburn Acid regurgitation b coef. 95% CI b coef. 95% CI b coef. 95% CI Drug )0.134 ()0.64, 0.37) )0.233 ()0.75, 0.28) )0.107 ()0.61, 0.39) Syracuse ()0.13, 1.1) 0.694à (0.04, 1.3) )0.471 ()1.1, 0.15) Los Angeles ()0.47, 0.78) ()0.55, 0.7) )0.463 ()1.0, 0.16) Age ()0.00, 0.03) ()0.00, 0.02) )0.004 ()0.02, 0.01) Gender ()0.63, 1.4) ()0.55, 1.5) )0.520 ()1.5, 0.49) Overall model for daytime heartburn v 2 (5) = 5.98, P = 0.30, for night-time heartburn v 2 (5) = 8.23, P = 0.14 and acid regurgitation v 2 (5) = 4.48, P = à P < Complete relief of daytime and night-time heartburn was reported by 10 (22.7%) patients from the lansoprazole group vs. 8 (17.4%) and 12 (26.1%), respectively, from the omeprazole group (P ˆ N.S.). Acid regurgitation completely resolved in 9 (20.5%) patients receiving lansoprazole and 8 (17.6%) patients receiving omeprazole (P ˆ N.S.). The average time for complete resolution of daytime and night-time heartburn was 17.3 and 24.5 days, respectively, in the lansoprazole group and 14.3 and 22.1 days in the omeprazole group. Acid regurgitation completely resolved after a mean of 20.3 days in the lansoprazole group and 17.3 days in the omeprazole group. Antacid consumption Figure 3 demonstrates the mean antacid tablet consumption during the 6 weeks of the study. A total of 2146 antacid tablets were used by the lansoprazole group and 1718 by the omeprazole group throughout the study (P ˆ N.S., an average of 49 and 37 pills per patient, respectively). The proportion of patients reporting the use of any antacids was signi cantly greater among omeprazole patients in week 1 (P ˆ 0.045), but not in any other week. Paradoxically, these proportions did not match the number of pills taken by each group in a given week (i.e. there were often more pills taken by fewer subjects). The proportion taking any antacids in the lansoprazole and omeprazole groups, respectively, were 50% and 50% at baseline, 42% and 63% at week 1 (P ˆ 0.045), 51% and 57% at week 2 (P ˆ 0.6), 47% and 57% at week 3 (P ˆ 0.3), 42% and 54% at week 4 (P ˆ 0.2), 39% and 52% at week 5 (P ˆ 0.2) and 37% and 51% at week 6 (P ˆ 0.2). Side-effects Side-effects were reported by 18 (39.1%) patients from the omeprazole group and 14 (31.1%) from the lansoprazole group (P ˆ N.S.). There was no signi cant difference in the overall frequency of side-effects between the two groups or for each individual sideeffect. Table 4 demonstrates the comparative distribution of side-effects between the two groups. The percentages reporting side-effects in the two treatment groups were compared using v 2 tests. Gastrointestinal

7 OMEPRAZOLE VS. LANSOPRAZOLE TREATMENT IN GERD 1601 Figure 3. Mean antacid tablet consumption for both lansoprazole and omeprazole groups during the six week period of the study. There was no statistical difference (P > 0.05) in the mean antacid consumption for each of the 6 weeks. Antacid consumption remained relatively constant throughout the study. The absolute level of antacid use between the two drug groups was analysed by t-tests for each week, as well as v 2 tests on the proportions taking any antacids for each week. Table 4. A comparative distribution of side-effects between patients receiving omeprazole 40 mg q.d.s. and lansoprazole 30 mg b.d. Side-effect Lansoprazole (%) n =44 Omeprazole (%) n =46 P-value Diarrhoea 4 (9.1%) 8 (17.4%) 0.25 Abdominal 4 (9.1%) 4 (8.7%) 0.95 pain/discomfort Bloating/gas 3 (6.8%) 5 (10.9%) 0.50 Vomiting 3 (6.8%) 4 (8.7%) 0.74 Nausea 2 (4.6%) Headache 2 (4.6%) 3 (6.5%) 0.68 Dizziness 2 (4.6%) 1 (2.2%) 0.53 Fatigue 2 (4.6%) Constipation 0 2 (4.4%) 0.16 symptoms were the most common side-effect reported by both groups. In the omeprazole group, diarrhoea was the most common side-effect reported by eight (17.4%) patients, followed by bloating/gas by ve (10.9%) patients and vomiting and abdominal pain/discomfort by four (8.7%) patients. In the lansoprazole group, diarrhoea and abdominal pain/discomfort were equally common, affecting four (9.1%) patients, followed by vomiting and bloating/gas by three (6.8%) patients. Nausea and fatigue were only reported by patients receiving lansoprazole and constipation by patients receiving omeprazole. Side-effects were transient in nature and did not result in the drop-out of any patient from the study. One patient from the omeprazole group developed angina pectoris due to coronary artery disease and requested to be withdrawn from the study. In the lansoprazole group the frequency of side-effects reported per week remained stable during the rst 4 weeks (20%, 20%, 20%, 22.2%) and dropped to 8.9% and 11.1% in the fth and sixth weeks, respectively. In the omeprazole group the frequency of reported side-effects remained almost the same throughout the 6 weeks of the study (23.9%, 30.4%, 17.4%, 26.1%, 21.7% and 28.3%). In the fth and sixth weeks, omeprazole was associated with an increased odds of side-effects (OR ˆ 4.4, P ˆ and OR ˆ 4.3, P ˆ 0.025, respectively). DISCUSSION Most of the comparative studies between omeprazole and lansoprazole have focused on surrogate markers such as intragastric ph and gastric acid secretion. Not uncommonly, these studies were conducted in normal subjects. 3±8, 14 Only a handful of studies have evaluated clinically applicable end-points, such as oesophageal mucosal healing and symptom improvement. 9±13 This study is one of the rst to compare double dosing of lansoprazole vs. omeprazole, using clinical end-points. Demonstrating a better intragastric ph or acid secretion control, when comparing two drugs, may not automatically translate to a better clinical outcome. For patients and physicians alike, symptom improvement is the most important clinical end-point. Although, traditionally oesophageal mucosal healing has been the main clinical focus of many GERD treatment studies, symptom

8 1602 R. FASS et al. improvement is by far the most anticipated clinical outcome by patients. The results of this study demonstrated that omeprazole 40 mg daily was as effective as lansoprazole 30 mg twice daily in symptom control of GERD patients who failed the standard dose of proton pump inhibitor. Clinically, administration of double dose omeprazole in one 40 mg capsule rather than a split doseð20 mg twice dailyðstill provided similar symptom control to lansoprazole 30 mg twice daily. In contrast, Kuo et al. have demonstrated better gastric acid suppression using omeprazole 20 mg twice a day as compared to 40 mg once a day, concluding that a divided dose of omeprazole should be the preferred mode of administration. 14 However, in that study a surrogate markerðgastric ph, not symptom evaluation, was assessed. In addition, it was conducted in normal subjects and it also failed to demonstrate a signi cant difference in oesophageal acid exposure. Moreover, our study compared two different proton pump inhibitors rather than one proton pump inhibitor administered in different doses, which may also explain the lack of clinical advantage of lansoprazole. It should also be emphasized that the advantage of a split dose of proton pump inhibitor over once a day in controlling gastric acid secretion has never been demonstrated to provide an advantage in clinical practice, such as symptom improvement and oesophageal healing. Several studies comparing omeprazole and lansoprazole using standard doses, have documented a signi cantly better symptom control in the lansoprazole group 9, 10 that was limited solely to the rst week of therapy. This early clinical advantage has been explained by a better single dose bioavailability of lansoprazole 30 mg, which resulted in better gastric acid suppression over the rst few days of therapy. 16, 17 However, in this study we were unable to demonstrate a signi cant difference in symptom control between the two drugs during the rst week of therapy. This might be explained by the higher doses used in this study and the different patient population that included only those who had already failed a standard dose of lansoprazole. Additionally, we were unable to demonstrate a signi cant bene t of either drug over the other in daytime and night-time heartburn or acid regurgitation. The patient population in this study is considered to be a therapeutic challenge and a management dilemma for most physicians. Patients who have failed standard doses of proton pump inhibitor represent a small group with a more severe form of GERD. Thus, it was not surprising that symptom improvement rather than complete symptom resolution was demonstrated in most patients. Furthermore, antacid consumption was continued throughout the study by both groups of patients. Studies assessing the usage of antacids in patients receiving high doses of proton pump inhibitors are lacking. As compared to placebo, patients with GERD treated with standard doses of proton pump inhibitors consumed signi cantly less antacids. 9 It is unclear if increased doses of proton pump inhibitors result in an even greater reduction in antacid consumption. However, our study demonstrated a relatively high usage of antacids in both the lansoprazole and omeprazole groups, suggesting that the patients in our study were more resistant to proton pump inhibitor therapy and would require continuous antacid supplementation. On the other hand, by making antacids readily available to patients during a study, usage may be spuriously increased, either for typical GERD or even non-gerd related symptoms. Side-effects are not uncommon with proton pump inhibitor therapy and are reported by 30±50% of 10, 11, 19 patients. Similarly, in this study about onethird of the patients from each group experienced sideeffects. Furthermore, adverse events were commonly reported to involve the gastrointestinal tract. Interestingly, despite reports of diarrhoea being more common in patients treated with lansoprazole, in this study more patients on omeprazole developed diarrhoea as a sideeffect. 11 It is currently recommended that patients who continue to be symptomatic on standard doses of proton pump inhibitor should undergo an evaluation by ambulatory 24-h oesophageal ph monitoring on therapy. 2 However, in this study we did not perform ph testing and thus cannot determine how many of our subjects had abnormal acid exposure prior to randomization. An early report has suggested that in 61% of these patients, the total acid exposure time is within normal limits. 20 This intriguing nding does not exclude the need for a more potent anti-re ux therapy (double dose proton pump inhibitor), as some of the patients may need more profound acid suppression in order to obtain complete symptom relief. A limitation of this study was that patients were not blinded to the type of proton pump inhibitor that they received. The study was conducted at a VA setting and

9 OMEPRAZOLE VS. LANSOPRAZOLE TREATMENT IN GERD 1603 consequently includes mainly males of a relatively older age. Another possible limitation is the lack of information about the H. pylori status of our subjects. The presence of H. pylori may augment the therapeutic response of proton pump inhibitors. 21 However, in this study there were no age or demographic differences between the two groups, suggesting the possibility of a similar distribution of H. pylori. In conclusion, in patients who failed a standard dose of lansoprazole, administration of omeprazole 40 mg once daily resulted in similar symptom control as lansoprazole 30 mg twice daily. The use of once daily double dose proton pump inhibitor may be clinically more attractive due to better compliance. ACKNOWLEDGEMENTS This study was supported in part by a research grant from Astra-Zeneca, This study was presented in part at the annual meeting of the American College of Gastroenterology, Phoenix, Arizona, October REFERENCES 1 Hetzel DJ, Dent J, Reed WD, et al. Healing and relapse of severe peptic esophagitis after treatment with omeprazole. Gastroenterology 1988; 95: 903±12. 2 American Gastroenterological Association. American Gastroenterological Association medical position statement: guidelines on the use of esophageal ph recording. Gastroenterology 1996; 110: 1981±96. 3 Verdu EF, Fraser R, Armstrong D, Blum AL. Effects of omeprazole and lansoprazole on 24-hour intragastric ph in Helicobacter pylori positive volunteers. Scand J Gastroenterol 1994; 29: 1065±9. 4 Damman HG, Fuchs W, Richter G, Burkhardt F, Wolf N, Walter TA. Lansoprazole versus omeprazole: in uence on meal stimulated gastric acid secretion. Aliment Pharmacol Ther 1997; 11: 359±64. 5 Geus WP, Mulder GH, Nicolai JJ, VanDenBoomgaard DM, Lamers CBHW. Acid inhibitory effects of omeprazole and lansoprazole in Helicobacter pylori negative healthy subjects. Aliment Pharmacol Ther 1998; 12: 329±35. 6 Timmer W, Ripke H, Kleist P, et al. Effect of four lansoprazole dose levels and one dosage regimen of omeprazole on 24-hour intragastric ph in healthy subjects. Meth Find Exp Clin Pharmacol 1995; 17: 489±95. 7 Blum RA, Shi H, Karol MD, Greski-Rose P, Hunt RH. The comparative effects of lansoprazole, omeprazole, and ranitidine in suppressing gastric acid secretion. Clin Therapeutics 1997; 19: 1013±22. 8 Tolman KG, Sanders SW, Buchi KN, Karol MD, Jennings DE, Ringham GL. The effects of oral doses of lansoprazole and omeprazole on gastric ph. J Clin Gastroenterol 1997; 24: 65±70. 9 Castell DO, Richter JE, Robinson M, Sontag SJ, Haber MM, and the Lansoprazole Group. Ef cacy and safety of lansoprazole in the treatment of erosive re ux esophagitis. Am J Gastroenterol 1996; 91: 1749± Mee AS, Rowley JL, and the Lansoprazole Clinical Research Group. Rapid symptom relief in re ux oesophagitis: a comparison of lansoprazole and omeprazole. Aliment Pharmacol Ther 1996; 10: 757± Hatlebakk JG, Berstad A, Carling L, et al. Lansoprazole versus omeprazole in short term treatment of re ux oesophagitis. Scand J Gastroenterol 1993; 28: 224±8. 12 Janczewska I, Sagar M, Sjostedt S, Hammarlund B, Iwarzon M, Seensalu R. Comparison of the effect of lansoprazole and omeprazole on intragastric acidity and gastroesophageal re ux in patients with gastroesophageal re ux disease. Scand J Gastroenterol 1998; 33: 1239± Jaspersen D, Diehl KL, Schoeppner H, Geyer P, Martens E. A comparison of omeprazole, lansoprazole and pantoprazole in the maintenance treatment of severe re ux oesophagitis. Aliment Pharmacol Ther 1998; 12: 49± Kuo B, Castell DO. Optimal dosing of omeprazole 40mg daily: effect on gastric and esophageal ph and serum gastrin in healthy controls. Am J Gastroenterol 1996; 8: 1532±8. 15 Locke G, Talley N, Zinsmeister A. A new questionnaire for gastroesophageal re ux disease. Mayo Clinic Proc 1994; 69: 539± Fass R, Fennerty MB, Ofman JJ, et al. The clinical and economic value of a short course of omeprazole in patients with noncardiac chest pain. Gastroenterology 1998; 115: 42±9. 17 Fass R, Ofman JJ, Gralnek IM, et al. Clinical and economic assessment of the omeprazole test in patients with symptoms suggestive of gastroesophageal re ux disease (GERD). Arch Intern Med 1999; 159: 2161±8. 18 Liang K-Y, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986; 73: 13± Carling L, Axelsson CK, Forssell H, et al. Lansoprazole and omeprazole in the prevention of relapse of re ux oesophagitis: a long-term comparative study. Aliment Pharmacol Ther 1998; 12: 985± Fass R, Ofman JJ, Pulliam G, et al. Persistent symptoms of heartburn in patients on standard doses of proton pump inhibitors (PPI) are not due to acid re ux in most patients. Gastroenterology 1999; 116: G0694(Abstract). 21 Sachs G. Gastritis, Helicobacter pylori, and proton pump inhibitors. Gastroenterology 1997; 112: 1033±6.