Jones Department of General Practice, UMDS (Guy's and St Thomas's), 5 Lambeth Walk, London, SE11 6SP, UK.
: To compare the relative efficacies of lansoprazole 15 mg o.m. and omeprazole 10 mg o.m. in relieving heartburn and epigastric pain in patients with acid-related dyspepsia. In addition, the study compared the safety profiles of the two treatments.
: This double-blind, parallel group, randomised, multicentre study was conducted in 52 general practices in the UK. A total of 609 patients was recruited, 562 of whom were eligible for inclusion in the intention-to-treat analysis. All of the patients had experienced at least mild heartburn or mild epigastric pain persistently on at least 4 of the previous 7 days; patients with severe symptoms were excluded. 283 patients received lansoprazole 15 mg and 279 received omeprazole 10 mg, both for 4 weeks. The main efficacy measure was relief of symptoms, based on physician assessments.
: In the intention-to-treat population, a complete relief of overall primary symptoms of dyspepsia was achieved after 2 weeks in 53% of patients receiving lansoprazole and in 41% of patients receiving omeprazole (P = 0.007). After 4 weeks, 59% of the lansoprazole group and 51% of the omeprazole group had achieved complete symptom relief (P = 0.078). Antacids were taken for additional relief of symptoms in fewer patients given lansoprazole compared to the omeprazole group in the third and fourth weeks (P = 0.035) and also significantly fewer antacids were taken by patients in the lansoprazole group compared with patients in the omeprazole group (P = 0.033). The proportion of patients reporting adverse events was similar in both groups.
: Low-dose lansoprazole is more effective than low-dose omeprazole in the treatment of patients with mild heartburn or epigastric pain in general practice.
Dyspepsia is a common problem in the community, where 40% of adults experience significant symptoms each year,1 and in general practitioners' surgeries, where it accounts for 4–5% of consultations.2 The role of the general practitioner in the management of dyspepsia is to employ a system of `triage', in which the minority of patients with serious disease (ulcers and cancer) are distinguished from the majority who are likely to have self-limiting or less serious problems.3 To a large extent, this distinction can be made on the basis of the patient's clinical history, and in particular the presence or absence of alarm symptoms (new symptoms in patients aged over 45 years, difficulty or pain on swallowing, weight loss, early satiety, systemic illness, anaemia, etc.), whose presence mandates specialist referral or urgent investigation.4
In the remainder of patients whose dyspeptic symptoms persist, investigation is often necessary; even after investigation, a substantial number of patients with dyspepsia have symptoms for which no organic cause can be found, but which respond to acid suppressing drugs. This is `acid-related dyspepsia', and significant numbers of patients require long-term—generally intermittent rather than maintenance—prescriptions of antisecretory drugs to control their symptoms. Clearly, a cost-effective approach to prescribing is required, and one way to secure this is to identify the lowest effective dose of an antisecretory drug that can be used to achieve symptom relief.
The proton pump inhibitor class of drugs has been shown to have excellent therapeutic and symptomatic effects in dyspepsia that is treated in general practice, and studies have shown that useful levels of acid suppression and symptomatic relief can also be obtained by lower dosages of these drugs, generally half that used for ulcer healing.5,6
In this study we have compared the relative efficacies of low-dose lansoprazole (15 mg o.m.) with omeprazole (10 mg o.m.) in the relief of heartburn and epigastric pain in patients with acid-related dyspepsia.
This was a Phase III, double blind, randomised, parallel group, multicentre study conducted in 52 general practices in the UK.
Patients aged 18–80 years presenting to their general practitioner (GP) with persistent mild heartburn or epigastric pain which had been present on at least 4 of the previous 7 days were eligible for inclusion. Patients aged 45 years or over were not entered until the possibility of gastric or oesophageal carcinoma had been considered and excluded. Patients were also excluded if they presented with severe symptoms of dyspepsia, or if they had been previously investigated and found to have confirmed gastro-oesophageal reflux disease or peptic ulcer disease. In addition, patients with symptoms suggestive of disordered gut motility (diffuse pain at night, hunger but with premature satiety, epigastric fullness, nausea and vomiting), irritable bowel syndrome (significantly abnormal bowel habit, lower abdominal pain, diffuse abdominal pain) or biliary tract disease (current or in the past) were excluded. Patients were not allowed to have taken a continuous course of ulcer healing treatment within the month prior to entry, nor were they allowed concomitant use of corticosteroids, anticoagulants, phenytoin or nonsteroidal anti-inflammatory drugs. Patients with suspected gastric or oesophageal carcinoma, a current or past history of malignancy of the upper gastrointestinal tract, previous gastric surgery, a history of alcoholism or drug abuse, those who were night-shift workers, had symptoms of upper gastrointestinal bleeding, had serious or uncontrolled illness, were pregnant or lactating women, were women of childbearing potential not willing to use a nonhormonal method of contraception, had previously participated in this study or were in another study, were also excluded.
At the baseline visit (Visit 1) an information leaflet was provided for the patient and written informed consent was given by the patient. Patients were allocated at random, using a computer generated randomization list, to double-blind treatment with either lansoprazole 15 mg once daily or omeprazole 10 mg once daily, to be taken in the morning before breakfast; the study medication was presented for 4 weeks in matching capsule form. Antacids were also provided for additional symptom relief and patients were instructed to take these as required.
Demographic information was recorded, and a medical history taken. The type of dyspepsia was categorized at baseline by the physician according to the primary presenting symptoms (Investigator Categorization). If the patient's predominant symptom was heartburn, the patient was categorized as having `reflux-like dyspepsia', if predominantly epigastric pain the category assigned was `ulcer-like dyspepsia' and if the patient had both symptoms, the patients was categorized as `both'. The physician assessed the patients symptoms over the week prior to entry in terms of daytime and night-time heartburn, daytime and night-time epigastric pain and overall symptoms, scoring them on a 10-point categorical scale. This scoring was used to provide a derived categorization for the type of dyspepsia (Symptom Categorization). The physician also recorded the presence or absence of a number of secondary dyspeptic symptoms. The patient was provided with a diary card on which to record, on a daily basis, daytime and night-time heartburn and epigastric pain, again on a 10-point categorical scale.
At the second visit, 2 weeks later, primary and secondary symptoms experienced the week prior to the visit were assessed by the physician. Any changes in concomitant medication or adverse events were recorded, and a count was made of all returned unused capsules and antacids. Patient diary cards were also collected and reviewed.
At the final visit, after 4 weeks of treatment, the physician repeated the assessments of the primary and secondary symptoms of dyspepsia, reviewed and noted any changes to concomitant medication and adverse events, counted the returned capsules and antacids, and collected the patient diary card.
Sample size and analysis
The number of patients required for the study was based on the assumption that 85% of the patients on lansoprazole would be relieved of their symptoms after 2 weeks of treatment, compared to 75% of the omeprazole patients. Hence, for a type I error of 5% (two-sided) with a power of 80%, approximately 250 patients per treatment arm would be needed in order to detect a treatment difference of 10%. Allowing for a drop-out rate of 20%, it was calculated that a further 100 patients would be needed.
The primary end-point for the analysis was the complete relief of symptoms using the physician's overall assessment of symptom relief after 2 weeks of treatment. Symptom scores were measured on a 10-point scale, but reduced to a dichotomous variable representing complete relief or not (no symptoms vs. symptoms). Additional analyses focused on the individual symptoms of dyspepsia and the safety of the treatments throughout the study. All statistical tests were two-sided and significance was assessed at the 5% level.
The results presented here are by intention-to-treat and include all patients randomized.
A total of 609 patients from 52 general practices were enrolled into the study over a 13-month period; of these, 38 were ineligible and a further 114 were nonevaluable at the second visit. A further 46 were not evaluable at Week 4, giving a total number of patients in the per protocol population of 457 at Week 2 and 411 at Week 4. A total of 562 patients were included in the intention-to-treat population, of whom 283 received lansoprazole and 279 omeprazole (see Figure 1).
The profile of the two treatment groups in terms of demographic features and clinical characteristics was comparable (see Table 1), and in addition there was no difference between the groups in their smoking and alcohol drinking habits.
Table 1. . Demographics and clinical characteristics of study patients
Most patients had a history of dyspepsia—78% of the lansoprazole group and 74% of the omeprazole group—and the mean duration of the current dyspepsia episode was 2 months in both groups. Less than half of the patients with a history of dyspepsia had had a previous barium meal or endoscopy—82 (39%) of the lansoprazole patients and 90 (44%) of the omeprazole patients—and these investigations were normal in the majority of patients (130, 74%). Approximately 60% of the subjects in each group had taken dyspepsia medication in the past, with antacids and alginates accounting for about 90% of the treatments.
Type of dyspepsia
The Investigator Categorization of the type of dyspepsia indicated that most patients (50%) presented with reflux-like dyspepsia as their predominant symptom; however, the Symptom Categorization calculated from the investigator's assessment of individual symptoms indicated that 57% of patients had both reflux-like and ulcer-like symptoms at presentation ( Figure 2).
Overall symptom relief
There were no differences in terms of overall symptoms at baseline, with most patients in both treatment groups presenting with mild to moderate (grade 4–5 on the 10-point scale) symptoms. A significant difference was seen between the two treatments after 2 weeks of treatment, with 53% of patients in the lansoprazole 15 mg group being symptom free compared to 41% of patients in the omeprazole 10 mg group (χ2 = 7.36, P = 0.007). Using the Odds ratios, the lansoprazole patients had over 1.5-fold greater chance of complete relief compared with those taking omeprazole (Odds ratio 1.6 (95% CI 1.15–2.25)).
After 4 weeks of treatment, the difference between the two treatment groups did not reach statistical significance; 59% of patients in the lansoprazole group were symptom free compared with 51% in the omeprazole group (χ2 = 3.11, P = 0.078).
The influence on overall symptom relief of six possible prognostic factors (treatment, smoking, alcohol, gender, antacid use and type of dyspepsia) was analysed using logistic regression. An interaction was seen between treatment and the type of dyspepsia identified by the investigator: at Week 4, patients with ulcer-like dyspepsia receiving lansoprazole were more likely to obtain symptomatic relief than those with reflux-like dyspepsia (χ2 = 6.64, P = 0.01) and, conversely, patients with reflux-like dyspepsia receiving omeprazole were more likely to become symptom-free than those presenting with ulcer-like dyspepsia. None of the other factors had a significant effect on relief of symptoms.
Other primary symptoms
A significant difference was seen between the two treatment groups after 2 weeks of treatment, with greater symptom control being seen in the lansoprazole patients ( Figure 3).
Notably, a significantly greater effect was seen in patients receiving lansoprazole compared with omeprazole with respect to daytime heartburn (n= 164 (70%) vs. n= 131 (58%); χ2 = 6.52, P = 0.011) and on daytime epigastric pain (129 (63%) vs. 88 (46%); χ2 = 10.67, P = 0.001); the observed differences for night-time heartburn and epigastric pain did not reach statistical significance.
After 4 weeks of treatment, more patients treated with lansoprazole were free of heartburn and epigastric pain than those treated with omeprazole, but this difference was not statistically significant.
Consumption of antacids
A comparison was made between the treatment groups with respect to the number of patients who had taken antacids, by each visit and with respect to the number of antacids taken between each visit. A significantly greater number of patients took antacids in the omeprazole group (n= 161, 58%) compared to the lansoprazole group (138, 49%) between the Week 2 and Week 4 visits (χ2 = 4.45, P = 0.035) (see Table 2). In addition, the mean number of antacid tablets taken in the omeprazole group (9.8) was significantly higher compared with the lansoprazole group (8.0) between the Week 2 and Week 4 visits (Z statistic = – 1.84, P = 0.033) (see Table 3).
Table 2. . Number of patients taking antacids in lansoprazole and omeprazole groups
Table 3. . Mean number of antacids taken in lansoprazole and omeprazole groups
Both drugs were well tolerated. Six-hundred and forty-six events were reported by 330 patients during the study. Ninety-one events were noted as drug-related in the lansoprazole group and 135 in the omeprazole group. Approximately three-quarters of the events reported were gastrointestinal in nature, a number of which are also considered to be secondary symptoms of dyspepsia and part of the intermittent pattern of dyspepsia. Four patients reported serious adverse events, all in the omeprazole group, of which only one (severe chest pain) was noted as being related to treatment, and this relationship was deemed remote. There were no significant differences between the treatments, although the trend was towards a higher incidence of more severe adverse events in the omeprazole group.
This study indicates that daily treatment with 15 mg lansoprazole results in a significantly higher proportion of patients reporting complete relief of dyspeptic symptoms than those treated with a low dose of omeprazole, 10 mg daily. This superior efficacy appears to apply both to ulcer-like symptoms and to heartburn experienced during the daytime. In addition, both drugs appear to be well-tolerated and safe in this patient population.
Lansoprazole treatment resulted in a significantly higher proportion of patients in the intention-to-treat population experiencing relief of overall symptoms after 2 weeks of treatment compared with omeprazole treatment; this was the primary end-point of the study. The reduction of the 10-point scale to a dichotomous variable—symptoms vs. no symptoms—was not considered to bias the results in favour of either treatment arm. After 4 weeks of treatment, a greater proportion of patients receiving lansoprazole had achieved complete relief of symptoms compared with the patients on omeprazole, although this difference did not reach statistical significance.
In addition, some individual symptoms of dyspepsia were relieved in a higher proportion of patients taking lansoprazole compared with those taking omeprazole. A greater proportion of patients taking lansoprazole experienced relief of their daytime heartburn symptoms after 2 weeks of treatment compared with those patients taking omeprazole. Similarly, the relief of daytime epigastric pain was achieved in a higher proportion of patients after 2 weeks in those taking lansoprazole compared with those taking omeprazole. There were no significant differences between the treatment groups with respect to the proportion of patients in whom night-time symptoms of epigastric pain or heartburn were relieved. However, the higher proportion of symptom-free patients in the lansoprazole group compared with the omeprazole group may imply that a smaller number of patients treated with lansoprazole were being woken in the night by their heartburn, which is arguably a specific advantage in favour of lansoprazole. It is possible that the difference between relief of daytime and night-time symptoms reflects the timing of administration of the study medication.
The significantly lower proportion of patients in the lansoprazole group taking antacids compared with the proportion in the omeprazole group reflects the higher efficacy afforded by lansoprazole treatment in relieving symptoms. This is further confirmed by the significantly greater number of antacids consumed by the patients taking omeprazole compared with those taking lansoprazole.
Prognostic factor analyses identified the type of dyspepsia as a significant factor in the relief of overall primary symptoms, so that patients treated with lansoprazole who had ulcer-like dyspepsia were more frequently symptom-free than those with reflux-like dyspepsia, whereas patients who had reflux-like dyspepsia who were treated with omeprazole were more frequently symptom-free than those who had ulcer-like dyspepsia. This was confirmed by the observation that a significantly higher proportion of patients with ulcer-like dyspepsia who were treated with lansoprazole had overall symptom relief compared with those patients treated with omeprazole.
The results of this study reflect those observed when a higher dose of lansoprazole (30 mg once daily) was used to treat patients with dyspepsia.7,8 Whilst a direct comparison of the studies cannot be made, since the comparator drug in the previous study was ranitidine, similar results have been observed with respect to the proportion of patients receiving lansoprazole who had relief of daytime symptoms. However, the proportion of patients with relief of night-time heartburn and epigastric pain was higher for patients in the previous study (i.e. during treatment with lansoprazole 30 mg) compared with that observed in patients in the current study.
These results will be of interest to general practitioners faced with large numbers of patients with non-ulcer dyspepsia requiring antisecretory therapy. There are many concerns about the rising costs of medication for the treatment of acid-related disorders, so it would be helpful to know that a lower dose of lansoprazole provides a good level of symptomatic relief in this patient group. Although this study did not investigate the pattern of tablet taking in these patients, who are likely to take medication in response to symptoms rather than on a daily basis,9 the study confirms that lansoprazole 15 mg represents an effective approach to the management of acid-related dyspepsia in general practice.
The authors wish to thank the following general practitioners who participated in this study: Dr S. C. Arora, Brixton; Dr M. D. Ballon, Levenshulme; Dr P. D. Barnard, Wirral; Dr C. Barrett, Rutherglen; Dr L. Bidwell, Dumbarton; Dr M. Blagden, Chesterfield; Dr R. G. Bowen, Cardiff; Dr N. Brennan, Barnstable; Dr M. C. Brown, Westhoughton; Dr J. P. O. Chapman, Sutton Coldfield; Dr G. Charlwood, Tunbridge Wells; Dr P. T. Chue Hong, Kinghorn; Dr M. R. B. Cottrill, Leigh; Dr I. D. Cracknell, Hinckley; Dr F. R. Cranfield, St Dials; Dr M. Davidson, Dronfield; Dr J. F. Dromey, Abergele; Dr V. S. Dwivedi, Rock Ferry; Dr J. D. A. Evans, Abergele; Dr J. D. R. Gordon, Glasgow; Dr P. J. Griffiths, Middleton; Dr M. G. Grillage, Shirley; Dr B. Hamilton, Stanstead; Dr A. Harris, Sutton Coldfield; Dr G. Hosie, Glasgow; Dr M. Hossain, Peckham; Dr T. M. Hughes, St Dials; Dr J. S. Hutchison, Glasgow; Dr W. R. S. Jones, Deeside; Dr I. C. Kemp, Emmer Green; Dr L. King, Wakefield; Dr A. G. Lane, Cardiff; Dr H. L. Ling, Flitwich; Dr E. Livingston, Glasgow; Dr J. D. McCallum, Edinburgh; Dr D. F. McCann, Parkstone; Dr G. S. McGregor, Bishop Auckland; Dr H. J. Patel, London; Dr R. Pool, Woking; Dr J. M. Price, Chichester; Dr E. Raubitschek, Addingham; Dr I. Reid, Glasgow; Dr J. A. Renwick, Claughton; Dr G. Roberts, Parkstone; Dr J. D. V. Roberts, Paignton; Dr I. J. Robertson, Knowle; Dr A. Sanderson, Spennymoor; Dr S. K. Sharma, Chorlton; Dr K. Sher, Birstall; Dr R. W. Spence, Redland; Dr Vilath, London; Dr D. E. Ward, Poole; Dr M. A. Whitehead, Thornton Heath.
The preliminary results were presented at the meeting of the American Gastroenterological Association, Washington, in May 1997.
This study was supported by Wyeth Laboratories, UK.