Management of Reflux Esophagitis: Does the Choice of Proton Pump Inhibitor Matter?

J. Labenz; D. Armstrong; A. Leodolter I. Baldycheva

Int J Clin Pract. 2015;69(7):796-801.

Abstract and Introduction

Abstract

Background: Proton pump inhibitors (PPIs) are the treatment of choice for reflux esophagitis (RE). The effectiveness of PPIs throughout RE management, from healing to maintenance, has not been fully studied.

Aim: To compare esomeprazole with lansoprazole or pantoprazole for RE management using a management model.

Methods: Data from six studies comparing esomeprazole with lansoprazole or pantoprazole for healing (4–8 weeks) or maintenance of healing (6 months) of RE were incorporated into hypothetical management models to determine the proportion of patients in endoscopic remission after sequential healing and maintenance therapy, assuming that patients received the same PPI throughout. The number needed to treat (NNT) to achieve one more patient in remission with esomeprazole vs. other PPIs was estimated. The hypothetical model was validated using results from the EXPO study, which compared esomeprazole with pantoprazole for RE healing and maintenance.

Results: Overall, esomeprazole 40 mg produced higher rates of healing (life-table estimates) than lansoprazole 30 mg (82.4–92.6% vs. 77.5–88.8%; p < 0.01) or pantoprazole 40 mg (95.5% vs. 92.0%; p < 0.001) and higher rates of endoscopic and symptomatic remission at 6 months than lansoprazole (83.0–84.8% vs. 74.0–75.9%; p < 0.001; life-table estimates) or pantoprazole (70.9% vs. 59.6%; p < 0.0001; observed rates). In the hypothetical management model, the NNT for esomeprazole was 9 vs. lansoprazole and 8 vs. pantoprazole. The actual NNT for esomeprazole vs. pantoprazole in the EXPO study was 9 (95% confidence interval: 6; 16).

Conclusions: In this management model, esomeprazole was more effective than either lansoprazole or pantoprazole for maintaining remission after sequential healing and 6 months' maintenance therapy for RE.

Introduction

Gastroesophageal reflux disease (GERD) is a chronic condition^[1,2] associated with troublesome symptoms that significantly reduce quality of life.^[3,4] Possible complications include reflux esophagitis (RE), which may affect up to 50% of GERD patients.^[5] For patients with RE, short-term management goals include prompt resolution of symptoms and healing of erosions.^[6] However, following initial healing of RE, few patients remain in endoscopic and symptomatic remission unless provided with maintenance therapy.^[7] Therefore, long-term maintenance therapy is necessary for the prevention of recurrent RE and GERD symptoms in a high proportion of patients.

Proton pump inhibitors (PPIs) are the treatment of choice for both healing and maintenance therapy of RE.^[8–10] Numerous studies have demonstrated the efficacy of PPIs for initial healing of RE^[11–18] and numerous separate studies have demonstrated the efficacy of PPIs for maintaining endoscopic remission.^[8,10] The EXPO study,^[14,15] which compared esomeprazole with pantoprazole, is one of the few clinical trials to follow patients with RE through the entire course of management, i.e. initial healing of RE and subsequent maintenance of remission of healing. However, comparative data for other PPIs are lacking.

The aim of this analysis was to compare the outcome of RE treatment (i.e. initial healing for 4–8 weeks followed by maintenance for 6 months) with esomeprazole, lansoprazole or pantoprazole, using a hypothetical management model. This hypothetical model was then validated using data from the EXPO study (an 'actual' management model).^[14,15]

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References
Sidebar

Table 1. Estimated proportions of reflux esophagitis patients healed after 4–8 weeks' treatment with esomeprazole 40 mg, lansoprazole 30 mg, or pantoprazole 40 mg by baseline severity of reflux esophagitis (Kaplan–Meier analysis) ^†
Comparator/study	Treatment arm	Patients healed (%)^†
		Baseline LA grade
		A	B	C	D	All
Lansoprazole 30 mg
Castell et al. (11) and Vakil (19)	Esomeprazole	97.2 (n = 962)	92.0 (n = 1022)	88.3 (n = 482)	81.4 (n = 158)	92.6 (n = 2624)***
Castell et al. (11) and Vakil (19)	Lansoprazole	97.0 (n = 916)	91.0 (n = 1054)	77.3 (n = 477)	64.4 (n = 169)	88.8 (n = 2617)
Fennerty et al. (12)	Esomeprazole	Not applicable		80.3 (n = 390)^‡	67.6 (n = 108)^‡	82.4 (n = 498) **
Fennerty et al. (12)	Lansoprazole	Not applicable		74.9 (n = 403)^‡	66.3 (n = 98)^‡	77.5 (n = 501)
Pantoprazole 40 mg
Labenz et al. (EXPO study) (15)	Esomeprazole	97.3 (n = 523)	96.9 (n = 665) *	91.3 (n = 304)**	88.1 (n = 70)*	95.5 (n = 1562)***
Labenz et al. (EXPO study) (15)	Pantoprazole	97.1 (n = 478)	93.1 (n = 716)	87.6 (n = 303)	73.6 (n = 92)	92.0 (n = 1589)

^†Life-table estimates, except where stated. ^‡Observed. *p < 0.05, **p < 0.01, ***p < 0.001 vs. comparator (log-rank test).

Table 2. Estimated proportions of reflux esophagitis patients in remission at 6 months, following treatment with esomeprazole 20 mg, lansoprazole 15 mg, or pantoprazole 20 mg, by baseline severity of reflux esophagitis (Kaplan–Meier analysis)
Comparator/study	Treatment arm	Patients in remission (%)
		Baseline LA grade
		A	B	C	D	All
Lansoprazole 15 mg
Lauritsen et al. (16)	Esomeprazole	87.0 (n = 232)	83.0 (n = 269)**	75.0 (n = 95)*	77.0 (n = 19)*	83.0 (n = 615)****
Lauritsen et al. (16)	Lansoprazole	84.0 (n = 229)	72.0 (n = 278)	61.0 (n = 82)	50.0 (n = 20)	74.0 (n = 609)
DeVault et al.^† (20)	Esomeprazole	87.2 (n = 380)^‡		77.6 (n = 121)^‡		84.8 (n = 501)***
DeVault et al.^† (20)	Lansoprazole	78.7 (n = 369)^‡		68.4 (n = 131)^‡		75.9 (n = 500)
Pantoprazole 20 mg
Labenz et al. (EXPO study) (14)	Esomeprazole	89.8 (n = 447)***	88.0 (n = 607)****	84.1 (n = 263)****	69.4 (n = 60)	87.0 (n = 1377)****
Labenz et al. (EXPO study) (14)	Pantoprazole	81.0 (n = 451)	75.2 (n = 621)	67.3 (n = 251)	58.8 (n = 66)	74.9 (n = 1389)

^†Patients with LA grade A–D disease were included in the study but only data from patients with LA grade C or D disease were extracted for the hypothetical management model in Table 3. ^‡p-values were not calculated for LA grade A/B and C/D. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 vs. comparator (log-rank test).

Table 3. Management models: estimated proportions of patients who had healed reflux esophagitis (after 4–8 weeks' treatment) and remained in remission (endoscopic and symptomatic) after 6 months' maintenance therapy with esomeprazole or comparator proton pump inhibitors
Model/study	Treatment arm	Patients (%)					NNT*
		Baseline LA grade
		A	B	C	D	All
Hypothetical management models ^†
Castell et al. (11) and Lauritsen et al. (16)	Esomeprazole 40 mg/20 mg	84.6	76.4	66.2	62.7	76.9	9
Castell et al. (11) and Lauritsen et al. (16)	Lansoprazole 30 mg/15 mg	81.5	65.5	47.2	32.2	65.7	9
Fennerty et al.^‡ (12) and Lauritsen et al.^§ (16)	Esomeprazole 40 mg/20 mg					62.6	6
Fennerty et al.^‡ (12) and Lauritsen et al.^§ (16)	Lansoprazole 30 mg/15 mg					45.7	6
Fennerty et al.^‡ (12) and DeVault et al.^§ (20)	Esomeprazole 40 mg/20 mg					63.9	9
Fennerty et al.^‡ (12) and DeVault et al.^§ (20)	Lansoprazole 30 mg/15 mg					53.2	9
Labenz et al. (15) and Labenz et al. (14) (EXPO study)	Esomeprazole 40 mg/20 mg	87.4	85.3	76.8	61.2	83.1	8
Labenz et al. (15) and Labenz et al. (14) (EXPO study)	Pantoprazole 40 mg/20 mg	78.7	70.0	59.0	43.3	68.9	8
Actual management model ^¶
Labenz et al. (15) and Labenz et al. (14) (EXPO study)	Esomeprazole 40 mg/20 mg	76.4	72.8	61.6	52.6	70.9	9 (95% CI: 6; 16)
Labenz et al. (15) and Labenz et al. (14) (EXPO study)	Pantoprazole 40 mg/20 mg	68.1	58.2	53.7	44.0	59.6	9 (95% CI: 6; 16)

*Number needed to treat (NNT) with esomeprazole to avoid one treatment failure with the comparator agent. Calculation of the corresponding 95% confidence interval (CI) was not appropriate for the hypothetical management models, given that the NNT was based on the difference between two products of estimates (each of which was based on more than one study). ^†Life-table estimates (generated by multiplication of separate healing and remission rates from different studies or from different phases of the same study). ^‡Patients with Los Angeles (LA) grade C or D disease were included in the study. ^§Patients with LA grade A–D disease were included in the study but only data from patients with LA grade C or D disease were extracted for the maintenance model. ^¶Observed rates.