Reflux Esophagitis: Does the Choice of PPI Matter?
Reflux Esophagitis: Does the Choice of PPI Matter?
When all patients were considered together, the proportion of patients achieving healing of RE was significantly higher for esomeprazole than for either lansoprazole or pantoprazole [life-table estimates, p < 0.01 (Table 1)].
While healing rates tended to decrease as the baseline severity of RE increased, they remained higher with esomeprazole than with lansoprazole or pantoprazole across all grades of RE (Table 1). Healing rates were significantly higher (p < 0.05) with esomeprazole than with pantoprazole for patients with LA grade B, C or D at baseline (Table 1).
Among all patients, life-table estimates of the proportions of patients maintaining healing and symptomatic remission at 6 months were significantly higher with esomeprazole than with either lansoprazole [83.0% vs. 74.0% in the Lauritsen study; p < 0.0001 and 84.8% vs. 75.9% in the DeVault study; p < 0.001] or pantoprazole [87.0% vs. 74.9% in the EXPO study; p < 0.0001.]
Esomeprazole was associated with higher endoscopic and symptomatic remission rates than lansoprazole and pantoprazole, regardless of the severity of RE prior to healing therapy (Table 2). With the exception of LA grade A, observed remission rates were significantly greater (p < 0.05) with esomeprazole than with lansoprazole across all grades of RE (Table 2). Esomeprazole was also significantly more effective than pantoprazole for LA grades A, B and C (Table 2).
Management model results are shown in Table 3 and Figure 1. For lansoprazole, the hypothetical management model yielded an NNT of 9, indicating that esomeprazole therapy would avoid one treatment failure for every 9 patients with RE of LA grade A–D treated with lansoprazole. As data were also available for studies including patients with more severe disease (LA grade C–D), it was possible to generate an NNT for this population. The hypothetical management model yielded an NNT of 6 for these patients.
(Enlarge Image)
Figure 1.
Management models: overall proportion 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. *Life-table estimates; Observed rates
For pantoprazole (the EXPO study), the hypothetical management model NNT was 8 for esomeprazole vs. pantoprazole. The EXPO trial results (the 'actual' management model) yielded a similar NNT of 9 (95% confidence interval: 6; 16) for esomeprazole vs. pantoprazole.
Results
Healing Rates
When all patients were considered together, the proportion of patients achieving healing of RE was significantly higher for esomeprazole than for either lansoprazole or pantoprazole [life-table estimates, p < 0.01 (Table 1)].
While healing rates tended to decrease as the baseline severity of RE increased, they remained higher with esomeprazole than with lansoprazole or pantoprazole across all grades of RE (Table 1). Healing rates were significantly higher (p < 0.05) with esomeprazole than with pantoprazole for patients with LA grade B, C or D at baseline (Table 1).
Relapse Rates During Maintenance Therapy
Among all patients, life-table estimates of the proportions of patients maintaining healing and symptomatic remission at 6 months were significantly higher with esomeprazole than with either lansoprazole [83.0% vs. 74.0% in the Lauritsen study; p < 0.0001 and 84.8% vs. 75.9% in the DeVault study; p < 0.001] or pantoprazole [87.0% vs. 74.9% in the EXPO study; p < 0.0001.]
Esomeprazole was associated with higher endoscopic and symptomatic remission rates than lansoprazole and pantoprazole, regardless of the severity of RE prior to healing therapy (Table 2). With the exception of LA grade A, observed remission rates were significantly greater (p < 0.05) with esomeprazole than with lansoprazole across all grades of RE (Table 2). Esomeprazole was also significantly more effective than pantoprazole for LA grades A, B and C (Table 2).
Management Models
Management model results are shown in Table 3 and Figure 1. For lansoprazole, the hypothetical management model yielded an NNT of 9, indicating that esomeprazole therapy would avoid one treatment failure for every 9 patients with RE of LA grade A–D treated with lansoprazole. As data were also available for studies including patients with more severe disease (LA grade C–D), it was possible to generate an NNT for this population. The hypothetical management model yielded an NNT of 6 for these patients.
(Enlarge Image)
Figure 1.
Management models: overall proportion 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. *Life-table estimates; Observed rates
For pantoprazole (the EXPO study), the hypothetical management model NNT was 8 for esomeprazole vs. pantoprazole. The EXPO trial results (the 'actual' management model) yielded a similar NNT of 9 (95% confidence interval: 6; 16) for esomeprazole vs. pantoprazole.
