Healing of NSAID-Associated Gastric Ulcers in Patients Continuing NSAIDs

views the best pictures

Results

Patients and Disposition

The study was conducted from February 2001 to April 2003. In all, 440 patients were randomized at 75 study sites as follows: Bulgaria, 89 patients from nine sites; Indonesia, 51 patients from two sites; Romania, 45 patients from seven sites; the Ukraine, eight patients from three sites, and the US, 247 from 54 sites. Patient disposition in the study is shown in Figure 1. Of the 440 patients randomized, 30 were not included in the primary efficacy analysis because of no study drug use, the lack of a documented baseline GU, or a GCP violation at a single site identified after the site was audited. Of 410 patients included in the primary efficacy analyses, 88 (21%) were in Bulgaria, 51 (12%) were in Indonesia, 45 (11%) were in Romania, 8 (2%) were in the Ukraine and 218 (53%) were in the United States.

Figure 1.  (click image to zoom)

Patient disposition. GCP, good clinical practice; NSAID, non-steroidal anti-inflammatory drug; GU, gastric ulcer. * Some patients had more than one type of protocol deviation.      

Baseline demographics and characteristics for patients included in the efficacy analysis are summarized in Table 1 . The study population included more women than men, and most patients were white. The proportion of patients with positive H. pylori status, as determined by histology, was similar among the three groups and included 13.8% (30 of 218) of patients in the United States and 40.1% (77 of 192) of patients in the non-US countries. The most common chronic condition for which patients were taking NSAIDs was osteoarthritis (51% of the patients). Across the three treatment groups, most patients (88%) used non-selective NSAIDs; the most commonly used non-selective NSAIDs were aspirin, diclofenac, ibuprofen, and naproxen. Of these patients, approximately 14% (51/362) used only low-dose aspirin (80 to 325 mg per day). Baseline OGD findings showed that the mean maximum GU size was approximately 8.2 mm, and 129 patients (31%) had GUs > /=10 mm ( Table 2 ). Also, four patients (< 1%) had baseline GUs of maximum size < 5 mm. The GU healing results for these four patients were included in the mITT analysis for completeness. Excluding the healing results for these patients would not have changed the results of the analysis. Only 34 patients included in the efficacy population had concurrent baseline DUs ( Table 2 ).

Overall, for patients in the efficacy analysis, compliance with study medication was 97% (396 of 410 patients) and was similar among the three treatment groups. NSAID compliance was also generally similar among the treatment groups with an overall compliance rate of 88% (361 of 410) of the patients.Efficacy

Table 3 reports the GU healing rates at 4 and 8 weeks. Although the observed GU healing rates at week 8 were numerically higher with the esomeprazole treatments than with ranitidine, these results were not statistically significant. In contrast, at week 4, the observed GU healing rates were significantly higher for patients treated with esomeprazole 40 and 20 mg than for patients treated with ranitidine ( Table 3 ). The observed GU healing rates in the esomeprazole groups compared with the ranitidine group are shown for subgroups evaluated according to age group (Figure 2), baseline H. pylori status determined by histology (Figure 3), and non-selective vs. selective baseline NSAID used (Figure 4).

Figure 2.  (click image to zoom)

Gastric ulcer (GU) healing rates by age. E40, esomeprazole 40 mg once daily; E20, esomeprazole 20 mg once daily; R150, ranitidine 150 mg twice daily (mITT population).      

Figure 3.  (click image to zoom)

Gastric ulcer (GU) healing rates by baseline Helicobacter pylori status determined by histology. E40, esomeprazole 40 mg once daily; E20, esomeprazole 20 mg once daily; R150, ranitidine 150 mg twice daily. †Baseline H. pylori status was missing for one patient in the esomeprazole 40-mg group (mITT population).      

Figure 4.  (click image to zoom)

Gastric ulcer (GU) healing rates by baseline non-steroidal anti-inflammatory drug (NSAID) type. The non-selective NSAID group includes patients who were taking a cyclooxygenase (COX) 2-selective NSAID if they were also taking a non-selective NSAID (including low-dose aspirin). E40, esomeprazole 40 mg once daily; E20, esomeprazole 20 mg once daily; R150, ranitidine 150 mg twice daily. †One patient in the ranitidine 150-mg group was not taking an NSAID before or during the study (mITT population).      

The observed GU healing rates were reanalysed including the 15 patients (five in each treatment group) that were excluded from the mITT population because of GCP violations. The results of this reanalysis provided healing rates that were almost identical to those seen in Table 3 . Therefore, removing these patients from the mITT analysis did not affect the findings of this study.

As a secondary analysis of GU healing, the estimated GU healing rates through the final visit (based on Kaplan-Meier estimation) were 92.1% (95% CI, 87.4%-96.8%), 94.6% (95% CI, 90.7%-98.5%) and 89.2% (95% CI, 83.7%-94.7%) in the esomeprazole 40-mg, esomeprazole 20-mg and ranitidine groups, respectively. Estimated GU healing rates through week 4 (based on Kaplan-Meier estimation) were 71.6% (95% CI, 63.9%-79.4%), 75.2% (95%CI, 67.8%-82.5%), and 58.4% (95% CI, 49.9%-66.8%) for the esomeprazole 40-mg, esomeprazole 20-mg, and ranitidine groups, respectively. Comparisons of the time-to-event curves showed that the time to first healing of GUs was significantly different for esomeprazole 40 mg and esomeprazole 20 mg compared with ranitidine, (P = 0.047 and P = 0.002, respectively).

For the 34 patients in the efficacy analysis who had concurrent DUs at baseline, the DU healing rates at week 8 were 90% (nine of 10 patients) with esomeprazole 40 mg, 68.8% (11 of 16 patients) with esomeprazole 20 mg, and 87.5% (seven of eight patients) with ranitidine.Safety

The safety population included 432 patients: 140 patients in the esomeprazole 40-mg group, 145 patients in the esomeprazole 20-mg group, and 147 patients in the ranitidine 150-mg group. The overall percentage of patients with AEs in this study was 56% (79/140) with esomeprazole 40 mg, 58% (84/145) with esomeprazole 20 mg and 58% (85/147) with ranitidine 150 mg. The number of patients with AEs considered by the site investigator to be related to study drug was also similar among the three treatment groups: 12 (9%) for esomeprazole 40 mg, 13 (9%) for esomeprazole 20 mg and 10 (7%) for ranitidine 150 mg. The most commonly reported AEs for esomeprazole 40 mg, esomeprazole 20 mg and ranitidine 150 mg were gastrointestinal and included gastritis [22 (16%), 25 (17%) and 25 (17%) patients, respectively], flatulence [18 (13%), 27 (19%) and 20 (14%), patients respectively], new onset or worsening of existing dyspepsia [14 (10%), 19 (13%) and 18 (12%) patients, respectively], and new onset or worsening of existing nausea [17 (12%), 11 (8%) and 17 (12%) patients, respectively].

Fourteen of the 440 randomized patients had 16 serious AEs (four patients in the esomeprazole 40-mg group, six in the esomeprazole 20-mg group and four in the ranitidine group), but none was the same and none was considered related to treatment. There was one death that occurred due to an unknown cause on the third day of treatment with esomeprazole 20 mg; the investigator did not consider the death to be related to study medication. Of 440 patients, 17 (3.9%) discontinued the study due to AEs; the number in each treatment group was similar. There were no clinically relevant trends in the results of laboratory tests, physical examinations or vital signs.  Printer- Friendly Email This

Aliment Pharmacol Ther.  2007;26(8):1101-1111.  ©2007 Blackwell Publishing
This is a part of article Healing of NSAID-Associated Gastric Ulcers in Patients Continuing NSAIDs Taken from “Nexium Generic Esomeprazole” Information Blog