Dr M. B. Fennerty, Division of Gastroenterology, Oregon Health & Science University, Mail Code PV-310, 3181 SW Sam Jackson Park Rd, Portland, OR 97239-3098, USA. E-mail: email@example.com
Background : Secondary analyses from previous studies indicated that esomeprazole was more effective than lansoprazole and omeprazole in healing moderate or severe (Los Angeles grades C or D) erosive oesophagitis (EE).
Aim : To compare prospectively healing rates with esomeprazole vs. lansoprazole in patients with moderate to severe EE.
Methods : In this multicentre, randomized, double-blind, parallel-group trial, adult patients with endoscopically confirmed moderate or severe EE received esomeprazole 40 mg (n = 498) or lansoprazole 30 mg (n = 501) once daily for up to 8 weeks. The primary end point was EE healing through week 8. Secondary assessments included investigator-assessed resolution of symptoms and safety and tolerability.
Results : Time to healing was significantly different (P = 0.007), favouring esomeprazole. Estimated healing rates at week 8 were 82.4% with esomeprazole 40 mg and 77.5% with lansoprazole 30 mg. Heartburn resolved at week 4 in 72% and 64% of patients who received esomeprazole and lansoprazole, respectively (P = 0.005). Control of other GERD symptoms was similar between treatments. Both treatments were well tolerated.
Conclusions : With 8 weeks’ treatment, esomeprazole 40 mg once daily heals moderate to severe EE faster and in more patients, and resolves heartburn in more patients after 4 weeks of treatment, than lansoprazole 30 mg once daily.
Erosive oesophagitis (EE), caused by gastro-oesophageal reflux, is a common medical problem. Data from large clinical trials indicate that the prevalence of moderate to severe [Los Angeles (LA) Classification grades C and D] EE at the time of screening is as high as 13–15% or more in patients with frequent heartburn.1, 2 The severity of oesophagitis is in large part affected by the duration of oesophageal acid exposure, with greater acid exposure being associated with more severe mucosal lesions.3 Currently, therapy for gastro-oesophageal reflux disease (GERD) largely focuses on the pharmacological reduction of gastric acid secretion. This approach to therapy for GERD has been effective in eliminating many of the symptoms of GERD, such as heartburn and acid regurgitation, and effectively heals the oesophageal mucosa in the majority of patients. The goal of acid suppression therapy in patients with EE is to achieve sufficient acid control necessary to relieve GERD-related symptoms and optimize oesophageal healing rates. Oesophagitis healing rates have been correlated with the amount of time that the intragastric pH level can be maintained above a pH of 4.4
Proton pump inhibitors (PPIs) are widely recognized as the most effective agents for treating GERD and have been the mainstay of pharmacological GERD management.5 These drugs produce greater inhibition of intragastric acid secretion, more rapid symptom control, and better healing of EE than H2-receptor antagonists.5–7 Esomeprazole, the S-enantiomer of omeprazole (a racemic mixture of R- and S-enantiomers) has increased bioavailability compared with the R-enantiomer and omeprazole because of stereoselective differences in hepatic metabolism by cytochrome P-450 isoenzymes 2C19 and 3A4.8 Treatment with esomeprazole resulted in significantly greater inhibition of intragastric acid secretion vs. other PPIs as well as higher EE healing rates than other currently available PPIs at doses approved for healing EE.1, 2, 9–11 In post-hoc analyses, the differences in healing rates between esomeprazole and other PPIs observed in these studies were more pronounced in patients with moderate to severe disease (LA grades C and D), which typically is more difficult to heal with antisecretory therapy than less severe states of disease (LA grades A and B) [data on file, AstraZeneca LP (Wilmington, DE, USA), reference codes, DA-NEX-17, DA-NEX-18, DA-NEX-24, DA-NEX-25]. One of these previous analyses showed that the observed healing rates after 8 weeks of treatment in patients with moderate or severe EE (LA grades C or D) were 86.6% with esomeprazole 40 mg and 73.9% (P = 0.0013) with lansoprazole 30 mg (data on file, AstraZeneca LP, reference code, DA-NEX-18 and Castell et al.9) No previous study has used healing of moderate to severe EE as the primary outcome measure. We hypothesized that, because of the better antisecretory activity of esomeprazole 40 mg compared with that of lansoprazole 30 mg and because the results of the previously described post-hoc analyses showed better healing in patients with LA grades C and D EE using esomeprazole, healing of moderate to severe EE would be more effective using esomeprazole 40 mg vs. lansoprazole 30 mg when evaluated as a primary end point of a randomized clinical trial. Thus, the primary objective of this study was to compare prospectively the healing rates achieved with esomeprazole 40 mg vs. lansoprazole 30 mg over 8 weeks among patients with endoscopically determined moderate or severe EE (LA grades C or D). Secondary objectives included healing at 4 weeks, comparison of the resolution of investigator-evaluated GERD symptoms and patient-recorded heartburn symptoms, and evaluation of the safety and tolerability of these therapies.
Materials and methods
The primary inclusion criteria for this randomized, double-blind, double-dummy, parallel-group, multicentre study (D9612L00046/Study 322), were LA grades C or D EE confirmed by an oesophagogastroduodenoscopy within 1 week before randomization and heartburn (described as a burning feeling, rising from the stomach or lower part of the chest up towards the neck) for ≥2 of the 7 days before randomization. Qualified patients received up to 8 weeks of oral therapy with esomeprazole 40 mg or lansoprazole 30 mg taken once daily in the morning with a glass of water before breakfast. Study medications were provided as single capsules. To maintain double-blind conditions, patients were instructed to take both an active treatment, a capsule containing esomeprazole or lansoprazole, and a dummy capsule that was identical to the comparator. Patients were enrolled sequentially within each site using predetermined, computer-generated randomization schedules, using a block size of 4 and by-site allocation that were concealed from both the patients and investigators by using an opaque panel on the study drug label assigned to that patient that could only be broken by the investigator in case of medical emergency and, if done, was documented. Endoscopy was repeated after approximately 4 weeks of treatment, and patients with no evidence of EE, based on the LA classification of oesophagitis, were considered healed and were withdrawn because they had achieved the end point of healing. The patients who still had EE continued their assigned treatment and returned after 4 more weeks (8 weeks from study entry) for a repeat endoscopic evaluation. Once again, the absence of EE based on the LA classification was used as the determinant of healing.
Patients also received 200-mg antacid tablets [Gelusil; Warner-Lambert Consumer Healthcare (Parke-Davis), Morris Plains, NJ, USA] and were instructed to take no more than six tablets per day only for the relief of acute intolerable heartburn symptoms. Patients returned all unused study drugs at each visit and returned capsule counts were used to assess treatment compliance.
Independent Ethics Committee or Institutional Review Board approval was obtained before enrolment of any patient into the study, which was performed in accordance with the Declaration of Helsinki, Good Clinical Practice, and applicable regulatory requirements. Signed informed consent was obtained from all patients before study entry.
Men and non-pregnant, non-lactating women between the ages of 18 and 75 years were included. A negative pregnancy test at study baseline and a medically acceptable form of birth control were required for women of childbearing potential. Patients were excluded if they experienced any bleeding disorder or signs of gastrointestinal bleeding at the time of the baseline endoscopy or within 3 days before randomization; had a history of gastric or oesophageal surgery, except for simple closure of perforated ulcer; or had current or historical evidence (within 3 months) of Zollinger–Ellison syndrome, primary oesophageal motility disorders (achalasia, scleroderma and/or primary oesophageal spasm), inflammatory bowel disease, pancreatitis, malabsorption, generalized bleeding disorders resulting from haemorrhagic diathesis, oesophageal stricture, duodenal ulcer, gastric ulcer, evidence of upper gastrointestinal malignancy, endoscopic Barrett's oesophagus (>3 cm), significant dysplastic changes in the oesophagus, or any other severe concomitant disease. Patients who used a PPI within 28 days before the baseline visit or daily histamine H2-receptor antagonists in doses exceeding standard approved prescription-strength doses within the last 3 months were excluded, as were patients who had need for continuous concurrent therapy with warfarin or other anticoagulants, prostaglandin analogues, antineoplastic agents, salicylates (unless ≤165 mg daily for cardiovascular prophylaxis), steroids (oral or intravenous), pro-motility drugs, sucralfate, non-steroidal anti-inflammatory drugs, phenytoin, or tegaserod. Use of Helicobacter pylori eradication therapy or a concomitant medication that depends on the presence of gastric acid for optimal absorption was not permitted. Compliance assessments were based on returned drug counts. Patients were considered compliant if they consumed >90% of the study drug.
Endoscopic confirmation of the EE healing was the primary efficacy assessment. An oesophagogastroduodenoscopy was performed at baseline, at week 4, and at week 8 (if EE was not healed at the week-4 visit). Healing was defined as no mucosal breaks according to the LA classification. It was assumed for this study that patients healed at week 4 (but who did not continue in the study) would have remained healed through week 8 (if they had continued therapy). The proportion of patients healed by week 8 was estimated by the Kaplan–Meier method. The observed healing rates at weeks 4 and 8 were also calculated.
Investigators evaluated GERD symptoms (heartburn, acid regurgitation, dysphagia, and epigastric pain) for the 7 days before the patient's study visit on a four-point scale from ‘none’ to ‘severe’ [incapacitating symptom, with inability to perform normal activities (including sleep)]. For each symptom, resolution was defined as no symptoms (severity = none) during the 7 days before the clinical visit.
During the first 4 weeks of the study, patients kept a diary in which they assessed their heartburn symptoms each morning by rating the most intense episode during the previous 24 h using the same scale as that used by the investigators to assess symptoms. Resolution of heartburn was defined as no symptoms. Sustained resolution of heartburn was defined as no symptoms for 7 consecutive days. The first day of the 7 days was used to calculate the days to sustained resolution. In their diaries, patients also indicated whether or not they had nocturnal heartburn (during sleeping hours). From these diary data, the proportions of heartburn-free days and heartburn-free nights were calculated for each patient. Patients were asked to complete diary cards only during the first 4 weeks because of the anticipated discontinuation of many patients after the week-4 visit due to healing of EE.
Adverse events spontaneously reported by the patient or in response to an open question from the study personnel or revealed by physical or laboratory observations were recorded. Investigators recorded the intensity (mild, moderate, or severe) of adverse events and whether or not there was a reasonable possibility that the adverse event was caused by the study drug. Fasting blood samples for serum chemistry and haematological measurements were obtained by standardized techniques at baseline and final visits and were analysed by a central laboratory. An H. pylori serology test (FlexSure HP; Beckman Coulter, Inc., Fullerton, CA, USA) was performed at the site during the baseline visit. A pregnancy test (dipstick in urine) was performed for women of childbearing potential. Physical examinations and vital signs were recorded and evaluated.
Efficacy analyses included all randomized patients who took at least one dose of study medication and had LA grade C or D EE (intention-to-treat analysis). The safety evaluation included all patients who took at least one dose of study medication.
All statistical tests were two-sided. A difference between the two treatment groups was considered statistically significant if P ≤ 0.05. The analysis was performed using SAS software, version 8.2 (SAS Institute Inc, Cary, NC, USA).
The primary efficacy variable was healing of EE by week 8. The EE healing rate through 8 weeks was estimated by the Kaplan–Meier method. The primary analysis comparing the time-to-healing curves of esomeprazole 40 mg and lansoprazole 30 mg was performed using a log-rank test, which considers both the time of healing and the magnitude of difference between treatment groups. In addition, observed healing rates were calculated for weeks 4 and 8 separately. Healed patients who discontinued from the study at week 4 were included in a cumulative week-8 healing rate, on the basis of the assumption that they would remain healed if treatment continued for another 4 weeks.
The secondary analysis comparing the observed healing rates between treatment groups was performed using a Cochran–Mantel–Haenszel (CMH) test that was stratified by baseline severity of EE. The proportions of patients with resolution of GERD symptoms at week 4 as recorded by investigators were compared between treatment groups using a CMH test stratified by the baseline severity of that symptom. Concurrence between investigator-assessed resolution of GERD symptoms (resolved, not resolved) and EE healing status (healed, not healed) was summarized. Time-to-event curves for time to sustained resolution and time to first resolution of heartburn from diary data were compared between treatments using a log-rank test. Analysis of variance (anova) was used to compare treatment groups for the percentage of heartburn-free days and heartburn-free nights.
Because the planning of this study was based on the subgroup results of a previous study,9 the results from that study were used as a reference for the sample-size calculation. An estimated 474 patients randomized to each group were needed to detect a 10% difference in EE healing rate (85% in the esomeprazole 40-mg group and 75% in the lansoprazole 30-mg group), with a 5% significance level, 95% power, and allowing up to 10% withdrawals.
Of 4015 patients screened, 1001 patients were randomized at 163 US centres (Figure 1). The first patient enrolled in December 2002 and the last patient completed the study in August 2003. One patient in each group did not take any study medication, and these patients were not included in the efficacy or safety analyses. Overall, 94% of those randomized completed the study (Figure 1). The most common reasons for withdrawal were lost to follow-up in the esomeprazole group and lost to follow-up and adverse events in the lansoprazole group. Baseline demographic and clinical characteristics were similar between groups (Table 1). Compliance with study medication was similar between groups [92% (458 of 498) and 90% (452 of 501) by esomeprazole and lansoprazole recipients respectively]. During the first 4 weeks of treatment, mean (standard deviation [s.d.]) antacid tablets used were 0.52 (0.72) per day in the esomeprazole group and 0.57 (0.77) per day in the lansoprazole group.
Table 1. Baseline demographic and clinical characteristics
Esomeprazole 40 mg once daily (n = 498)
Lansoprazole 30 mg once daily (n = 501)
Mean age in years (s.d.)
Sex, n (%)
Race, n (%)
GERD history, n (%)
H. pylori status by serology, n (%)
Baseline LA grade, n (%)
Investigator-recorded GERD symptoms at baseline
Heartburn, n (%)
Acid regurgitation, n (%)
Dysphagia, n (%)
Epigastric pain, n (%)
During 8 weeks of treatment, the time-to-healing curves were significantly different (P = 0.007) favouring esomeprazole, which indicated that esomeprazole healed EE faster and in more patients than lansoprazole. The estimated healing rates were 82.4% with esomeprazole 40 mg once daily and 77.5% with lansoprazole 30 mg once daily (Table 2). The secondary analysis showed that the observed healing rates for all patients at 4 weeks were also significantly higher in the esomeprazole group than in the lansoprazole group (55.8% vs. 47.5% respectively; P = 0.005), but the difference was not significant at 8 weeks (77.5% vs. 73.3% respectively; P = 0.099) (Table 2). Both in the group of patients with grade C EE and the group with grade D EE, at weeks 4 and 8, observed healing rates were higher in the esomeprazole group compared with the lansoprazole group (Table 2).
Table 2. Estimated life-table and observed EE percentage healing rates after 4 or 8 weeks of treatment with once-daily esomeprazole 40 mg or lansoprazole 30 mg
Esomeprazole 40 mg once daily (n = 498)
Lansoprazole 30 mg once daily (n = 501)
* Compared using a Cochran–Mantel–Haenszel test.
† Comparison of the time-to-healing curves using a log-rank test.
Estimated % (95% CI)
58.6 (54.1, 63.0)
49.4 (44.9, 53.8)
Observed % (95% CI)
C and D
55.8 (51.5, 60.2)
47.5 (43.1, 51.9)
Estimated % (95% CI)
82.4 (78.9, 85.9)
77.5 (73.7, 81.2)
Observed % (95% CI)
C and D
77.5 (73.8, 81.2)
73.3 (69.4, 77.1)
Control of symptoms
Investigator assessments showed that a significantly greater proportion of patients (P = 0.005) had resolution of heartburn symptoms with esomeprazole than with lansoprazole (Table 3). Resolution of the other symptoms of GERD was not significantly different between treatments (Table 3). Among patients with resolution of a given symptom, 60–65% of those treated with esomeprazole were also healed of EE, and the corresponding range for those treated with lansoprazole was 49–59%.
Table 3. Percentage of patients with resolution of symptoms at week 4 as assessed by the investigator
Percentage of patients
Esomeprazole 40 mg once daily (n = 478)
Lansoprazole 30 mg once daily (n = 483)
* From the Cochran–Mantel–Haenszel test.
As shown in Figure 2, the difference in sustained resolution of heartburn started to emerge after 1 week of treatment (54.2% for esomeprazole vs. 51.7% for lansoprazole) and remained apparent throughout the 4-week diary period, although the difference was not significant. The mean percentages of heartburn-free days and nights were numerically higher with esomeprazole but were not significantly different between treatments (Table 4).
Table 4. Mean (s.d.) percentage of heartburn-free days and heartburn-free nights based on patients’ daily diaries
Esomeprazole 40 mg once daily
Lansoprazole 30 mg once daily
* From analysis of variance.
† Includes only patients for whom results could be calculated from the diary data.
Mean (s.d.) %
Mean (s.d.) %
Overall, adverse events were similar between groups and mostly mild or moderate in severity. Adverse events were reported by 33.1% of patients in the esomeprazole group and 36.9% of those in the lansoprazole group; no treatment-related serious adverse events were reported. The most common adverse events (occurring in >2% of patients) were Barrett's oesophagus, gastritis, diarrhoea, and headache, all of which were reported by <5% of patients in each group. The most common adverse event leading to study withdrawal was abdominal pain (two in each group, one of which in each group was considered related to treatment). There were no clinically significant differences in the results of clinical laboratory tests or vital signs either within or between treatments.
This study in patients with moderate to severe EE (LA grades C or D) confirmed that esomeprazole 40 mg once daily provided significantly faster healing and a higher healing rate than lansoprazole 30 mg once daily when taken for up to 8 weeks of treatment. Comparison of the time-to-EE healing curves suggests that healing occurred significantly earlier and in more patients with esomeprazole than with lansoprazole. This finding was supported by the analysis of the observed healing rates for which esomeprazole 40 mg had higher healing rates than lansoprazole 30 mg at both weeks 4 and 8. When timing of the healing was not taken into account, the difference in observed healing rates between treatment groups at week 4 (8.3%) was greater and was significant; however, the difference at week 8 (4.3%) was not significant. Investigator assessments indicated that esomeprazole resolved heartburn, the most common GERD symptom, in a significantly higher percentage of patients than did lansoprazole. The symptoms of acid regurgitation, dysphagia, and epigastric pain were resolved similarly with both treatments.
This study had a similar design and included patients who were similar to the subgroup of patients with moderate to severe EE in the study of Castell et al.9 In that study,9 which included more than 5200 patients with EE of all grades of severity, a significantly higher healing rate was observed for up to 8 weeks of once-daily treatment with esomeprazole 40 mg than with lansoprazole 30 mg (92.6% vs. 88.8%; P = 0.0001). The greatest differences between treatments in EE healing rates were for patients with moderate or severe EE at baseline (differences of 17% for LA grade D, 11% for grade C). In addition, a significantly higher percentage of patients in the esomeprazole group than in the lansoprazole group had complete resolution of heartburn by week 4.9 The present study prospectively confirms the significantly faster EE healing and greater number of patients healed of EE with esomeprazole 40 mg than with lansoprazole 30 mg in a patient population exclusively with moderate to severe EE. However, the difference between treatment groups in healing rates in this study was not as great as the difference in the Castell study.9 A possible explanation is the different populations that were studied.
In other randomized clinical trials, esomeprazole 40 mg once daily also demonstrated significantly higher healing rates and more rapid symptom resolution than omeprazole 20 mg once daily, irrespective of patients’ age, gender, race, or H. pylori status (although, by design, most patients enrolled in these trials were H. pylori-negative).1, 2 Moreover, esomeprazole was the first PPI to show efficacy better than that of the standard recommended dose of omeprazole for healing reflux oesophagitis. By contrast, data from randomized controlled trials have shown that lansoprazole 30 mg once daily and omeprazole 20 mg once daily have similar efficacy for the healing of EE.12–14
In the present study, the healing rates with both esomeprazole and lansoprazole in patients with moderate EE (grade C) were higher than in those with severe EE (grade D). This observation is consistent with those of previous studies of patients with EE using these and other PPIs.9, 13–21
Studies that compare EE healing rates between different PPIs can provide useful information for the practising clinician to help guide therapeutic decisions. The difference in estimated healing rates between treatments at week 8 observed in this study (5%) and the difference in heartburn resolution rates at week 4 (8%) are potentially clinically relevant differences because GERD and EE are common disorders. But factors other than efficacy, such as cost of treatment, often have to be also considered when making treatment decisions for a disease. As no formal cost analysis was performed as part of this study, we can base our findings only on estimated differences in healing rates that was the primary outcome of the study. Further studies or analyses looking at cost-effectiveness of the various PPIs for managing EE may aid management decisions in these patients.
Patients with moderate or severe EE are often not readily identified in clinical practice, because the presence and severity of symptoms do not correspond with disease severity as assessed during endoscopy, and most patients who have symptoms typical of uncomplicated GERD do not undergo an endoscopy.5, 22, 23 In addition, empirical treatment with a PPI in clinical practice is common. Therefore, an ideal treatment should provide a high degree of efficacy across all grades of disease severity. The findings of this study and of the previous trial9 show that esomeprazole heals more consistently than lansoprazole across all grades of disease. This demonstrated increased speed in healing EE when using esomeprazole also suggests that this agent may be a better choice when using a PPI as an empirical therapeutic trial in patients with GERD.
Data from this study show that esomeprazole 40 mg once daily heals EE faster and in more patients with moderate to severe EE (LA grade C or D) than does lansoprazole 30 mg once daily. Furthermore, esomeprazole resolved heartburn, the most commonly reported symptom of GERD, in a significantly higher proportion of patients than lansoprazole at week 4. Both esomeprazole 40 mg once daily and lansoprazole 30 mg once daily are well tolerated. Therefore, these results suggest that esomeprazole is a treatment of choice for patients with moderate to severe EE if healing and heartburn relief are the primary goals of therapy.
This study was supported by AstraZeneca LP, Wilmington, DE. We gratefully acknowledge the study participants and study coordinators at the investigational sites, Colleen Jensen for study management, Donna Curtis and Mary Wiggin for editorial assistance (funded by AstraZeneca LP), and the following study investigators: D. Aarons, Lodi, CA; H. Allende, San Antonio, TX; I. Alam, Austin, TX; M.S. Avila, Hialeah, FL; D.L. Avner, Salt Lake City, UT; F. Avni, Lake Worth, FL; L. Bank, Binghamton, NY; C. Barish, Raleigh, NC; I. Bassan, Miami Beach, FL; J.G. Beckwith, Fort Worth, TX; S. Behar, Hialeah, FL; M. Bennett, San Diego, CA; C. Berggreen, Baton Rouge, LA; P. Berggreen, Phoenix, AZ; M.F. Beshay, Mission Hills, CA; T. Bianchi, Tallassee, AL; B.L. Bleau, Tacoma, WA; E. Bonapace, Great Neck, NY; B. Borkar, Jeffersonville, IN; D. Brandon, San Diego, CA; W.C. Bray, Winston-Salem, NC; J. Breiter, Manchester, CT; A. Caos, Ocoee, FL; S. Castillo, Chicago, IL; T. Chami, Zephyrhills, FL; R. Chasen, Laurel, MD; P. Chen, Corpus Christi, TX; D. Chua, Oakbrook Terrace, IL; G. Ciambotti, Hillsborough, NJ; N. Cirillo, Wilkesboro, NC; L. Cohen, New York, NY; F. Cortese, Butte, MT; M. Davis, Rochester, NY; S. Desautels, Salt Lake City, UT; A. Dhand, Ormond Beach, FL; J. Doyle, Spokane, WA; D.W. Dozer, Milwaukee, WI; M. Draelos, High Point, NC; S. Duckor, Orange, CA; M. Eisner, Zephyrhills, FL; P. Eweje, Jacksonville, NC; D. Fenner, Johnson City, TN; V. Fishbein, Hillsboro NJ; F. Fowler, Charlotte, NC; S. Frank, Chattanooga, TN; S. Gaddam, Garden Grove, CA; E. Gallup, Overland Park, KS; H. Giller, Clive, IA; D. Goddard, Upland, CA; A.C. Goetsch, Huntsville, AL; J. Goff, Arvada, CO; J. Goldstein, Willingboro, NJ; M.R. Grossman, Oklahoma City, OK; A.D. Haidar, Picayune, MS; G. Halow, El Paso, TX; M. Hamad, Cary, NC; R. Hansen, Golden, CO; V. Hee, Vancouver, WA; E. Hirsh, Atlanta, GA; R. Holmes, Winston-Salem, NC; W. Ignatowicz, Brooklyn, NY; A.K. Jain, Slidell, LA; V.S. Jayanty, Houston, TX; B.R. Johnson, San Diego, CA; J. Jolley, San Rafael, CA; J. Jones Jr, Ruston, LA; S.R. Jones, Little Rock, AR; S. Jones, Salt Lake City, UT; R. Kaplan, Greensboro, NC; K. Karimi, Indianapolis, IN; S. Kero, Spring Hill, FL; P. Kiyasu, Portland, OR; T. Klein, Wichita, KS; E. Klorfein, West Palm Beach, FL; V. Kodali, Fayetteville, NC; V. Kolli, El Paso, TX; J. Kopp, Anderson, SC; G. Koval, Portland, OR; R. Krause, Chattanooga, TN; S. Krumholz, West Palm Beach, FL; F. Kucer, Sellersville, PA; F. Lanza, Houston, TX; J. Leavitt, Miami, FL; M. LeVine, Marietta, GA; R. Lindenberg, Torrington, CT; D. Lipkis, San Diego, CA; J.E. Lowe, South Ogden, UT; H. Maimon, Dayton, OH; S. Malhorta, Alexandria, VA; R. Marks, Alabaster, AL; J. Medoff, Greensboro, NC; T. Mendolia, Elkin, NC; A. Mihas, Richmond, VA; P. Milman, Lake Success, NY; P. Monroe, Richmond, VA; D. Morin, Bristol, TN; S. Moussa, Tucson, AZ; R. Movva, Moline, IL; G. Mula, Covington, LA; R. Murphy, Little Rock, AR; P. Nichols, Lake Charles, LA; J. Novick, Towson, MD; E. Ojo, Corsicana, TX; J. Orchard, Knoxville, TN; N. Patel, Kettering, OH; R. Patel, Lancaster, CA; L. Peters, High Point, NC; W. Powell, Newark, DE; V. Pratha, San Diego, CA; N.M. Price, Nashville, TN; S. Prohaska, Kansas City, MO; M.V. Provenza, Shreveport, LA; M. Raikhel, Torrance, CA; A. Reymunde, Ponce, PR; M.D. Reynolds, Dallas, TX; D. Ricci, Port Orange, FL; D. Riff, Anaheim, CA; M. Ringold, Christiansburg, VA; R.E. Ringrose, Guthrie, OK; M.M. Rodriguez, Bradenton, FL; M. Rosenberg, Los Angeles, CA; A. Roth, Rancho Cucamonga, CA; S.D. Rubin, Merrick, NY; L. Saco, Lakeland, FL; S. Safavi, Irving, TX; L.A. Salas, Baltimore, MD; J. Salcedo, Washington, DC; J. Santoro, Egg Harbor, NJ; S.L. Scheinert, St Petersburg, FL; M. Schmalz, Milwaukee, WI; L. Schmidt, Tucson, AZ; C. Schmitt, Chattanooga, TN; J. Schneier, Edmonds, WA; D. Schumacher, Columbus, OH; H. Schwartz, Miami, FL; M. Schwartz, Jupiter, FL; A. Shaffi, Colorado Springs, CO; A. Shah, Prince Frederick, MD; U. Shah, Hollywood, MD; B. Shivakumar, Davenport, IA; D. Sikes, Zephyrhills, FL; D. Silvers, Metairie, LA; T. Simmons, Los Angeles, CA; G.A. Spiegelman, Knoxville, TN; E. Spiotta Jr, Memphis, TN; D. Stanton, Orange, CA; M. Stern, Decatur, GA; C. Strout, Mt Pleasant, SC; M. Swaim, Jackson, TN; R. Tamayo, Longwood, FL; R. Tobin, Seattle, WA; N. Trent, Hanover, PA; J. Turse, Melbourne, FL; J. Van Valkenburg, Medford, OR; R. Wade, Bountiful, UT; J. Wang, Jefferson City, MO; M. Weiss, Clearwater, FL; J. Winder, Sylvania, OH; B. Winston, Houston, TX; R. Wholman, Bellevue, WA; S.J. Woods, Bridgeport, CT; L. Wruble, Memphis, TN.
AstraZeneca was the sponsor of this multicentre clinical trial.