`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`OMEPRAZOLE VS. RANITIDINE FOR ULCERS ASSOCIATED WITH NONSTEROIDAL ANTIINFL AMMATORY DRUGS
`
`A COMPARISON OF OMEPRAZOLE WITH RANITIDINE FOR ULCERS ASSOCIATED
`WITH NONSTEROIDAL ANTIINFLAMMATORY DRUGS
`
`.D.,
`, P
` R
`, P
` J
`, P
` T
`, M.D., Z
` D. Y
`N
`.D., L
`.D., I
`H
`ÁCZ
`H
`UHÁSZ
`H
`ULASSAY
`SOLT
`EOMANS
`EVILLE
`ÁSZLÓ
`STVÁN
` M. H
`, M.D., C
` J.
` R
`, M.M
`.(I
`.), A
` J. S
`, M.B.,
`J
`WANNELL
`NTHONY
`ENSBURG
`VAN
`HRISTOFFEL
`OWARD
`OHN
`ED
`NT
` C
` J. H
`, D.M.,
`
` A
` S
` T
`: R
`
` O
`
`AND
`HRISTOPHER
`AWKEY
`FOR
`THE
`CID
`UPPRESSION
`RIAL
`ANITIDINE
`VERSUS
`MEPRAZOLE
` NSAID-A
` U
` T
` (ASTRONAUT) S
` G
`*
`FOR
`SSOCIATED
`LCER
`REATMENT
`TUDY
`ROUP
`
`A
`BSTRACT
`Background
`Suppressing acid secretion is thought
`to reduce the risk of ulcers associated with regular
`use of nonsteroidal antiinflammatory drugs (NSAIDs),
`but the best means of accomplishing this is un-
`certain.
`Methods
`We studied 541 patients who required
`continuous treatment with NSAIDs and who had ul-
`cers or more than 10 erosions in either the stomach
`or duodenum. Patients were randomly assigned to
`double-blind treatment with omeprazole, 20 mg or
`40 mg orally per day, or ranitidine, 150 mg orally
`twice a day, for four or eight weeks, depending on
`when treatment was successful (defined as the res-
`olution of ulcer and the presence of fewer than five
`erosions in the stomach and fewer than five erosions
`in the duodenum, and not more than mild dyspep-
`sia). We randomly assigned 432 patients in whom
`treatment was successful to maintenance treatment
`with either 20 mg of omeprazole per day or 150 mg
`of ranitidine twice a day for six months.
`Results
`At eight weeks, treatment was successful
`in 80 percent (140 of 174) of the patients in the group
`given 20 mg of omeprazole per day, 79 percent (148
`of 187) of those given 40 mg of omeprazole per day,
`and 63 percent (110 of 174) of those given ranitidine
`⬍
`(P
`0.001 for the comparison with 20 mg of omepra-
`⫽
`zole and P
`0.001 for the comparison with 40 mg of
`omeprazole). The rates of healing of all types of le-
`sions were higher with omeprazole than with raniti-
`dine. During maintenance therapy, the estimated
`proportion of patients in remission at the end of six
`months was 72 percent in the omeprazole group and
`59 percent in the ranitidine group. The rates of ad-
`verse events were similar between groups during
`both phases. Both medications were well tolerated.
`Conclusions
`In patients who use NSAIDs regular-
`ly, omeprazole healed and prevented ulcers more ef-
`fectively than did ranitidine. (N Engl J Med 1998;338:
`719-26.)
`©1998, Massachusetts Medical Society.
`
`N
`
`ONSTEROIDAL antiinflammatory drugs
`(NSAIDs) are commonly used to treat
`pain and inflammation.
` Their use is fre-
`1
`quently limited by gastrointestinal side
`effects, ranging from dyspeptic symptoms to life-
`threatening bleeding or perforation of gastroduode-
`nal ulcers,
` especially in the elderly.
`2-4
`4
`NSAIDs are thought to cause mucosal injury by
`several mechanisms.
` Some have a direct toxic action
`5
`
`on the gastric mucosa that is exacerbated by acidity,
`since acidity promotes the absorption of NSAIDs in
`their non-ionized form.
` They also impair prosta-
`6
`glandin-dependent mucosal protective mechanisms.
`When the surface cells have been damaged by either
`of these mechanisms, a second wave of injury medi-
`ated by luminal acid often occurs and generates
`deeper ulcerative lesions.
` Prior approaches to pre-
`7
`venting the side effects of NSAIDs have included
`-receptor antagonists to
`treatment with histamine H
`2
`inhibit acid secretion and the administration of pros-
`taglandin analogues to replace the depleted endog-
`enous prostaglandins. However, H
`-receptor antago-
`2
`nists are not very effective for healing or preventing
`NSAID-associated gastric ulcers during continued
`therapy with NSAIDs, although they speed healing
`8
`and help prevent duodenal ulcers.
` The use of
`8-10
`prostaglandin analogues such as misoprostol is lim-
`ited by gastrointestinal side effects such as diarrhea
`and abdominal cramps.
`11
`In short-term studies, the proton-pump inhibitor
`omeprazole prevented aspirin-induced gastric muco-
`sal damage and lesions.
` Omeprazole heals ulcers
`12,13
`effectively and is equally efficacious for gastric or
`duodenal ulcers in the presence or absence of NSAID
` We compared the efficacy of omepra-
`treatment.
`14,15
`zole and ranitidine in patients with gastroduodenal
`ulcers and erosions associated with continuous NSAID
`therapy.
`
`METHODS
`Study Design and Recruitment
`The study, conducted between August 1992 and April 1995,
`was an international double-blind, randomized evaluation of
`
`From the Department of Medicine, University of Melbourne, Western
`Hospital, Melbourne, Australia (N.D.Y.); the Second Department of Inter-
`nal Medicine, Semmelweis University Medical School, Budapest, Hungary
`(Z.T.); the Second Department of Medicine, Borsod County Teaching
`Hospital, Miskolc, Hungary (L.J.); the First Department of Medicine, Petz
`Aladár Teaching County Hospital, Gyor, Hungary (I.R.); the Division of
`Gastroenterology, Department of Medicine, London Health Sciences,
`London, Ont., Canada (J.M.H.); the Department of Gastroenterology,
`Tygerberg Hospital, Tygerberg, South Africa (C.J.R.); and the Department
`of Rheumatology and Rehabilitation, City Hospital (A.J.S.), and the Divi-
`sion of Gastroenterology, University Hospital (C.J.H.), Nottingham, Unit-
`ed Kingdom. Address reprint requests to Dr. Yeomans at the Department
`of Medicine, University of Melbourne, Western Hospital, Footscray, Victo-
`ria 3011, Australia.
`*Other members of the ASTRONAUT Study Group are listed in the Ap-
`pendix.
`
`Volume 338 Number 11
`
`ⴢ
`
`719
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 1
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 1
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`The New England Journal of Medicine
`
`omeprazole and ranitidine for healing and preventing gastro-
`duodenal lesions in patients receiving long-term treatment with
`NSAIDs. The study was conducted in 73 centers in 15 countries,
`in accordance with the principles of good clinical practice
` and
`16
`the Declaration of Helsinki (Tokyo amendments).
`The study had two phases: a healing phase and a maintenance
`phase.
`
`Healing Phase
`Patients of either sex, 18 to 85 years of age, who had any con-
`dition requiring continuous therapy with NSAIDs above specified
`therapeutic doses (no maximal dose), and not more than 10 mg
`of prednisolone or its equivalent per day, were recruited. The
`minimal (and mean) daily oral doses of the commonly used
`NSAIDs were as follows: 50 mg (113 mg) of diclofenac, 50 mg
`(111 mg) of indomethacin, and 500 mg (775 mg) of naproxen.
`The patients underwent endoscopy, and those found to have any
`or all of the following were invited to enter the study: ulcers that
`were 3 mm or more in diameter, more than 10 erosions in the
`stomach, and more than 10 erosions in the duodenum. Erosions
`were assessed with the modified Lanza scale.
` Major exclusion
`9,10
`criteria were neck instability that would compromise endoscopy,
`concurrent erosive or ulcerative esophagitis, pyloric stenosis, ma-
`jor active gastrointestinal bleeding, or disorders that might mod-
`ify the absorption of study drugs.
`All patients provided written informed consent. They were ran-
`domly assigned (in blocks of three per site) to receive oral double-
`blind treatment with 20 mg of omeprazole (Losec, Astra Hässle,
`Mölndal, Sweden) per day, 40 mg of omeprazole per day, or 150
`mg of ranitidine (Zantac, Glaxo Wellcome, Research Triangle Park,
`N.C.) twice daily for four or eight weeks, depending on when
`treatment was successful. Treatment success was defined in ad-
`vance as the disappearance of ulcer and the presence of fewer than
`five erosions in the stomach, fewer than five erosions in the duo-
`denum, and not more than mild dyspeptic symptoms. Patients in
`whom treatment remained unsuccessful after eight weeks received
`40 mg of omeprazole per day for a further four or eight weeks.
`The patients visited the clinic monthly and continued to take
`NSAIDs throughout the study.
`
`Maintenance Phase
`Patients in whom treatment was successful in the healing phase
`were eligible for the maintenance phase if they used therapeutic
`doses of NSAIDs at least five days per week, had no concurrent
`erosive or ulcerative reflux esophagitis, and had no abnormalities
`in the laboratory tests regarded as clinically important by the in-
`vestigator. The patients were randomly assigned (in blocks of two
`per site) to treatment with either 20 mg of omeprazole per day
`or 150 mg of ranitidine twice daily for six months. They were
`evaluated after one, three, and six months and if they had trou-
`blesome dyspeptic symptoms or adverse events. Treatment failure,
`specified in advance, during the maintenance phase was defined
`as a finding of any of the following: gastric or duodenal ulcer,
`more than 10 erosions in the stomach, more than 10 erosions in
`the duodenum, moderate or severe symptoms of dyspepsia, or ad-
`verse events resulting in the discontinuation of treatment. When
`a patient withdrew from the study, both randomized treatment
`and assessment of its effectiveness ceased.
`
`Clinical and Laboratory Assessments
`At all visits, the esophagus, stomach, and duodenum were ex-
`amined endoscopically for erosions or ulcers and biopsy speci-
`mens were obtained from gastric ulcers to rule out malignant
`conditions. Antral-biopsy specimens were obtained at the begin-
`
`ning of the healing phase to evaluate the patients’ Helicobacter py-
`
`lori status with the urease enzyme test (CLO test, Delta West,
`
`
`Bentley, Australia). Patients infected with H. pylori were not treat-
`ed for the infection, because there was no evidence that this ben-
`efited them
` and because
`–stimulated mucosal synthe-
`H. pylori
`17,18
`
`720
`
`ⴢ
`
`March 12, 1998
`
`sis of prostaglandin has been suggested to be beneficial for
`NSAID-associated ulcers.
`19
`Patients were asked standardized questions about their overall
`upper gastrointestinal symptoms and dyspeptic symptoms (epi-
`gastric or abdominal pain, heartburn, nausea, vomiting, upper
`abdominal bloating, and empty feeling in stomach) during the
`preceding seven days. Their overall symptoms and individual
`symptoms were graded as absent, mild (easily tolerated), moder-
`ate (interfering with normal activities), or severe (incapacitating;
`leaving the patient unable to perform normal activities). Safety
`assessments were based on the reported symptoms, adverse
`events, and the results of standard blood screening. We assessed
`compliance by measuring the amount of medication that was re-
`turned.
`
`Statistical Analysis
`We compared the rates of successful treatment in all patients
`who received at least one dose of medication using a Mantel–
`Haenszel life-table test combining data at four and eight weeks.
`We also calculated the rates of healing of gastric ulcer, duodenal
`ulcer, and erosions separately and compared the rates between
`treatment groups using the Mantel–Haenszel test. A logistic-
`regression analysis of prognostic factors that may have influenced
`the success of treatment at four weeks included treatment, site
`and type of lesion, ulcer size,
`
`H. pylori status at entry, blood
`group, type of arthritic disease, smoking status, age, and sex. Up-
`per gastrointestinal symptoms were analyzed by Wilcoxon’s test,
`with stratification according to severity at enrollment.
`The length of time to treatment failure in the maintenance
`phase was compared between treatment groups with remission
`curves estimated by Kaplan–Meier life-table analysis and the log-
`rank test. To evaluate possible prognostic factors, a Cox regression
`model was used with the duration of remission as the dependent
`variable.
`The study was designed to have a power of at least 80 percent
`for two-sided tests at the 1.7 percent level of significance in the
`healing phase (with the Bonferroni adjustment for the three pair-
`wise comparisons) and at the 5 percent level in the maintenance
`phase. The primary efficacy analysis used an intention-to-treat ap-
`proach that included all patients meeting major entry criteria who
`had taken at least one dose of medication. The safety analysis in-
`cluded all patients who received at least one dose of medication
`and for whom there were safety data, regardless of whether they
`met the entry criteria for the trial. For these reasons there were
`small differences in the numbers of patients included in the effi-
`cacy and safety analyses. No interim analyses were conducted. For
`cases in which data censoring arising from the use of the life-table
`approach to data analysis prevented valid statistical comparisons,
`P values are not presented.
`
`RESULTS
`Characteristics of the Patients
`
`Healing Phase
`Of 541 patients randomly assigned to treatment,
`535 were included in the efficacy analysis according
`to treatment actually received (174 received 20 mg
`of omeprazole per day, 187 received 40 mg of
`omeprazole per day, and 174 received 150 mg of ra-
`nitidine twice daily). Three patients who received no
`trial drug, one patient who did not receive an
`NSAID, one whose NSAID dosage was below the
`required threshold, and one in whom inflammation
`was not verifiable were not evaluated. Demographic
`characteristics, endoscopic findings,
`
`H. pylori status,
`and the various underlying arthritic diseases requir-
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 2
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 2
`
`

`

`OMEPRAZOLE VS. RANITIDINE FOR ULCERS ASSOCIATED WITH NONSTEROIDAL ANTIINFL AMMATORY DRUGS
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`T
` 1.
`ABLE
`
`B
`ASE
`
` P
`
`-L
` C
`
`THE
`OF
`INE
`HARACTERISTICS
` E
` A
`.
`NALYSIS
`THE
`FFICACY
`
` I
`
`ATIENTS
`NCLUDED
`
`IN
`
`C
`HARACTERISTIC
`
`H
`EALING
`
` P
`HASE
`
`M
`AINTENANCE
`
`,
`
`,
`
`,
`
`,
`
`,
`
`OMEPRAZOLE
`20 mg/
`DAY
`(
`174)
`⫽
`N
`
`OMEPRAZOLE
`40 mg/
`DAY
`(
`187)
`⫽
`N
`
`RANITIDINE
`150 mg
`
`TWICE
`DAILY
`(
`174)
`⫽
`N
`
`OMEPRAZOLE
`20 mg/
`DAY
`(
`210)
`⫽
`N
`
` P
`HASE
`RANITIDINE
`150 mg
`
`TWICE
`DAILY
`(
`215)
`⫽
`N
`
`49/125
`
`73/114
`
`53/121
`
`64/146
`
`66/149
`
`57
`20–82
`
`56
`31–78
`
`56
`20–80
`
`76
`71
`
`173
`161
`
`177
`53
`16
`
`41
`
`81
`
`3
`
`85
`
`25
`60
`40
`
`104
`85
`21
`
`99
`67
`16
`11
`11
`6
`
`78
`68
`
`172
`161
`
`176
`68
`15
`
`56
`
`74
`
`10
`
`75
`
`17
`81
`42
`
`100
`96
`19
`
`88
`71
`11
`12
`18
`15
`
`56
`25–80
`
`80
`70
`
`173
`161
`
`57
`26–80
`
`79
`68
`
`172
`160
`
`74
`68
`
`172
`160
`
`no. of patients
`
`139
`51
`17
`
`157
`53
`17
`
`144
`60
`19
`
`36
`
`70
`
`7
`
`61
`
`18
`53
`42
`
`83
`77
`14
`
`79
`59
`10
`8
`11
`7
`
`42
`
`67
`
`5
`
`73
`
`20
`55
`39
`
`91
`76
`20
`
`74
`69
`13
`8
`14
`9
`
`42
`
`70
`
`5
`
`57
`
`17
`70
`30
`
`72
`86
`16
`
`81
`54
`7
`8
`14
`10
`
`Sex (M/F)
`Age (yr)
`Mean
`Range
`Mean weight (kg)
`Men
`Women
`Mean height (cm)
`Men
`Women
`
`Previous gastrointestinal
`disease
`Dyspeptic symptoms
`Peptic ulcer
`Bleeding
`Current gastrointestinal
`disease
`Duodenal ulcer
`with
`or without
`
`erosions
`or
`Gastric ulcer with
`
`without erosions
`Duodenal ulcer and
`gastric ulcer with
`or
`
`without erosions
`Erosions only
`Maximal ulcer diameter
`(all sites)
`5 mm
`⬍
`5–9 mm
`10 mm
`⭓
`
`Positive
`Negative
`Unknown
`Disease requiring NSAIDs
`Rheumatoid arthritis
`Osteoarthritis
`Psoriatic arthritis
`Ankylosing spondylitis
`Others
`Combinations
`
`Helicobacter pylori status
`
`ing treatment with NSAIDs were similar at base line
`among the three groups (Table 1). The most com-
`monly used NSAIDs in the group as a whole were
`diclofenac (29 percent of patients), indomethacin
`(23 percent), and naproxen (16 percent). Most pa-
`tients had rheumatoid arthritis or osteoarthritis.
`
`Maintenance Phase
`At the end of the healing phase, 432 patients were
`enrolled in the maintenance phase after randomiza-
`tion, of whom 425 patients were included in the ef-
`ficacy analysis according to treatment received. Sev-
`en patients could not be evaluated: six because they
`declined to continue in the study, and one because
`
`of noncompliance with the NSAID regimen. The
`patients had similar base-line characteristics (Table
`1) and were well balanced with respect to treatment
`received during the healing phase (data not shown).
`
`Clinical Efficacy
`
`Healing Phase
`
`At eight weeks, treatment was successful in 80
`percent (140 of 174) of the patients in the group
`given 20 mg of omeprazole, 79 percent (148 of
`187) of those given 40 mg of omeprazole, and 63
`percent (110 of 174) of those given ranitidine
`(P
`0.001 for the comparison with 20 mg of omep-
`⬍
`
`Volume 338 Number 11
`
`ⴢ
`
`721
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 3
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 3
`
`

`

`The New England Journal of Medicine
`
`razole and P=0.001 for the comparison with 40 mg
`of omeprazole). The success rates were similar when
`all 541 patients were analyzed according to the
`treatmentassigned.
`The rates of healing ofall types of lesions were
`higher in patients treated with 20 mg or 40 mg of
`omeprazole than in those treated with ranitidine
`(Fig. 1). The rates of gastric-ulcer healing during
`the eight-week period were significantly higher with
`20 mg of omeprazole (84 percent, P<0.001) and
`40 mg of omeprazole (87 percent, P<0.001) than
`with ranitidine (64 percent). When the data were
`analyzed to include patients with concurrent duode-
`nal ulcers initially, the respective values were 81 per-
`cent (62 of 77 patients), 86 percent (62 of 72), and
`64 percent (48 of 75). The rates of healing of duo-
`denal ulcers were also higher with 20 mg of omep-
`razole (92 percent, P=0.03) and 40 mg of omepra-
`zole (88 percent, P=0.17) than with ranitidine (81
`percent). When the data were analyzed to include
`patients with concurrent gastric ulcers initially, the
`respective values were 93 percent (40 of 43), 87 per-
`cent (41 of 47), and 79 percent (37 of 47). The
`rates of healing of erosions were also higher with 20
`mg of omeprazole (89 percent, P=0.008) and 40
`mg of omeprazole (86 percent, P=0.19) than with
`ranitidine (77 percent). There were no significant
`differences between the two doses of omeprazole for
`the three types of lesions. The numberofpatients in
`whomtreatmentwasnotsuccessful after eight weeks
`as judged by the persistence of moderate-to-severe
`symptoms was low: nonein the group given 20 mg
`of omeprazole, one in the group given 40 mg of
`omeprazole, and two in the ranitidine group. The
`major reasons for treatmentfailure were therefore
`lack of healing of ulcers or erosions.
`
`Most patients obtained relief from upper gastro-
`intestinal symptoms withinthefirst four weeks ofei-
`ther treatment. However, the percentage of patients
`with improvementofoverall symptomsat four weeks
`was significantly greater in the group given 20 mg
`of omeprazole than in the group given ranitidine
`(P=0.04 by Wilcoxon’s test). Only 6 percent of
`those treated with 20 mg of omeprazole had mod-
`erate-to-severe symptoms at four weeks, as com-
`pared with 52 percentat base line, whereas 12 per-
`cent of those treated with ranitidine had such
`symptoms, as compared with 50 percentat base line.
`By eight weeks, most patients had mild symptoms or
`were free of symptoms, and the differences between
`the groups were notsignificant.
`Maintenance Phase
`
`During the maintenance phase, patients treated
`with 20 mg of omeprazole were in remission (de-
`fined as the absence ofa relapse of lesions, dyspeptic
`symptoms, and adverse events leading to the discon-
`tinuation of treatment) significantly longer than
`those treated with ranitidine (P=0.004 by the log-
`rank test). The estimated proportion ofpatients in
`remission at the end of six months was 72 percent
`in the omeprazole group and 59 percentin the ra-
`nitidine group (Fig. 2).
`All categories of relapse were more common with
`ranitidine than with omeprazole. Gastric ulcers re-
`curred in 5.2 percent of the patients treated with
`omeprazole (11 of 210), as compared with 16.3 per-
`cent (35 of 215) of those treated with ranitidine.
`Therespective proportions with a recurrence of duo-
`denal ulcers were 0.5 percent (1 of 210 patients)
`and 4.2 percent (9 of 215) and with erosions only,
`5.7 percent (12 of 210) and 7.0 percent (15 of 215).
`
`
`
`Gastric Ulcers
`
`Duodenal Ulcers
`
`Erosions
`@ Ranitidine
`
`
`
`Week
`
`a 20 mgof omeprazole
`
`100
`80
`60
`40
`20
`0
`
`= 40 mg of omeprazole
`
`
`
` _
`
`3
`2
`S
`=
`
`Week
`
`No.oF PATIENTS HEALED/TOTAL
`
`20 mgof
`omeprazole
`40 mg of
`omeprazole
`Ranitidine
`
`47/70
`
`59/70
`
`45/67
`
`58/67
`
`35/70
`
`45/70
`
`32/36
`
`33/36
`
`35/42
`
`37/42
`
`31/42
`
`34/42
`
`50/61
`
`54/61
`
`51/73
`
`63/73
`
`37/57
`
`44/57
`
`Figure 1. Cumulative Rates of Healing of Gastric Ulcers, Duodenal Ulcers, and Erosions at Four and
`Eight Weeks during Treatment with 20 mg of Omeprazole Daily, 40 mg of Omeprazole Daily, or 150
`mgof Ranitidine Twice Daily.
`
`722 - March 12, 1998
`
`MYLAN PHARMS.INC. EXHIBIT 1061 PAGE 4
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 4
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 4
`
`

`

`OMEPRAZOLE VS. RANITIDINE FOR ULCERS ASSOCIATED WITH NONSTEROIDAL ANTIINFL AMMATORY DRUGS
`
`with ulcers showed that those with smaller ulcers at
`the initial endoscopy had a significantly higher prob-
`ability of successful treatment in the healing phase
`(P⫽0.003) and of continued remission during the
`maintenance phase (P⫽0.008) than those with larg-
`er ulcers. For this analysis, ulcer size was used as a
`continuous variable.
`
`Safety
`Unfavorable changes in any laboratory variable re-
`garded by the investigator as clinically relevant were
`classified as adverse events. Data on 533 patients in
`the healing phase were available for the evaluation of
`adverse events. The overall rates were high: 30 per-
`cent in the group given 20 mg of omeprazole, 38
`percent in the group given 40 mg of omeprazole,
`and 40 percent in the ranitidine group. In part this
`may reflect the fact that most of the patients had
`chronic arthritis.
`Data on 426 patients in the maintenance phase
`were available for the evaluation of adverse events.
`The mean duration of exposure to the study drug
`was longer in the omeprazole group than the raniti-
`dine group (151 days vs. 131 days). The frequencies
`of adverse events were similar in the omeprazole
`group (64 percent) and the ranitidine group (58
`percent), despite the longer duration of exposure to
`the study drug in the omeprazole group. The six
`most common adverse events in each phase of the
`study are summarized in Table 2, along with the
`proportions of patients who discontinued the study
`because of adverse events or for other reasons. A
`bleeding duodenal ulcer developed in one patient
`after 10 days of maintenance treatment with omep-
`razole. The patient was hospitalized and treated
`with 20 mg of omeprazole twice daily orally and
`2 units of blood, and the bleeding stopped.
`
`DISCUSSION
`When peptic ulcers develop in patients taking
`NSAIDs, the preferred approach is to stop the NSAID
`when possible.20 Ulcers heal slowly during treatment
`with H2-receptor antagonists when NSAIDs are con-
`tinued and more quickly when they are stopped.8,14,21
`However, stopping the NSAID therapy may exacer-
`bate the underlying arthritis. We compared the abil-
`ities of omeprazole and ranitidine to heal and pre-
`vent NSAID-associated gastroduodenal lesions in
`patients receiving continuous NSAID therapy. Be-
`cause we studied patients with lesions at base line,
`who are at higher risk for ulceration than patients
`without lesions,22 our study is relevant to clinical sit-
`uations in which ulcer prophylaxis would be consid-
`ered. Omeprazole has also been shown to be effec-
`tive in placebo-controlled trials in preventing ulcers
`in patients who are taking NSAIDs but who are at
`lower risk for ulceration.23,24
`The main end point for the healing phase was suc-
`
`Volume 338 Number 11
`
`ⴢ 723
`
`Omeprazole
`
`Ranitidine
`
`P ⫽ 0.004
`
`0
`
`42
`
`6 8 10 12 14 16 18 20 22 24
`Week
`
`26
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`Patients in Remission (%)
`
`NO. AT RISK OF RELAPSE
`Omeprazole
`210
`Ranitidine
`215
`
`205
`205
`
`177
`160
`
`145
`114
`
`Figure 2. Estimated Rates of Remission among Patients Treated
`with 20 mg of Omeprazole Daily or 150 mg of Ranitidine Twice
`Daily for up to 26 Weeks.
`P⫽0.004 by the log-rank test for the difference between groups.
`
`Prognostic Factors
`Healing Phase
`The factors associated with a significantly greater
`likelihood of successful treatment in the healing
`phase were treatment with 20 mg of omeprazole
`(P⫽0.04 for the comparison with ranitidine), the
`presence of duodenal ulcers or erosions (both
`P⬍0.001 for the comparison with gastric ulcers), fe-
`male sex (P⫽0.04 for the comparison with male
`sex), and H. pylori–positive status (P ⫽ 0.05 for the
`comparison with negative status). At eight weeks the
`respective rates of successful treatment among H. py-
`lori–positive patients and H. pylori–negative patients
`were 83 percent (69 of 83 patients) and 75 percent
`(58 of 77) in the group given 20 mg of omeprazole,
`82 percent (75 of 91) and 71 percent (54 of 76) in
`the group given 40 mg of omeprazole, and 72 per-
`cent (52 of 72) and 55 percent (47 of 86) in the ra-
`nitidine group.
`
`Maintenance Phase
`In the maintenance phase, treatment with 20 mg
`of omeprazole and a positive test for H. pylori were
`both associated with a significantly greater likeli-
`hood of remaining in remission. The estimated pro-
`portions of H. pylori–positive patients and H. pylori –
`negative patients who remained in remission at six
`months were 79 percent and 60 percent, respective-
`ly, in the group given 20 mg of omeprazole and 66
`percent and 53 percent, respectively, in the raniti-
`dine group. An analysis that included only patients
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 5
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 5
`
`

`

`The New England Journal of Medicine
`
`TABLE 2. INCIDENCE OF MODERATE-TO-SEVERE ADVERSE EVENTS
`IN ALL PATIENTS AND REASONS FOR DISCONTINUATION
`OF TREATMENT AND FOLLOW-UP IN PATIENTS INCLUDED
`IN THE EFFICACY ANALYSIS.
`
`VARIABLE
`
`Mean duration of
`exposure to
`drug (days)
`Most common ad-
`verse events
`(% of patients)
`Diarrhea
`Arthritis
`Headache
`Nausea
`Constipation
`Flatulence
`Discontinuation of
`treatment
`(% of patients)
`Adverse event or
`lack of efficacy
`Other reasons†
`
`Mean duration of
`exposure to
`drug (days)
`Most common ad-
`verse events
`(% of patients)
`Arthritis
`Rheumatoid
`arthritis
`Vomiting
`Abdominal pain
`Diarrhea
`Respiratory tract
`infection
`Discontinuation of
`treatment
`(% of patients)
`Adverse event
`Other reasons‡
`
`HEALING PHASE
`RANITIDINE,
`150 mg
`TWICE DAILY
`(N⫽175)
`
`OMEPRAZOLE,
`40 mg/DAY
`(N⫽185)
`
`OPEN-LABEL
`OMEPRAZOLE,
`40 mg/DAY*
`(N⫽81)
`
`OMEPRAZOLE,
`20 mg/DAY
`(N⫽173)
`
`36
`
`37
`
`40
`
`41
`
`1.7
`2.3
`0.6
`0.6
`0.6
`0.6
`10.2
`
`2.8
`
`7.4
`
`1.6
`0.5
`1.1
`1.6
`0.5
`0.5
`10.1
`
`3.2
`
`6.9
`
`4.0
`3.4
`2.3
`1.7
`2.3
`2.3
`14.1
`
`8.5
`
`5.6
`
`2.5
`2.5
`3.7
`2.5
`1.2
`1.2
`3.9
`
`1.1
`
`2.8
`
`MAINTENANCE PHASE
`OMEPRAZOLE,
`RANITIDINE, 150 mg
`20 mg/DAY
`TWICE DAILY
`(N⫽210)
`(N⫽215)
`
`151
`
`131
`
`2.4
`4.3
`
`2.9
`2.9
`3.3
`2.4
`
`14.5
`
`6.1
`8.4
`
`3.2
`1.4
`
`2.3
`1.9
`1.4
`2.3
`
`13.3
`
`3.2
`10.1
`
`*All other drugs were given in a double-blind manner.
`†The main other reason was unwillingness to continue in the study (four
`patients in the group given 20 mg of omeprazole, four in the group given
`40 mg of omeprazole, six in the ranitidine group, and three given open-
`label omeprazole).
`‡The main other reason was unwillingness to continue in the study (sev-
`en patients in the omeprazole group and seven in the ranitidine group).
`
`724 ⴢ March 12, 1998
`
`cessful treatment, defined as complete healing of ul-
`cers, the disappearance of erosions or a marked re-
`duction in the number of erosions, and the presence
`of no more than minimal symptoms of dyspepsia.
`Treatment failure in the maintenance phase was de-
`fined as reappearance of these features or the discon-
`tinuation of treatment due to adverse events. We
`chose these end points before the study began be-
`cause we believed that patients would consider suc-
`cess in all these areas to be the most desirable out-
`come. These goals were more demanding than those
`in most previous studies,8-11,14,15,22-25 especially since
`the primary assessment was based on all patients
`treated. Nevertheless, treatment was successful in 80
`percent of the patients given 20 mg of omeprazole
`daily for eight weeks in the healing phase, and this
`treatment was clearly superior to 150 mg of raniti-
`dine twice daily. The rates of healing of gastric ul-
`cers, duodenal ulcers, and erosions were all better
`with either 20 mg or 40 mg of omeprazole than
`with ranitidine. Use of the higher dose of omepra-
`zole provided no obvious advantage overall (rate of
`successful treatment, 79 percent).
`Our study was not designed to investigate compli-
`cations of ulcers, since very large numbers of pa-
`tients would be required to yield significant results.
`A bleeding duodenal ulcer developed in one patient
`after 10 days of maintenance treatment with omep-
`razole. In the whole program, including the study
`reported by Hawkey et al.26 in the next article in this
`issue as well as an assessment of prophylaxis among
`patients who were initially free of ulcers,24 only 1 of
`654 patients who were taking 20 mg of omeprazole
`and 2 of 331 patients who were taking placebo had
`such complications.
`In previous studies, the use of standard doses of
`H2-receptor antagonists has been disappointing with
`respect to preventing NSAID-associated ulcers. A
`400-mg dose of cimetidine each night for 10 months
`was ineffective.25 Two large placebo-controlled trials
`of ranitidine showed a substantial reduction in the
`incidence of duodenal ulcers during NSAID treat-
`ment, but no effect on gastric ulcers.9,10 This is a
`problem, since most ulcers that develop during
`NSAID treatment are gastric,27-29 as we also found.
`A large, prospective evaluation showed that therapy
`with H2-receptor antagonists does not reduce the
`risk of ulcer complications in NSAID users.30 Find-
`ings such as these led to the conclusion that NSAID-
`associated gastric ulcers are not caused by acid.31
`However, our findings challenge this view. The rate
`of continued remission was significantly higher with
`omeprazole than with ranitidine, presumably be-
`cause of the greater inhibition of acid induced by
`omeprazole.32 In keeping with the view that acid
`suppression is required to prevent NSAID-associated
`gastric damage, Taha et al. recently showed that high-
`dose famotidine (40 mg twice daily) gave better
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 6
`
`MYLAN PHARMS. INC. EXHIBIT 1061 PAGE 6
`
`

`

`OMEPRAZOLE VS. RANITIDINE FOR ULCERS ASSOCIATED WITH NONSTEROIDAL ANTIINFL AMMATORY DRUGS
`
`protection against NSAID-associated gastric ulcer
`than the standard dose (20 mg twice daily) or pla-
`cebo.28 Although our study did not include a place-
`bo group, three other recent trials showed substan-
`tial reduction in the incidence of NSAID-associated
`gastroduodenal ulcer in patients treated with 20 mg
`of omeprazole as compared with placebo for three
`or six months.23,24,26
`The analysis of prognostic factors showed that
`treatment was more likely to be successful in patients
`with smaller ulcers at the initial endoscopy. This was
`no surprise, since ulcer size has often been noted to
`affect the healing rate, with larger ulcers healing more
`slowly,33 but we did not expect those with larger ul-
`cers initially to be more likely to relapse during the
`maintenance phase. This finding should be added to
`the list of risk factors that may be clinically useful
`when one is deciding whether to recommend preven-
`tive therapy in patients who are receiving NSAIDs.
`There is disagreement whether H. pylori infection
`increases the risks associated with NSAIDs.34,35 Al-
`though it would seem logical to remove all poten-
`tially ulcerogenic factors, when the study started we
`were aware of data showing that H. pylori could ab-
`rogate NSAID-associated reductions in gastric mu-
`cosal synthesis of prostaglandins19 and that H. pylori
`status did not influence the risk of NSAID-associat-
`ed bleeding ulcers.17,18 We therefore decided not to
`eradicate H. pylori. We found higher rates of success
`in H. pylori–positive patients than in those who were
`negative for H. pylori during both phases of the
`study regardless of treatment assignment. We stud-
`ied a selected group of patients, and proving that
`H. pylori infection is beneficial to NSAID users
`would require a formal randomized study.
`In conclusion, in patients taking NSAIDs daily,
`we found that treatment with 20 mg of omeprazole
`daily is superior to ranitidine with respect to healing
`and preventing gastroduodenal ulcers and erosions,
`as well as controlling dyspeptic symptoms.
`
`Supported by Astra Hässle, Mölndal, Sweden.
`Dr. Yeomans serves as a consultant for Searle Australia.
`
`We are indebted to Dr. M. Frame for valuable he