The
`„n New England
`Journal of Medicine
`
`Abstracts in the
`advertising
`sections
`
`A 3 0 ,96
`
`Established in 1812 as The NEW ENGLAND JOURNAL OF MEDICINE AND SURGERY
`
`VOLUME 332
`
`FEBRUARY 2, 1995
`
`NUMBER 5
`
`Original Articles
`
`Decline in Semen Quality among Fertile Men
`in Paris during the Past 20 Years (cid:9)
`J. AucER, J.M. KUNSTNIANN, F. CZYGLIK,
`AND P. JOUANNET
`Association between Plasma Homocysteine
`Concentrations and Extracranial
`Carotid-Artery Stenosis (cid:9)
`J. SELHUB AND OTHERS
`Methotrexate for the Treatment of Crohn's
`Disease (cid:9)
`B.G. FEAGAN AND OTHERS
`Infection with a Babesia-like Organism in
`Northern California (cid:9)
`al-1. PERSING AND OTHERS
`
`Images in Clinical Medicine
`
`Acupuncture-Needle Fragments (cid:9)
`E.S. (cid:9)
`AND J.H.M. AUSTIN
`
`Review Articles
`
`Mechanisms of Disease: Bone Marrow, Cyto-
`kines, and Bone Remodeling — Emerg-
`ing Insights into the Pathophysiology
`of Osteoporosis (cid:9)
`MANOLAGAS AND R.1,..J11.1:A
`
`Current Concepts: Treatment of Male
`Infertility (cid:9)
`S.S. HowARDs
`
`Molecular Medicine
`
`Human DNA Polymorphism (cid:9)
`D. HOUSMAN
`
` 281
`
`
`
`286
`
` 292
`
`
`
`
`
`298
`
`304
`
` 305
`
` 312
`
` 318
`
`Clinical Problem-Solving
`Diagnostic Strategy — The Shotgun versus
`the Arrow (cid:9)
`G.E. TI-JIBAULT
`
`Editorials
`Are Semen Quality and Male Fertility
`Changing? (cid:9)
`SHERINS
`
`Can Lowering Homocysteine Levels Reduce
`Cardiovascular Risk? (cid:9)
`STAMITER AND M.R. MALINow
`Methotrexate for Chronic Diseases in Adults
`M.E. MINI:ATT'
`
` 321
`
`
`
`327
`
` 328
`
`330
`
`
`
`332
`
`
`334
` 335
` 335
`
`
`
`336
`
`
`
`336
`337
` 338
` 338
` 339
` 342
`
`Correspondence
`Clinical Problem-Solving: Still Hazy after All These
`Years (cid:9)
`Post-remission Chemotherapy for Acute Myeloid
`Leukemia (cid:9)
`Roth's Spots in Leukemic Retinopathy (cid:9)
`Management of Cancer of the Prostate (cid:9)
`More on Primary Biliary Cirrhosis in Monozygotic
`Twins (cid:9)
`Drug Malabsorption and Resistant Tuberculosis in
`HIV-Infected Patients (cid:9)
`More on Hantavirus in New England and New York
`The Terrier and the Tendinitis (cid:9)
`The Cat and the Catheter (cid:9)
`Book Reviews (cid:9)
`Notices (cid:9)
`
`Corrections
`Management of Cancer of the Prostate (cid:9)
`Fluoxetine (cid:9)
`Books Received (cid:9)
`Information for Authors (cid:9)
`
`336
`
`
`343
` 343
` 344
`
`Owned, Published, and OCopyrighted, 1995, by the Massachusetts Medical Society
`th-
`Tint NEw ENci.ANn .lurRNAI. ur NIEDICINE (ISSN 0028-179:n is i
`lislaid weekly from editorial Aires to 10 Shattuck Street, Huston, 11\.
`021 11-609-I. Subscription price: 51 1)9.00 per year. Sceond-class posi-
`age paid at Boston and at additional wailing offices. POSTNIASTER:
`Send address changes io RD. Bus 803, Waltham, MA 02254-0803,
`
`2
`C (cid:9)
`2 (cid:9)
`C (cid:9)
`2 (cid:9)
`C (cid:9)
`2
`C
`
`VBXNEVOG******** 5-DIN T 63706
`10550158/4A 122895 020295 AY 75
`UN SCFI PHARM
`PHARM LIBR
`25 N CHARTER Si
`4MADISON
`
`WI 53706-1508
`
`1111111111111111111111111111111111111111111111111111
`
`N
`E
`w
`S
`P
`A
`P
`E
`R
`
`8666-6
`
`- • -
`
`Sandoz v. AbbVie
`Sandoz Ex. 1084
`
`? (cid:9)
`
`A (cid:9)
`
`LIBRAEY
`
`Ex. 1084 - Page 1
`
`

`

`maval (cid:9)
`
`Fotw1.1. Iept.....,111, 17 41 1.10 I
`
`292
`
`THE NEW ENGLAND JOURNAL OF MEDICINE (cid:9)
`
`Feb. 2, 1995
`
`METHOTREXATE FOR THE TREATMENT OF CROHN'S DISEASE
`
`BRIAN G. FEAGAN, M.D., JAMES ROCHON, PH.D., RICHARD N. FEDORAK, M.D., E. JAN IRVINE, M.D.,
`GARY WILD, M.D., LLOYD SUTHERLAND, M.D., A. HILLARY STEINHART, (cid:9)
`GORDON R. GREENBERG, M.D.,
`RICHARD GILLIES, M.D., MARYBETI-I HOPKINS, R.N., STEPHEN B. HANAUER, M.D.,
`AND JOHN W.D. MCDONALD, M.D., FOR THE NORTH AMERICAN CROHN'S STUDY GROUP INVESTIGATORS*
`
`Abstract Background. Although corticosteroids are
`highly effective in improving symptoms of Crohn's dis-
`ease, they may have substantial toxicity. In some pa-
`tients, attempts to discontinue corticosteroids are unsuc-
`cessful.
`Methods. We conducted a double-blind, placebo-
`controlled multicenter study of weekly injections of
`methotrexate in patients who had chronically active
`Crohn's disease despite a minimum of three months of
`prednisone therapy. Patients were randomly assigned to
`treatment with intramuscular methotrexate (25 mg once
`weekly) or placebo for 16 weeks. The patients also re-
`ceived prednisone (20 mg once a day), which was ta-
`pered over a period of 10 weeks unless their condition
`worsened. The primary outcome measure was clinical re-
`mission at the end of the 16-week trial. Remission was
`defined by the discontinuation of prednisone and a score
`of -..150 points on the Crohn's Disease Activity Index.
`Results. A total of 141 patients were randomly as-
`signed in a 2:1 ratio to methotrexate (94 patients) or pla-
`cebo (47 patients). After 16 weeks, 37 patients (39.4 per-
`
`CROHN'S disease is an inflammatory disorder that
`
`commonly involves the small bowel and colon. Al-
`though corticosteroids are highly effective in improv-
`ing symptoms,1,2 attempts to discontinue therapy are
`unsuccessful in approximately 20 percent of patients."
`Patients treated with corticosteroids continue to have
`both complications of the disease and chronic toxicity
`from the therapy.`'-' Either mercaptopurine or azathio-
`prine is sometimes prescribed to reduce the require-
`ments for corticosteroids, but the toxicity of these med-
`ications is of concern." Low-dose cyclosporine is not
`effective in this situation." Alternative treatments
`are desirable.
`Methotrexate, an antiinflammatory drug, has been
`used to treat rheumatoid arthritisi1,15 and psoriasis.H307
`After a report of improvement in patients with Crohn's
`disease who were treated with methotrexate,1" we fur-
`ther assessed the efficacy of methotrexate therapy in
`chronically active Crohn's disease.
`
`From the Department of Medicine, Division of Gastroenterology, University of
`Calgary, Calgary, Alta. (L.S.); the Department of Medicine, Section of Gastroen-
`terology, University or Chicago, Chicago (S.B.H.); the Department of Medicine,
`Division of Gastroenterology, University of Alberta, Edmonton (R.N.F.); the De-
`partment of Medicine, Division of Gastroenterology, McMaster University,
`Hamilton, Ont. (E.J.I.); the Departments of Medicine (B.G.F., J.W.D.M., MM.)
`and Epidemiology and Biostatistics (cid:9)
`J.R.), University of Western Ontario,
`London, Ont.; the Department of Medicine, Division of Gastroenterology, McGill
`University, Montreal (G.W.); the Department of Medicine, Division of Gastroen-
`terology, University of Thronto, Toronto (A.H.S., G.R.G.); and the Department of
`Medicine, University of Ottawa, Ottawa, Ont. (R.G.), Address reprint requests to
`Dr. Feagan at 6 OF 12 University Hospital. 339 Windemerc Rd., London. ON
`N6A 5A5. Canada.
`Supported by research grants from the Medical Research Council or Canada,
`the Crohn's and Colitis Foundation of America through donations front the David
`and Minnie Berk Foundation, and the Crohn's and Colitis Foundation of Canada.
`*The persons and institutions participating in the North American Crohn's
`Study Group are listed in the Appendix.
`
`cent) were in clinical remission in the methotrexate group,
`as compared with 9 patients (19.1 percent) in the placebo
`group (P=0.025; relative risk, 1.95; 95 percent confi-
`dence interval, 1.09 to 3.48). The patients in the metho-
`trexate group received less prednisone overall than those
`in the placebo group (P= 0.026). The mean (±SE) score
`on the Crohn's Disease Activity Index after 16 weeks of
`treatment was significantly lower in the methotrexate
`group (162±12) than in the placebo group (204±17,
`P = 0.002). The changes in quality-of-life scores and se-
`rum orosomucoid concentrations were similar. In the
`methotrexate group, 16 patients (17 percent) withdrew
`from treatment because of adverse events (including
`asymptomatic elevation of serum aminotransferase
`7 and nausea in 6), as compared with 1 patient (2 per-
`cent) in the placebo group.
`Conclusions.
`In a group of patients with chronically
`active Crohn's disease, methotrexate was more effec-
`tive than placebo in improving symptoms and reducing
`requirements for prednisone. (N Engl J Med 1995;332:
`292-7.)
`
`METHODS
`
`A randomized, double-blind, placebo-controlled study was con-
`ducted at seven university medical centers between November 1992
`and February 1994. The protocol was approved by the investigational
`review board at each center. All the patients gave written informed
`consent..
`
`Patients
`
`The medical records of potentially eligible patients were reviewed
`by a clinician, a radiologist, and a pathologist to confirm the diagno.
`sis of Crohn's disease. Eligible patients had chronically active disease
`with at least three months of symptoms despite daily doses of at least
`12.5 mg of prednisone with at least one attempt to discontinue treat-
`ment. Patients who had received long-term prednisone therapy at low
`doses (---5.1() nag per clay) were ineligible, as were critically ill patients.
`Patients with the following risk factors for methotrexate toxicity'
`were ineligible: preexisting hepatic disease (biopsy-proved cirrhosis,
`chronic active hepatitis, or serum aspartate aminotransferase, biliru-
`bin, or alkaline phosphatase concentrations at least twice the upper
`limit of normal), renal dysfunction (serum creatinine concentration
`greater than 1.7 mg per deciliter [150 µmot per liter]), clinically im-
`portant lung disease as determined subjectively, systemic infection.
`pregnancy or a desire to become pregnant, history of cancer, high
`alcohol consumption (more than seven drinks per week), hypersensi-
`tivity to methotrexate, erythrocyte macrocytosis, body weight 411
`percent higher than normal, diabetes mellitus, a requirement for non-
`steroidal antiinfiammatory drugs, or the use of immunosuppressivc
`drugs in the past three months. Patients with an estimated survival
`of less than one year and those who were unwilling to comply with
`the protocol were also ineligible for the study.
`
`Base-Line Studies
`
`Three weeks before randomization, potentially eligible patients
`were instructed in the use of a diary card to score the Crohn's Dis-
`ease Activity Index.20.21 This index incorporates eight items: the num-
`ber of liquid or very soft stools, abdominal pain, general well-being.
`extraintestinal manifestations of Crohn's disease, the use of opiates
`to treat diarrhea, abdominal mass, hematocrit, and body weight:
`these yield a composite score ranging from 0 to approximately 600.
`Higher scores indicate more disease activity; patients with scores of
`
`Ex. 1084 - Page 2
`
`(cid:9)
`

`

`Vol. 332 No. 5 (cid:9)
`
`METHOTREXATE FOR THE TREATMENT OF CROHN'S DISEASE (cid:9)
`
`293
`
`150 or less are considered to have inactive disease, whereas those
`with scores above 450 are critically ill. A clinic visit was scheduled
`one week later (two weeks before randomization), at which time a
`physical examination and blood tests were performed and base-line
`demographic information, scores on the Crohn's Disease Activity In-
`dex, and data on prednisone use were obtained. Quality of life was
`measured with the Inflammatory Bowel Disease Questionnaire, a
`previously validated instrument with four parts (on bowel function,
`emotional status, systemic symptoms, and social function); the total
`score on this index ranges from 32 to 224, with higher scores inclicat-
`ilig better quality of life. The scores of patients in remission usually
`range from 170 to 190.22.23 Patients were then treated with 20 mg of
`peednisone once daily. A uniform dose was chosen to control for the
`effects of a primary determinant of disease activity and to permit a
`common starting point from which to measure differences in predni-
`sone use between groups.
`
`Randomization
`The patients were randomly assigned, in a 2:1 ratio, to receive ei-
`ther 25 mg of methotrexate (Rheumatrex, Lederle Laboratories,
`Pearl River, N.Y.) or a placebo weekly for 16 weeks if they had not
`required increases in their prednisone close to 20 mg daily in the pre-
`ceding two weeks. Medication was given by intramuscular injection
`to ensure drug absorption and minimize nausea. The placebo was
`identical in appearance to the active drug. Between each patient's
`visits to the study clinic, the injection was administered by a family
`physician. The investigators were unaware of the treatment assign-
`ments. Patients who were receiving 20 mg or more of prednisone dai-
`ly two weeks before randomization were randomized in a separate
`stratum (the high-prednisone stratum) from those who had their
`dose increased to 20 mg (the low-prednisone stratum). Stratification
`was used because we predicted that patients who had required higher
`prednisone doses in the past to control symptoms would have a worse
`prognosis.
`
`Prednisone Therapy
`For two weeks after randomization, no attempt was made to de-
`crease the prednisone dose. After the first follow-up visit (at week 2),
`the daily dose of prednisone was decreased by 2.5 mg each week.
`Prednisone was discontinued by week 12 of the study if the patient's
`condition remained stable or improved. Patients whose condition
`worsened had their prednisone dosage increased to a maximal daily
`do,e of 40 mg. After a close increase, prednisone tapering was re-
`sumed at a rate of 5 mg a week until a daily dose of 20 mg was
`reached. The tapering regimen described above was then begun
`again.
`
`Other Treatments for Crohn's Disease
`The patients were not permitted to use aminosalicylates, budeso-
`nide, immunosuppressive agents, antibiotics for perianal disease,
`tufie feeding, parenteral nutrition, or topical aminosalicylates or cor-
`ticosteroids. The use of hydrocortisone ointment was allowed for
`perianal disease.
`
`Fcllow-up
`Patients were seen 2 and 4 weeks after randomization and every
`4 weeks thereafter for 16 weeks. At each visit, the patient's scores on
`the Crohn's Disease Activity Index and the Inflammatory Bowel Dis-
`ea::e Questionnaire were calculated, and the serum orosomucoid con-
`centration (a laboratory measure of inflammatory activity) and the
`total prednisone dose were measured. Patients who discontinued
`their medication because of adverse reactions or treatment failure
`wer.e followed in the same way as those who continued to receive in-
`jections.
`A physician who had no contact with patients and did not assess
`outcomes, but who was aware of the group assignments, monitored
`serum aminotransferase concentrations each month and complete
`blood counts every two weeks. These results were not made available
`to the attending physicians and nurses. If leukopenia developed
`(white-cell count, ...c.3.8X109 per liter), the study drug was withheld
`for one week and the daily dose was decreased to 17.5 mg the follow-
`ing week. The study drug was discontinued if persistent leukopenia
`developed. An identical algorithm was followed if the serum amino-
`
`transferase concentrations increased to twice the upper limit of nor-
`mal. Matching close adjustments were made in the placebo group.
`
`Outcome Measures
`The primary outcome measure was the presence of clinical remis-
`sion, as defined by the discontinuation of prednisone therapy and a
`score on the Crohn's Disease Activity Index of 150 points at the
`end of the trial (16 weeks). Secondary outcomes were the daily close
`of prednisone, the mean scores on the Crohn's Disease Activity Index
`and the Inflammatory Bowel Disease Questionnaire, and the mean
`serum orosomucoid concentration."
`
`Statistical Analysis
`
`Statistical comparisons were made with SAS software?'' A two-
`sided P value of 0.05 was the criterion for statistical significance. All
`analyses were performed according to the intention-to-treat princi-
`ple. The medical center and stratum of the prednisone close were
`used as the stratification variables. Base-line characteristics meas-
`ured on a nominal or ordinal scale were compared by Fisher's exact
`test or the chi-square test, and continuous variables were compared
`by analysis of variance.
`In the primary analysis, the proportions of patients in the two
`study groups who successfully discontinued prednisone and remained
`in remission at 16 weeks were compared with use of the Mantel—
`Haenszel chi-square test. Differences between the high-prednisone
`and low-prednisone strata with regard to this outcome were com-
`pared by logistic regression analysis. The daily prednisone close,
`scores on the Crohn's Disease Activity Index and the Inflammatory
`Bowel Disease Questionnaire, and the mean serum orosomucoid con-
`centrations were compared by repeated-measures analysis of vari-
`ance." In these analyses the overall effect of treatment was assessed
`by comparing trends over time; differences between study groups at
`the end of follow-up were assessed by comparing the values predicted
`for the two groups in linear models. The distribution of prednisone
`use was skewed toward higher daily doses; repeated-measures anal-
`ysis performed on ranks was used to analyze these data.
`The number of patients withdrawn from therapy because of ad-
`verse reactions or treatment failure was compared between study
`groups by Fisher's exact test. The number of adverse events was com-
`pared with the use of a Poisson regression mode1.2"
`We anticipated that 20 percent of the patients receiving placebo
`would remain in remission. The randomization of 135 patients al-
`lowed 80 percent power to detect an absolute difference of 25 percent
`in this outcome between study groups.
`
`RESULTS
`
`Between September 1992 and November 1993, 193
`patients were assessed to determine whether they were
`eligible for the study. The most common reasons for
`exclusion from the study were an inability or unwill-
`ingness to give informed consent (10 patients), the
`presence of risk factors for methotrexate toxicity
`(8 patients), and a requirement for a contraindicated
`medication (7 patients). Sixteen patients were excluded
`for other reasons, leaving a total of 152 eligible pa-
`tients, Eleven of these patients were not randomized
`because of a refusal to participate by the patient or the
`patient's physician (eight patients), an increase in the
`prednisone dose above 20 mg before randomization
`(two patients), or the occurrence of a new illness (deep
`venous thrombosis in one patient). The patients who
`were eligible but who were not randomized did not dif-
`fer significantly with respect to age, sex, and duration
`of disease from the patients who entered the study. Of
`the 141 study patients, 94 were randomly assigned to
`receive methotrexate and 47 to receive placebo. Eighty-
`nine patients (59 assigned to the methotrexate group
`and 30 to the placebo group) were included in the high-
`
`Ex. 1084 - Page 3
`
`

`

`294 (cid:9)
`
`THE NEW ENGLAND JOURNAL OF MEDICINE
`
`Feb. 2, 1995
`
`prednisone stratum, and 52 patients (35 in the metho-
`trexate group and 17 in the placebo group) were in-
`cluded in the low-prednisone stratum. The base-line
`characteristics of the groups were similar (Table 1).
`
`Primary Outcome
`No patients were lost to follow-up. The same propor-
`tion of patients in the two groups (28 percent) was
`withdrawn from treatment prematurely (26 of 94 re-
`ceiving methotrexate, as compared with 13 of 47 re-
`ceiving placebo; P = 0.99). The proportion of patients
`withdrawn because of treatment failure was significant-
`ly lower in the methotrexate group (7 of 94 receiving
`methotrexate [7 percent], as compared with 11 of 47
`receiving placebo [23 percent]; P = 0.014). After 16
`weeks (Fig. 1) the proportion of patients who had dis-
`continued prednisone therapy and remained in remis-
`sion was higher in the methotrexate group than in the
`placebo group: 37 of 94 (39 percent) as compared with
`9 of 47 (19 percent; P = 0.025; relative likelihood of
`entering remission, 1.95; 95 percent confidence inter-
`val, 1.09 to 3.48). In the high-prednisone stratum, this
`outcome occurred in 23 of 59 patients receiving meth-
`otrexate (39.0 percent), as compared with 3 of 30
`patients receiving placebo (10.0 percent; P =0.003;
`relative likelihood of entering remission, 3.88; 95 per-
`cent confidence interval, 1.60 to 9.43). In contrast,
`14 of 35 patients receiving methotrexate in the low-
`prednisone stratum (40 percent) had this primary
`outcome, as compared with 6 of 17 receiving placebo
`(35 percent; P = 0.92; relative likelihood of entering re-
`mission, 0.96; 95 percent confidence interval, 0.43 to
`2.17). When the percentage of response in the placebo
`group was subtracted from that in the methotrexate
`group, the difference in therapeutic gain between the
`prednisone strata (20 percent in the high-prednisone
`
`Table 1. Base-Line Characteristics of the Study Patients.*
`
`CHARACTERISTIC
`
`Age — yr
`Male sex
`Disease site
`Colon
`Small bowel
`Both
`Months since diagnosis
`Months of continuous disease activity
`Abdominal mass
`Previous surgery for Crohn's disease
`Cigarette smoker
`CDAI score
`IBDQ score
`Serum orosomucoid concentration — ing/dIt
`White-cell count — X10-Miter
`Hemoglobin — g/liter
`Platelet count — X10-9/liter
`
`Mut IOTREXATE
`(N = 94)
`
`34±- 1
`51 (54)
`
`15(16)
`30 (32)
`49 (52)
`93±8
`28±4
`13 (14)
`44 (47)
`46 (49)
`181±11
`162±3.4
`93±3.0
`11.7±0.3
`135-± 1.6
`367± 11
`
`PLACEBO
`(N = 47)
`
`36±2
`26 (55)
`
`9(19)
`8 (17)
`30 (64)
`98±12
`28±5
`3 (6)
`22 (47)
`22 (47)
`190±14
`159±5.2
`95±5.0
`11.6±0.5
`135±2.6
`363± 14
`
`.Plus-minus values are means ±SE, and all other values are numbers of patients followed
`in parentheses by the percentage of patients in the group. P >0.10 for all paired comparisons
`between groups. CDAI denotes the Crohn's Disease Activity Index, and IBDQ the Inflamma-
`tory Bowel Disease Questionnaire. Higher scores on the CDAI indicate greater disease activity,
`and higher scores on the IBDQ indicate better quality of life.
`tSerum orosomucoid concentrations increase with inflammation. The normal range is 33 to
`88 mg per deciliter.
`
`Placebo (cid:9)
`
`Wit Methotrexate
`
`50
`
`P = 0.025
`
`P = 0.003
`
`30.4
`
`39.0
`
`P = 0.92
`40.0
`
`25
`
`19.1
`
`a
`0
`ra
`
`.c
`a) c:n
`
`U
`
`0
`
`All Patients
`
`High-Stratum
`Prednisone
`
`Low-Stratum
`Prednisone
`
`Figure 1. Percentages of Patients in Remission at Week 16, Ac-
`cording to Study Group and Stratum of Daily Prednisone Dose
`before Entry into the Study.
`The high-prednisone stratum was receiving a daily dose of more
`than 20 mg of prednisone, and the low-prednisone stratum a dai-
`ly dose of 20 mg or less, more than two weeks before random-
`ization. The actual percentages are shown above the bars. P val-
`ues were derived by the Mantel—Haenszel chi-square test, with
`adjustment for study center. For the definition of remission, see
`the Methods section under "Outcome Measures."
`
`stratum minus 5 percent in the low-prednisone stra-
`tum) was significant (P = 0.04).
`Characteristics associated with the primary outcome
`were examined by stepwise logistic regression with th,:
`variables of age, sex, prednisone stratum, site of dis-
`ease, scores on the Crohn's Disease Activity Index ant .l
`the Inflammatory Bowel Disease Questionnaire, serum
`orosomucoid concentration, and smoking status. The
`base-line score on the Crohn's Disease Activity Index
`was inversely associated with the probability of dis-
`continuing prednisone and remaining in remission
`(P = 0.04; relative likelihood of entering remission, 1.30
`for each 50-point decrease in the score on the index).
`The other characteristics were not significantly associ-
`ated with the primary outcome.
`
`Prednisone Use
`The patients in the methotrexate group used less
`prednisone overall than those in the placebo group
`(P= 0.026). The difference in prednisone use was de-
`tectable in the 90th percentile of the distribution (hig11-
`er prednisone dose) by week 4 and in the 50th percen-
`tile by week 12 (Fig. 2). This difference was due to th..!
`increased use of high-dose prednisone therapy in the
`patients assigned to receive placebo whose condition
`worsened in the later weeks of the study. The difference
`was greatest from week 12 through week 16. At the end
`of the study, the 50th, 75th, and 90th percentiles of th.-,
`daily prednisone dose in the methotrexate group were
`0, 12.5, and 20 mg, respectively, as compared with J,
`20, and 30 mg in the placebo group (P= 0.003).
`
`Disease Activity
`The average of the mean (±SE) scores on the
`Crohn's Disease Activity Index (Fig. 3) over the entire
`follow-up period was significantly lower in the metho-
`
`Ex. 1084 - Page 4
`
`(cid:9)
`(cid:9)
`(cid:9)
`(cid:9)
`

`

`Vol. 332 No. 5 (cid:9)
`
`METHOTREXATE FOR THE TREATMENT OF CROHN'S DISEASE (cid:9)
`
`295
`
`50th Percentile
`
`30 -
`
`25 -
`
`20 -
`
`15-
`
`10-
`
`5-
`
`0-
`
`0 2 4 6 8 10 12 14 16
`
`75th Percentile
`
`30 -
`
`25 -
`
`20 -
`
`15-
`
`10-
`
`0 2 4 6 8 10 12 14 16
`
`30 -
`
`25 -
`
`20 -
`
`15 -
`
`10 -
`
`5 -
`
`90th Percentile
`
`o Placebo
`• Methotrexate
`
`-------..-•
`
`0 -
`0 2 4 6 8 10 12 14 16
`
`Weeks since Randomization
`
`Figure 2. The 50th, 75th, and 90th Percentiles of the Daily Pred-
`nisone Dose in the Study Patients, According to Group.
`The daily prednisone dose was 20 mg for all patients from week
`2 before randomization until week 2 after randomization, when it
`was tapered by 2.5 mg each week. Patients whose condition im-
`proved or remained stable discontinued prednisone therapy at
`week 12. Patients whose condition worsened at any time during
`the study were treated with high-dose prednisone (20 to 40 mg
`daily). The use of higher daily doses in these patients skewed
`the distribution of prednisone doses toward higher values, mak-
`ing a comparison based on means inappropriate. The overall
`P value was 0.026 by a repeated-measures log-rank test; the
`P value at week 16 was 0.003 by the log-rank test.
`
`trexate group (170±7) than in the placebo group
`(193±17, P = 0.003). There were significant differences
`from week 6 onward; at the end of the study, there was
`difference of 42 points (162± 12 in the methotrexate
`group vs. 204±17 in the placebo group, P = 0.002).
`At base line, the groups' mean quality-of-life scores
`were similar (methotrexate, 162±17; placebo, 159±5).
`After four weeks of therapy, a significant difference be-
`tween the groups developed (Fig. 3). The average of
`
`the mean quality-of-life scores was higher in the meth-
`otrexate group (methotrexate, 166±2; placebo, 155±3;
`P<0.001). At 16 weeks the mean values were 169±4 in
`the methotrexate group and 151±6 in the placebo
`group (P<0.002). Improvement in quality of life was
`evident in all four parts of the Inflammatory Bowel Dis-
`ease Questionnaire (P<0.01 for all comparisons).
`The mean serum orosomucoid concentrations de-
`creased in the methotrexate group and increased in the
`placebo group (Fig. 3). The average of the mean oroso-
`mucoid concentrations in patients treated with metho-
`trexate was 88±2 mg per deciliter, as compared with
`97±3 mg per deciliter in patients receiving placebo
`(P = 0,007). There were significant differences between
`the groups from 4 weeks onward; at 16 weeks, the val-
`ues were 82±3 in the methotrexate group and 97±6 in
`the placebo group (P = 0.003).
`
`Adverse Effects
`Among 94 patients treated with methotrexate, 16 (17
`percent) withdrew from treatment because of adverse
`events, as compared with 1 of 47 patients receiving
`placebo (2 percent, P = 0.012). The patient in the pla-
`cebo group had an episode of polyneuropathy that
`required hospitalization. The reasons for withdrawal
`in the methotrexate group were as follows: asympto-
`matic elevation of serum aminotransferase concentra-
`tions (seven patients), nausea (six), skin rash (one),
`pneumonia probably due to mycoplasma (one), and op-
`tic neuritis (one). Table 2 shows the frequency of drug-
`related adverse events that were not severe enough to
`warrant discontinuation of the study drug. The patients
`in the methotrexate group had 2.6 such events per pa-
`tient, as compared with 2.9 events per patient in the
`placebo group (P = 0.35).
`
`DISCUSSION
`We found that methotrexate was an effective and
`well-tolerated treatment for patients with chronically
`active Crohn's disease, At the time of randomization,
`the patients had moderately active disease despite re-
`ceiving 20 mg of prednisone each day. After treatment
`with methotrexate, significantly more patients were
`able to discontinue prednisone use than were patients
`receiving placebo. Because long-term prednisone ther-
`apy is associated with a variety of harmful consequenc-
`es, methotrexate represents an alternative treatment
`for patients who do not tolerate prednisone or in whom
`symptoms of Crohn's disease persist despite a moder-
`ately high dose of prednisone.
`Although they received less prednisone, the patients
`who received methotrexate had improvement with re-
`gard to disease activity and were more likely to enter
`clinical remission. After 16 weeks of treatment, the
`mean score on the Crohn's Disease Activity Index
`(162±12) approximated that in patients with inactive
`disease (--.150). Improvement in symptoms as assessed
`by the Crohn's Disease Activity Index and the Inflam-
`matory Bowel Disease Questionnaire was detectable by
`six weeks. This rapid response is in contrast to the rel-
`atively slow onset (three to six months) of therapeutic
`
`Ex. 1084 - Page 5
`
`(cid:9)
`

`

`296 (cid:9)
`
`THE NEW ENGLAND JOURNAL OF MEDICINE (cid:9)
`
`Feb. 2, 1995
`
`Table 2. Adverse Events in the Study Patients Ac-
`cording to Group.'"
`
`ADVERSE EVENT
`
`METHOTREXATE
`(N = 94)
`
`PLACEBO
`(N = 47)
`
`no. of patients (% o group)
`
`Nausea and vomiting
`Symptoms of cold
`Abdominal pain
`Headache
`Joint pain or arthralgia
`Fatigue
`Influenza-like illness
`Diarrhea
`Abdominal bloating or distension
`Skin rash
`Insomnia
`Other
`
`40 (42)
`22 (23)
`17 (18)
`16 (17)
`15 (16)
`15 (16)
`10(11)
`7 (7)
`6 (6)
`6 (6)
`2 (2)
`42 (45)
`
`18 (38)
`9 (19)
`12 (26)
`5 (11)
`6 (13)
`5 (11)
`6 (13)
`4 (8)
`3 (6)
`2 (4)
`2 (4)
`20 (42)
`
`'Adverse events related to treatment with the study drug that were not
`severe enough to warrant discontinuation of the study drug are shown. Pa-
`tients may have had more than one adverse event.
`
`the occurrence of this complication in one of our pa-
`tients was probably clue to chance, further study of pa-
`tients with Crohn's disease treated with methotrexato
`will be needed to exclude a causal relation. The risk
`of liver disease with long-term methotrexate therapy
`in patients with Crohn's disease is not known. To min-
`imize the risk of hepatic toxicity, we discontinued
`treatment if patients had persistently elevated serum
`aminotransferases, but this may have been unneces-
`sary. It might have been possible, for example, to re-
`duce the close of medication and follow the patients.'"
`However, in the absence of data specific to patients
`with Crohn's disease, we believe that the recommen-
`dations for the use of methotrexate in rheumatoid ar-
`thritis should be followed.3t,32 These recommendatiow•
`include not using the drug in patients with risk factors
`for hepatic toxicity (alcohol abuse, obesity, or preex-
`isting liver disease), monitoring serum aminotransfer-
`ase and albumin concentrations at monthly intervals,
`and performing a liver biopsy in patients with per-
`sistent enzyme elevations or hypoalbuminemia. Addi-
`tional risks associated with methotrexate are those of
`hypersensitivity pneumonitis,33 bone marrow depres-
`sion," and teratogenicity.35
`Effective drug therapy to maintain clinical remission
`in patients with Crohn's disease is currently unavail-
`able' Maintenance therapy is a research priority. Bu-
`desonide,'" the new aminosalicylates,"" and methotrex-
`ate should be evaluated in this regard. We are studying
`the efficacy and safety of 15 mg of methotrexate once
`weekly for the prevention of relapse of Crohn's disease
`in patients with quiescent disease.
`In conclusion, in our group of patients, methotrexate
`improved symptoms rapidly and reduced the require.
`meat for prechnsone in patients with chronically active
`Crohn's disease.
`
`We are indebted to the patients who participated in the study, to
`Karen Taylor-Dolmer for assistance in preparing the manuscript, try
`Beckman Scientific for providing orosomucoid-assay kits, to Lederle
`Laboratories for methotrexate, and to the Upjohn Company of Can-
`ada for prednisone.
`
`effect with the antimetabolites azathioprine and mer-
`captopurine.
`There was a significantly greater benefit of treatment
`in the high-prednisone stratum and in patients with
`lower scores on the Crohn's Disease Activity Index at
`base line. It was not, however, our hypothesis before
`the study that methotrexate would have such a differ-
`ential effect, and these analyses of subgroups should be
`interpreted with caution.
`Methotrexate tr