HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168361
`
`
`1 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION - OUTSIDE COUNSEL ONLY
`
`APO-FER-0168362
`
`
`2 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION - OUTSIDE COUNSEL ONLY
`
`APO-FER-0168363
`
`
`3 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION - OUTSIDE COUNSEL ONLY
`
`APO-FER-0168364
`
`
`4 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION - OUTSIDE COUNSEL ONLY
`
`APO-FER-0168365
`
`
`5 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION - OUTSIDE COUNSEL ONLY
`
`APO-FER-0168366
`
`
`6 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168367
`
`
`7 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168368
`
`
`8 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168369
`
`
`9 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168370
`
`
`10 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168371
`
`
`11 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168372
`
`
`12 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168373
`
`
`13 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
`REDACTED
`
` APOTEX EX. 2039
`
`

`

`Abstract: 1161
`December 8, 1997
`Oral Session: Clinical Sickle Cell Disease and Thalassemia
`(Room 15A, San Diego Convention Center)
`
`Final results of the randomized trial of deferiprone and deferoxamine.
`
`NF Olivier', GM Brittenham
`
`Toronto, Canada and Cleveland, OH
`
`Treatment with DFO reduces morbidity and improves survival in iron-loaded patients. Results of
`non-randomized trials of L1 suggest that long-term therapy may not adequately control body iron in most
`patients. A prospective randomized trial comparing the effectiveness of L1 and DFO was conducted in Toronto
`and Montreal and prematurely terminated in Toronto by Apotex Inc. in 1996. In Toronto, hepatic iron
`concentration determined by biopsy or magnetic susceptometry (SQUID) was monitored up to May 1997,
`when L1 was discontinued in all patients because of safety concerns. The primary endpoint of effectiveness,
`defined prospectively in the trial protocol, was the change in hepatic iron concentration from that determined
`by SQUID ..12 months prior to randomization, and that determined by SQUID or biopsy after 24 months of
`treatment on study. None of the 7 Montreal patients had completed 24 months on study by May 1997; results
`reported are those for patients randomized in Toronto only. Between study initiation in 11/93 and completion of
`enrollment in 9/95, 64 patients were randomized in Toronto; 13 withdrew before completion of 24 months [6
`from DFO due to: interruption of DFO due to poor growth or hearing loss (3); request for IV DFO (1); parental
`abandonment (1); relocation to the Middle East (1)]; and 7 from L1 due to: L1-induced agranulocytosis (2);
`pre-therapy neutropenia (2); therapy associated with neutropenia (1); request for IV DFO (1); cardiac
`transplantation (1)]. By May 1997 13 of the remaining 51 patients had still not completed 24 months on study,
`while 38 were eligible for evaluation of the primary endpoint. Final hepatic iron concentration was obtained in
`37/38 patients; because of obesity, final SQUID could not be obtained in 1 patient who refused biopsy. At
`baseline, mean hepatic iron did not differ significantly in the 18 DFO-treated patients (6.9 ± 0.9 milligrams iron
`per gram tissue, dry weight [mg/g]) and the 19 L1-treated patients (8.9 ± 1.2 mg/g; p 5_ 0.1). After 33 ± 1.0
`(range 24-43) months, hepatic iron of the DFO-treated patients had not changed (final mean hepatic iron: 7.9
`± 1.3 mg/g) even though, because of titration of dose against anticipated drug toxicity, the mean dose of DFO
`in this study was low (36.7 ± 2.8 mg/kg/night). By contrast, hepatic iron in Ll-treated patients increased
`significantly (final mean hepatic iron: 13.7 ± 1.2 mg/g; p 0.01). Compliance with L1, measured with
`computerized bottles (94.9 ± 1.1%), was significantly better than with DFO, measured using ambulatory
`pumps (71.6 ± 3.7%; P <0.005); only 4 of 18 DFO-treated patients administered 80% of prescribed drug.
`After 24 months, tissue iron exceeded 15 mg/g in 7 L1-treated and in 2 DFO-treated patients, placing these
`patients at heightened risks of cardiac disease and early death (NEJM 1994; 331: 567). Optimal control of
`body iron was observed in all DFO-treated patients who administered >70% of prescribed drug. These data,
`from the only randomized trial of these agents, demonstrate that mean body iron burden increases after two
`years of L1 despite excellent compliance; tissue iron reaches concentrations associated with iron-induced
`complications in 95% patients, even those who begin therapy with relatively low initial body iron burdens. By
`contrast, less regular compliance with low doses of DFO appears to maintain mean body iron burden within
`optimal range.
`
`Iron overload and hemochromatosis
`
`CONIFER,. Information System. CONIFER is a trademark of Excerpta Medics Medical Communications By, Amsterdam.
`
`1
`
`Copyright 1997 by the American Society of Hematology
`
`Pharmacia
`&Upjohn
`
`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168374
`
`
`14 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
` APOTEX EX. 2039
`
`

`

`Blood Cells, Molecules, and Diseases (2002) 28(2) Mar/Apr: 196-208
`doi:10.1006/beind.2002.0510. available online at littp://www.idealibraty.com on 1111Ette
`
`Maggio etal.
`
`Deferiprone versus Deferoxamine in Patients with Thalassemia
`Major: A Randomized Clinical Trial
`
`Submitted 01/24/02; revised 02/20/02
`(Communicated by G. Sternatoyattnopoulos, M.D., 03/04/02)
`
`Aurelio Maggio,' Gennaro D'Amic0,2 Alberto Morabit0,3 Marcell° Capra,4 Calogero Ciacci0,5
`Paolo Cianciulli,6 Felicia Di Gregorio,7 Giovanni Garozzo,8 Roberto Malizia,9 Carmelo .Magnano,1°
`Antonin° Mangiagli,11 Giovanni Quarta,12 Michele Rizzo,13 Domenico Giuseppe D'Ascola,14
`Aroldo Rizzo, 15 and Massimo Midiril6
`
`ABSTRACT: Deferiprone has been suggested as an effective oral chelation therapy for thalassemia major. To
`assess its clinical efficacy, we compared deferiprone with deferoxamine in a large multicenter randomized clinical
`trial. One-hundred forty-four consecutive patients with thalassemia major and serum ferritin between 1500 and
`3000 ng/ml were randomly assigned to deferiprone (75 mg/kg/day) (n = 71.) or deferoxamine (50 mg/kg/day) (n =
`73) for I year. The main measure of efficacy was the reduction of serum fenitin. Liver and heart iron contents were
`assessed by magnetic resonance. Liver iron content and fibrosis stage variations were assessed on liver biopsy by
`the ishak score in all patients willing to undergo liver biopsy before and after treatment. The mean serum ferritin
`reduction was 222 ± 783 ng/m1 in the deferiprone and 232 ± 619 ng/ml in the deferoxamine group (P = 0.81).
`No difference in the reduction of liver and heart iron content was found by magnetic resonance between the two
`groups. Thirty-six patients accepted to undergo repeat liver biopsy: 21 in the deferiprone and 15 in the
`deferoxamine group. Their mean reduction of liver iron content was 1022 ± 3511 p.,g/g of dry liver and 350 ±
`524, respectively (P = 0.4). No difference in variation of the Ishak fibrosis stage was observed between the two
`groups. Treatment was discontinued because of reversible side effects in 5 patients in the deferiprone group (3
`hypertransamin/asemia and 2 leukocytopenia) and in none in the cleferoxamine group. These findings suggest that
`deferiprone may be as effective as deferoxamine in the treatment. of thalassemia major with few mild and reversible
`side effects. ,o 2002 Elsevier Science (USA)
`
`Key Words: LI therapy; oral chelation: randomized clinical trial: dictation therapy: LI etlicacy; thalassemia major management.
`
`Correspondence and reprint requests to: A. Maggio, 1)ivisione di Ematologia II, Azienda Ospedaliera V. Cervello, 90146 Palermo. Italy. Fat:
`00.39.091.6880828. E-mail; aureliomaggio@virgitio.it.
`For the Multicenter 1.1 Study Group of the Society for the Study of Thalassemia and Hemoglobinopathies.
`Divisione di Ematologia lie Unita di Ricetra "Piera Cutino," Aziencla Ospedaliera V. Cervello, Palermo. July.
`z Divisione di Medicina, Azienda Ospedaliera V. Cervello, Palermo, Italy.
`3 Istituto di Biometria e Stalistica Medico., Llniversita di Milano, Milan. Italy.
`▪ I)ivisione di Pediatria VII. Ospedale dei Bambini, Palermo. Italy.
`5 Centro Trasfusionale, Ospedale Civile di Scia.cca, Sciacca, Italy.
`6 Ospedale S. Eugenio, Rome, Italy.
`o Centro Microcitemia. Policlinic° di Catania, Catania, Italy.
`Centro Microcitemia. Ospedale di Ragusa, Ragusa, Italy.
`°Centro Microcitemia, Ospedale Villa Sofia, Palermo, Italy.
`Centro Microcitemia, Ospedale Garibaldi, Catania. Italy.
`"Centre .Microciternia, Ospedale di Siracusa, Siracusa, Italy.
`" Divisione di Ematologia, Ospedale Summa. Brindisi. Italy.
`13 Centro Microcitemia, Ospedale Civile di Caltanissetta, Caltanissetta, Italy.
`54 °spectate Civile di Reggio Calabria, Reggio Calabria, Italy.
`'5 Anatomiti Patologica, Ospedale V. Cervello, Palermo, Italy.
`Istintto di Radiologia. Universita di Palermo. Palermo, Italy.
`
`(4)
`
`1 96
`
`1079-9796/02 535.00
`0 2002 Elsevier Science (USA)
`All rights reserved.
`
`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168375
`
`
`15 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
` APOTEX EX. 2039
`
`

`

`Maggio et al.
`
`INTRODUCTION
`
`Prognosis of patients with thalassemia major
`has dramatically improved in the past two decades
`as a consequence of improvement in transfu.sional
`and chelation therapy (1, 2). Deferoxamine B
`mesylate (DF), is widely accepted as the standard
`chelation therapy (2). However it requires over-
`night subcutaneous infusion and is associated
`with serious side-effects (3-11). For these rea-
`sons, several oral iron chelators have been stud-
`ied. Among these, deferiprone (1,2-dimethy1-3-
`hydroxypyrid-4-one, also called Li) appears to be
`promising. In the only randomized clinical, trial so
`far reported, including 20 patients, Ll, proved to
`have the same chelating effect as subcutaneous
`DF (12). However several other uncontrolled and
`small studies reported contrasting results which
`are difficult to interpret because of different pa-
`tients selection and different length of follow-up
`(.13-26). We therefore carried out a randomized
`clinical trial comparing LI with subcutaneous DR
`
`METHODS
`
`Patients
`
`All. the patients with thalassemia major con-
`secutively observed at the participating centers
`between September 1994 and October 1997 were
`considered eligible for the trial if they had a serum
`ferritin concentration equal to or lower than 3000
`ng/ml. Before the trial all patients were treated by
`deferoxamine B mesylate therapy at dosage of 50
`mg/kg given subcutaneously during a 12-h period
`usually overnight for 5 days a week.
`We decided to exclude patients with greater
`serum ferritin concentrations to minimize the in-
`clusion of patients with a serious risk of multiple
`organ damage from iron overload. The diagnosis
`of thalassemia major was based on accepted clin-
`ical and molecular criteria (27, 28).
`The exclusion criteria were (a) known intol-
`erance to one of the trial treatments; (b) presence
`of rheumatoid factor; (c) serum antinuclear-au-
`toantibody (ANA); (d) platelet count <100,000/
`mm3 or leukocyte <3000/mm3; (e) severe liver
`damage indicated by ascites; (0 clinical. evidence
`
`197
`
`Blood Cells, Molecules, and Diseases (2002) 28(2) MartApr: 196-208
`doi:10.1006/bcind.2002.0510, available online :a lutp/www.idealibrary.com on IDEtte
`
`of heart failure; (g) sepsis; and (h) a-interferon
`treatment (29). Eligibility and exclusion criteria
`were checked at each participating center in the
`outpatient or day-hospital section where the pa-
`tients were also seen throughout the whole fol-
`low-up period.
`
`Interventions
`
`The trial treatments were given according to
`following schedule: deferiprone, 75 mg/kg di-
`vided in three daily doses administered as 500 mg
`pills every 8 h; deferoxamine B mesylate, 50
`mg/kg given subcutaneously during a 12-h period,
`usually overnight for 5 days a week. Deferiprone
`was obtained by Inselspital (Berne, Switzerland),
`Lipomed (Basel, Switzerland), CIPLA Ltd. (In-
`dia) and Apotex (Toronto, Canada). The purity of
`the drug was assessed in random samples
`throughout the study at the School of Life, Basic
`Medical and Health Sciences—King's College
`London—by Professor R. C. Hider Laboratory,
`and always exceeded 98%.
`The planned duration of treatment was 1 year.
`It was established that at the end of the trial each
`patient had to continue chelation therapy by the
`conventional therapy (i.e., DF) until the study
`analysis was completed. We decided to assess the
`treatment efficacy over a 1-year period, consider-
`ing it unlikely that a clinically significant iron
`overload would develop in this relatively short
`time in patients treated with the experimental
`treatment with this baseline serum ferritin levels,
`even if Li was less effective than DF.
`Compliance with the trial treatment was as-
`sessed by counting the pills in each returned bag
`of deferiprone and by assessing the total dose of
`deferoxamine B mesylate consumed each week.
`Compliance was also checked by interviewing the
`patient relatives.
`Standard transfusional therapy was aimed at
`maintaining the hemoglobin blood concentration
`9.5 g/di.
`
`Objectives
`
`The study objective was to compare the two
`treatments in the reduction of iron overload or to
`prevent its increase.
`
`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168376
`
`
`16 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
` APOTEX EX. 2039
`
`

`

`Blood Cells, Molecules, and Diseases (2002) 28(2) Max/Apr: 196-208
`doi:10.1006/bemd.2002.0510, available online at http://wwwidealibrary.coin on IDE title
`
`Maggio a aL
`
`Outcomes
`
`The main measure of the treatment efficacy
`was the difference between the serum fenitin con-
`centration before and after 1. year of treatment.
`Secondary efficacy measures were (a) variation of
`liver iron content (LIC) measured as Ag/gram of
`dry weight in patients willing to undergo liver
`biopsy prior and after the treatment period; (b)
`variation of liver and heart iron content estimated
`by NMR performed by a 0.5-T superconducting
`unit (Vectra, General Eletric Medical Systems,
`Paris, France) using 0.24-cm2 operator-defined
`regions of interest (ROIs) and expressed as aver-
`age intensity signal ratio (ISR) (30-33). (NMR
`was performed at the Institute of Radiology, Uni-
`versity of Palermo for all patients); (c) heart func-
`tion as assessed by the following parameters re-
`corded on heart u.ltrasonography: left ventricular
`ejection fraction (LVEF), left ventricular shorten-
`ing fraction (LVSF) and the ratio of the right
`ventricle telediastolic to the telesystolic area
`(mm3) (RV.DSR); (d) variation in 24-h urinary
`iron excretion (UIE) measured during treatment.
`A liver biopsy before and after the treatment
`was performed in all the patients who accepted it,
`according to a standard technique, to assess liver
`iron content and fibrosis. Biopsies were blindly
`examined under code by two independent observ-
`ers experienced in liver histology, unaware of the
`type of treatment and of timing of biopsies. Liver
`inflammation and fibrosis were rated according to
`the ishak scoring system (34). Interobserver
`agreement beyond chance for the fibrosis score
`was assessed by the weighted kappa statistic (35).
`Liver iron content was assessed by atomic
`spectrophotometry on liver biopsy and expressed
`as amount of iron in µg/g dry liver weight. All
`patients were seen once or twice a month in the
`outpatient or day-hospital section of each partic-
`ipating center, according to the transfusional re-
`quirement. Clinical and biochemical assessment
`was repeated monthly, according to a prefixed
`data form.
`
`Adverse Events
`
`Any potential adverse event was recorded and
`the relationship with the trial treatment was care-
`
`fully investigated. Variation in. liver fibrosis was
`investigated in all patients accepting to undergo
`liver biopsy before and after the trial treatment.
`If an increase of ferritin levels more than 1000
`ng/ml, confirmed by two determinations apart
`with respect to the previous values was detected
`during the study period, the treatment was
`stopped and the alternative therapy was started.
`
`Sample Size
`
`The sample size estimate was based on the
`expected mean reduction in serum ferritin concen-
`tration at the end of 1 year of treatment. Based on
`previous experience at the coordinating center we
`knew that in patients with initial serum ferritin
`below 3000 ng/ml and treated with subcutaneous
`DF, the mean reduction in serum ferritin after one
`year of therapy was 250 ng/ml with a standard
`deviation of 65 ng/ml. We assumed that a differ-
`ence higher than 30 ng/ml with respect to this
`expected ferritin reduction with DF, would be
`clinically significant. Therefore we calculated that
`to detect a 30 ng/ml difference (i.e., from 250 to
`220), 70 patients should be included per group
`(two sided test; a = 0.05; p = 0.80).
`
`Randomization
`
`The randomization was based on a computer
`generated random list in permuted blocks of 10.
`The randomization sequence was generated at the
`Biometrics Institute of the University of Milano.
`To ensure allocation. concealment, treatment was
`assigned by telephone contact of each participat-
`ing center with a physician (FP) of the coordinat-
`ing center who kept the randomization sequence,
`but was not otherwise involved in the study.
`Treatment assignment was done when the inclu-
`sion and exclusion criteria per each consecutively
`observed patient were verified and treatment was
`started within the following 24 h.
`
`Assessments of Outcome
`
`Because of the modality of administration of
`deferoxamine B mesylate, a double blind design
`was considered unethical. However all the out-
`comes assessments (determination of serum fer-
`
`198
`
`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168377
`
`
`17 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
` APOTEX EX. 2039
`
`

`

`Maggio et at.
`
`Blood Cells, Molecules, and Diseases (2002) 28(2) MartApr: 196-208
`doi:10.1006/bcind.2002.0510, available online a lutp/www.idealibrary.com on
`
`Excluded. 102
`;dot meting inctaion cdtssia. 92
`R,,4 t, pseacifst*: 4
`Cstlost ?sums'
`
`Rossiomixed: 144
`
`Altos sted LI. 71
`Recowsd Amsted trestont
`
`fa-TAO DR73
`Rscsived 611=4404 is antossa 13
`
`ritin concentration, urinary iron excretion, liver
`iron content and fibrosis on liver biopsy, liver and
`heart iron content estimated by MR1, heart func-
`tion on ultrasound) were done under code by
`physicians blinded to the trial treatment. Also the
`statistical analysis was performed under code at
`the Biometrics Institute of the University of Mi-
`lano, by a biostatistician (A.M.) blinded to the
`trial treatment.
`
`Statistical Methods
`
`Means are reported with standard deviation
`(SD); proportions and differences between pro-
`portions are reported with 95% confidence inter-
`vals (CI). The statistical analysis was based on the
`intention to treat principle. Continue scale values
`were compared between the two study groups by
`paired t test or two-sample t test with equal vari-
`ances, as appropriate, by using a logarithmic
`transformation whenever this improved the ap-
`proximation to normal distribution. Differences in
`proportions observed on contingency tables were
`
`assessed by x2 analysis. A multiple linear regres-
`
`sion analysis by a step-wise backward procedure
`was planned to identify potential confounding
`factors affecting the mean serum ferritin reduction
`at the end of the treatment period. The following
`set of variables to be included in the multivariable
`analysis was defined a priori: sex, age, splenec-
`tomy, total number of blood units transfused in
`the last 12 months before randomization, initial
`serum ferritin concentration, 24-h urinary iron
`before randomization, cirrhosis,
`excretion
`HBsAg, anti-HCV, diabetes, left ventricular ejec-
`tion fraction, endocrine dysfunctions, number of
`transfusions during the study period, and trial
`treatment. All statistical analyses were performed
`by STATA 6 (1999 STATA Corp.).
`
`Ethics
`
`The study protocol conformed to the ethical
`guidelines of Declaration of Helsinki (36) and
`was approved by the local ethics committee for
`human investigations. The patients gave their
`written informed consent to participate in the
`study.
`
`199
`
`1
`
`g' Lost to Oollow-ug nuns
`Disoutimed intemotim
`issopecorlly. 4 (aorssout T AST 3, Infection 1)
`nteitnitively, 5 (petsissmi t AST 3.4403D Z
`
`Last to 5SPX.W-Rn stone
`Disocoosoosol inferrenlico.
`'it .7$3$41ilY 7 illitaint 1. AST I
`Sojestkot sue pos6)
`cistinslivily: Jort
`
`1
`fi
`1
`
` 1.h.l'ilr'21' 73 1
`
`:i
`
`
`
`FIG. I. Trial profile.
`
`RESULTS
`
`Participant Flow and Recruitment
`
`From September 1994 to October 1997, 246
`patients with thalassemia major were consecu-
`tively observed at the 15 participating centers.
`Among these patients 92 were not eligible be-
`cause of serum ferritin concentration above 3000
`ng/ml. Among the 154 eligible patients 3 were
`excluded because of ongoing a-interferon treat-
`ment, 3 because of rheumatoid factor positivity
`and 4 because of unwillingness to participate in
`the study. The remaining 144 patients were in-
`cluded: 71 were randomly assigned to LI and 73
`to DF. None of the patients was lost to follow-up
`(Fig. 1).
`Eleven patients in the Li group and 7 in the
`DF group with initial serum ferritin concentration
`over 3000 ng/ml were erroneously randomized.
`Individual data of these patients are shown in
`Table 5, and separate results are also reported for
`them.
`
`Baseline Data
`
`Clinically relevant patient characteristics at
`enrollment and corresponding values at the end of
`the one-year study period are shown in Tables
`1-3.
`
`Numbers Analyzed
`
`All 71 patients randomized to Li and 73 to DF
`were included in the analysis according to the
`"intention to treat" principle.
`
`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168378
`
`
`18 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
` APOTEX EX. 2039
`
`

`

`Blood Cells, Molecules, and Diseases (2002) 28(2) Mar/Apr: 196-208
`doi;10.1006/bond2002.0510, available online at littp://wwwidealibrary.com on lIEL
`
`Maggio el al.
`
`Characteristics
`
`Age, years"
`Sex (M/F)"
`Total blood transfusions in the year before
`randomization. ml (mean -± SD)"
`Total blood transfusions during the study,
`ml (mean ± SD)
`Mean serum ferritin in the year before
`randomization, ng/ml (mean ± SD)
`Liver iron concentration, ju,g/gr dry liver
`(mean S SD)
`TibsAg positive"
`Anti-HCV positive*
`Cirrhosis"
`Splenectomy°
`Diabetes"
`Hypogonadism"
`Hypothyroidism'
`Hypoparathyroidism"
`
`TABLE 1
`
`Patitl)ts' Characteristics at Baseline
`
`LI group
`(it = 71)
`
`20 5. 5.3
`37/34
`8302 ± 541
`
`LI group with
`liver biopsy
`(n = 21)
`
`19 -± 3.1
`11/10
`8,000 ± 430
`
`DF group
`(n = 73)
`
`21 54.2
`34/39
`8,965 ± 278
`
`DF group with
`liver biopsy
`(n = 15)
`
`20 ± 4
`7/8
`8,700 ± 650
`
`10,142 1071
`
`8,985 ± 1694
`
`9,181 227
`
`8,380 ± 1043
`
`2,159 ± 668
`
`2,3005 590
`
`2,074 .5 608
`
`2,200 ± 650
`
`3,363 ± 5,490b
`
`3,516 ± 2,974
`
`59
`2
`27
`1
`26
`15
`
`16
`
`8
`
`10
`7
`
`65
`7
`27
`2
`29
`11
`4
`
`II
`0
`7
`
`3
`
`Note. Data are numbers of patients unless otherwise stated.
`"Variables included in the multiple linear regression analysis of confounding factors for the treatment effect.
`"Liver iron concentration was measured in 20 patients of the LI group because of insufficient material in one patient.
`
`Treatment
`
`Fifty-five patients per each trial group took the
`prescribed dose of the trial treatment during the
`whole study period; four patients in the LI
`group and seven in the DF group took a reduced
`does because of low compliance. Twenty-four
`patients in the LI group were not willing to
`switch to the conventional DF treatment at the
`end of the study period. After the recent report
`of a study suggesting that LI may increase the
`risk of liver fibrosis (23), the patients who were
`still willing to continue LI treatment were
`asked to undergo a second liver biopsy before
`continuing treatment.
`Long-term variation of liver iron content and
`fibrosis of all these patients is reported (Tables 3
`and 4; Figs. 2 and 3).
`
`Adverse Events
`
`In the LI group, 24 patients developed side
`effects requiring temporaty dose reduction in. 3
`(nausea) and temporary treatment withdrawal in 4
`
`(transient hypertransaminasemia 3, infection 1).
`Five patients were definitively withdrawn from
`treatment because of recurrence of hypertransami-
`nasernia (>2 times the pretreatment values) (n =
`3, anti-HCV positive 1) or leukocytopenia (n = 2)
`even at reduced doses of the study drug. Mild
`hypertransa.minasemia, spontaneously recovering
`developed in 10 other patients and mild joint pain
`in two. Overall, 14 of the 16 patients who devel-
`oped hypeitransaminasemia, were anti-HCV pos-
`itive. Adverse events occurred in 11 DF treated
`patients: temporaq dose reduction was needed in
`6 patients because of pain and eiythema at the
`injection site and in one because of transient by-
`pertransaminasemia. Two patients developed in-
`fections (Yersinia enterocolitica) and two ototox-
`ici.ty, requiring temporary treatment withdrawal.
`Al.! of these patients continued on DF after tem-
`porary dosage reduction or temporary treatment
`withdrawal.
`One. patient per group was withdrawn from
`the trial treatment because of an increase of serum
`fenitin of more than 1000 ng/ml during the first 6
`
`200
`
`HIGHLY CONFIDENTIAL INFORMATION — OUTSIDE COUNSEL ONLY
`
`APO-FER-0168379
`
`
`19 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
` APOTEX EX. 2039
`
`

`

`Maggio et
`
`Blood Celts, Molecules, and Diseases (2002) 28(2) Mar/Apr: 196-208
`doi:10.1006/bctrid.2002.0510. available online at http://www.idealibrary.com on Mike
`
`Relevant Blood and Biochemisny Parameters at Baseline and at the End of Treatment in the 144 Patients Included in the Study
`
`Li group (a = 71)
`
`DF group (n = 73)
`
`TABLE 2
`
`Parameter
`
`Hemoglobin (g/dl)
`Leucocytes (X I 09/L)
`Platelets (x109/1.)
`Glucose (mgJd1)
`Urea (mg/d1)
`Creatinine (mg/d1)
`Uric acid (mg/d1)
`AST (U/L)
`Anti-14CV positive
`Anti-14CV negative
`ALT (U/I..)
`Anti-HCV positive
`Anti-HCV negative
`-KIT (U/1.)
`Anti-11CV positive
`Anti-HCV negative
`Bilirubin (mg/d1)
`Prothrombin activity (%)
`Albumin (g/d1)
`-y-Globulins (g/d1)
`Iron (jag/dl)
`Transferri El (mg/dl)
`
`Baseline
`
`10.3 ± 1.2
`9.4 ± 4.8
`369 ± 170
`98 ± 16
`31 ± 10
`0.7 ± 0.1
`4.1 ± 1.2
`37 ± 24
`37 ± 25
`30 ± 17
`58 ± 61
`61 ± 60
`29 ± 27
`21. ± 14
`23 ± 17
`12 ± 7
`1.4 ± 0.8
`84 ± 14
`4.5 ± 0.5
`2.3 ± 1.0
`214 ± 49
`21.5 ± 70
`
`After treatment
`
`10.0 ± 0.9*
`11.0 ± 7.9*
`396 ± 184
`95 ± 17
`30± 11
`0.7 ± 0.2
`4.2± 1.3
`55 ± 112
`59 ± 122
`30 ± 18
`80 ± 125
`86 ± 135
`41 ± 38
`35 ± 31."
`38 ± 33
`18 ± 9
`1.2 ± 0.6
`91 ± 115*
`4.5 ± 0.9
`2.2 ± 0.9
`216 ± 64
`223 ± 76
`
`Baseline
`
`10.2 :!: 1.4
`8.9 ± 5,8
`344± 135
`97 ± 36
`31 ± 10
`0.7 ± 0.2
`4.1 ± 13
`33 ± 23
`35 ± 24
`22 ± 12
`50 ± 47
`53 ± 48
`26 ± 31.
`21 -± 16
`22 ± 16
`12 ± 4
`1.4 ± 0.6
`81 ± 13
`4.5 ± 05
`2.4 ± 1.1
`207 ± 49
`215 ± 87
`
`Mr treatment
`
`10.0 :2:: 1.0
`9.1 ±7.4
`358 ± 147
`97 ± 18
`32 ± 9
`0.6 ± 0.2
`4.3 ± 1.1
`34 ± 24
`36 ± 24
`,Y) -I- 5
`48 ± 46
`51 ±47
`24 ± 12
`25 ± 24
`25 ± 25
`'12 ± 5
`1.5 ± 0.7
`85 ± 12
`4.5 ± 0.9
`2.3 ± 1.1
`201 ± 48
`213 ± 91
`
`Note. Data are means ± standard deviations. No significant differences were found at the end of treatment compared with baseline
`(paired t test) unless otherwise specified.
`* P < 0.05 compared with baseline (paired t test).
`P < 0.01 compared with baseline (paired test).
`
`months of trial treatment. Four patients in the Li
`group and seven in the DF group took a reduced
`dose because of low compliance.
`
`Outcomes
`
`Mean serum ferritin concentration before and
`after one year of treatment are reported in Table 3
`and Figs. 4-7. The mean reduction in serum fer-
`ritin concentration was 222 ± 783 ng/nil in the Li
`group and 232 ± 619 ng/ml in the DF group (P =
`0.81) (Table 3). Corresponding results in patients
`with baseline serum fenitin equal or lower than
`3000 ng,/m1 (n = 126) and above this value (n =
`18) are reported in Tables 4 and 5, respectively.
`
`Secondary Measures of Treatment Efficacy
`
`NMR assessment of liver and heart iron con-
`tent are reported in Table 3. A statistically signif-
`
`icant increase in ISR was found after both treat-
`ments for all the NMR measurements, suggesting
`that a significant reduction in the iron content in
`the heart and in the liver was associated with the
`two trial treatments (Tables 3 and 4). This in-
`crease was not statistically significant for the liver
`in the Li group (Tables 3 and 4) although the
`change in liver ISR after treatment was not sig-
`nificantly different between the two study groups
`(Tables 3 and 4).
`Assessment of heart function by ultrasound
`did not show appreciable variation with either
`treatment after the study period (Table 3). The
`values of other clinically relevant parameters at
`the end of treatment are reported in Table 2. It
`is important to note that a slight but not statis-
`tically significant increase of transaminases was
`found after Li treatment greater in anti-HCV
`positive patients; 1/GT significantly increased in
`
`201
`
`HIGHLY CONFIDENTIAL INFORMATION - OUTSIDE COUNSEL ONLY
`
`APO-FER-0168380
`
`
`20 of 71
`
`Taro Pharmaceuticals, Ltd.
`Exhibit 1037
`
`PROTECTIVE ORDER MATERIAL
`
` APOTEX EX. 2039
`
`

`

`Blood Cells, Molecules, and Diseases (2002) 28(2) Mar/Apr: 196-208
`doi:10.1006/bcmd2002.0510, available online at littp://www.idealibrary.com on
`
`Maggio etal.
`
`TABLE 3
`
`Summary of Treatment Efficacy Assessment
`
`Measures of treatment
`efficacy
`
`Scrum ferritin (ng/m1)
`Liver iron concentration
`(pg/g/ciry weight)b
`Anti-HCV positive
`Anti-HCV negative"
`Urinary iron excretion
`(mg/24 It)
`Liver NNW
`Heart septum NMRd
`Left ventricular NIVIRg
`Right ventricular NMRd
`Left ventricular Er a)"
`Left ventricular SF my
`Right ventricular area ratiof
`
`1..1 group (tt = 7! I
`
`DF group (a = 73)
`
`Baseline
`
`End of treatment
`
`Difference"
`
`Baseline
`
`End of treatment
`
`Difference"
`
`2283 ± 754g
`3363 ± 5490
`
`2061 ± 853*
`2341 ± 2197
`
`222 ± 783
`1022 ± 3511
`
`2019 ± 678
`3516 ± 2974
`
`1787 7i: 893*
`3166 ± 2519
`
`232 ± 619
`350 ± 524
`
`3651. ± 5928
`1731. ± 708
`11.4 ± 8.5g
`
`2506 ± 2321
`1353 ± 626
`15.8 ± 10.9*
`
`1145 ± 3812
`378 ± 192
`-4.4 ± 13.2
`
`3483 ± 3049
`3718 ± 3560
`15.7 ± 12.8
`
`3184 ± 2604
`3046 ± 2734
`19.9 ±. 13.6**
`
`299 ± 487
`672 ± 825
`-4.2± 12.5
`
`0.83 ± 0.32
`1.06 ± 0.20
`1.02 ± 0.26
`0.99 ± 0.24
`63 1: 6g
`41 ± 11
`1.9 ± 0.28
`
`0.89± 0.26
`1.18 ± 0.30*
`1.23 ± 0.46*
`1.22 ± 0.50**
`63 ± 6
`40 ± 8
`2.04 ± 0.32
`
`-0.06 ± 0.38
`-0.12 ± 0.32
`-0.21 ± 0.46
`-0.23 ± 0.50
`0 ± 8
`1 ± 10
`-0.14 ± 0.42
`
`0.85 ± 036
`0.98 ± 0.26
`0.99 ± 0.27
`0.97 ± 0.32
`62± 7
`40 ± 12
`1.9 ± 0.28
`
`0.98 ±.• 0.35"
`1.12± 0.29**
`1.12 ±
`1.16 ±
`61 ± 7
`38 ± 8
`2.0 ± 0.1.8
`
`-0.13 ± 0.28
`-0.14 ± 0.30
`-0.13 ± 0.31
`-0.19 ± 0.48
`1 ± 6
`2 ± 9
`0.1 ± 0.28
`
`Note. Data are means ± standard deviations.
`Values at randomization minus values at the end of treatment. Differences were not statistically different between the two study groups
`(two-sample test with equal variances).
`6 Liver iron concentration was measured in 20 and 15 patients in LI and Dr treatment groups.
`'All patients anti-14(N negative had baseline ferritin levels lower or equal to 3000 rigirnl.
`d NMR, nuclear magnetic resonance. Values are expressed a