`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`The New England
`Journal
` Medicine
`of
`
`© Copyr ight, 19 98, by the Massachusetts Medical Societ y
`
`VOLUME 339
`
`A
`
`U G U S T
`
` 13, 1998
`
`NUMBER 7
`
`LONG-TERM SAFETY AND EFFECTIVENESS OF IRON-CHELATION THERAPY
`WITH DEFERIPRONE FOR THALASSEMIA MAJOR
`
`.D.,
`, P
`L
` E. M
`, M.D., C
` M. B
`, M.D., G
` F. O
`N
`H
`AREN
`C
`HRISTINE
`RITTENHAM
`ARY
`LIVIERI
`ANCY
` M. T
`, P
`.D., M.D., R
` G. C
`, M.D., P
`.D., R
` A. M
`C
`OBERT
`OSS
`H
`EMPLETON
`AMERON
`H
`C
`LELLAND
`A
` D. B
`, M.D.,
` K
` A. F
`, D.P
`., M.B., C
`.B.
`LASTAIR
`URT
`AND
`ENNETH
`LEMING
`HIL
`H
`
`, P
`H
`
`.D.,
`
`D
`OUGLAS
`
`A
`BSTRACT
`Background
`Deferiprone is an orally active iron-
`chelation agent that is being evaluated as a treat-
`ment for iron overload in thalassemia major. Studies
`in an animal model showed that prolonged treat-
`ment is associated with a decline in the effectiveness
`of deferiprone and exacerbation of hepatic fibrosis.
`Methods
`Hepatic iron stores were determined year-
`ly by chemical analysis of liver-biopsy specimens,
`magnetic susceptometry, or both. Three hepato-
`pathologists who were unaware of the patients’ clin-
`ical status, the time at which the specimens were
`obtained, and the iron content of the specimens ex-
`amined 72 biopsy specimens from 19 patients treat-
`ed with deferiprone for more than one year. For com-
`parison, 48 liver-biopsy specimens obtained from 20
`patients treated with parenteral deferoxamine for
`more than one year were similarly reviewed.
`Results
`Of the 19 patients treated with deferiprone,
`18 had received the drug continuously for a mean
`(±SE) of 4.6±0.3 years. At the final analysis, 7 of the
`18 had hepatic iron concentrations of at least 80
`µmol per gram of liver, wet weight (the value above
`which there is an increased risk of cardiac disease
`and early death in patients with thalassemia major).
`Of 19 patients in whom multiple biopsies were per-
`formed over a period of more than one year, 14 could
`be evaluated for progression of hepatic fibrosis; of
`the 20 deferoxamine-treated patients, 12 could be
`evaluated for progression. Five deferiprone-treated
`patients had progression of fibrosis, as compared
`with none of those given deferoxamine (P=0.04). By
`the life-table method, we estimated that the median
`time to progression of fibrosis was 3.2 years in defer-
`iprone-treated patients. After adjustment for the ini-
`tial hepatic iron concentration, the estimated odds of
`progression of fibrosis increased by a factor of 5.8 (95
`percent confidence interval, 1.1 to 29.6) with each ad-
`ditional year of deferiprone treatment.
`Conclusions
`Deferiprone does not adequately con-
`trol body iron burden in patients with thalassemia
`and may worsen hepatic fibrosis. (N Engl J Med
`1998;339:417-23.)
`©1998, Massachusetts Medical Society.
`
`T
`
`RANSFUSIONS and iron-chelation thera-
`py have dramatically improved the lives of
`patients with thalassemia major.
` Transfu-
`1
`sions can prevent death and promote nor-
`mal development, but the iron in the transfused red
`cells accumulates and eventually damages the liver,
`heart, and other organs. Deferoxamine, the only iron-
`chelating agent approved for clinical use, prolongs
`survival and ameliorates iron-induced organ dam-
`age.
` Unfortunately, to be effective, treatment with
`2,3
`deferoxamine requires prolonged parenteral infusion,
`making compliance difficult. Considerable effort has
`been devoted to finding alternative treatments.
` One
`4
`candidate, deferiprone (1,2-dimethyl-3-hydroxypyri-
`din-4-one, or L1), was initially evaluated with an in-
`direct indicator of therapeutic effectiveness — the
`serum ferritin concentration.
` Subsequently, a di-
`5-11
`rect quantitative assessment of body iron burden
`demonstrated a favorable effect of deferiprone on
`iron balance.
` Recognized adverse effects of defer-
`12
`iprone include embryotoxicity, teratogenicity, arthri-
`tis, severe neutropenia, and agranulocytosis. Recently,
`concern about long-term treatment with deferiprone
`was aroused by studies of an animal model of iron
`overload
` in which long-term treatment with a
`13
`closely related compound, 1,2-diethyl-3-hydroxypyr-
`idin-4-one, was associated with a loss of effectiveness
`and an exacerbation of hepatic fibrosis.
`
`From the Departments of Medicine and Pediatrics (N.F.O.), Laboratory
`Medicine and Pathobiology (D.M.T., R.G.C.), and Chemistry (R.A.M.),
`University of Toronto, Toronto; the Departments of Pediatrics and Medi-
`cine, Columbia University College of Physicians and Surgeons, New York
`(G.M.B.); the Division of Epidemiology, Department of Medicine, Univer-
`sity of California, Irvine (C.E.M.); the University Department of Pathology,
`Royal Victoria Infirmary, Newcastle upon Tyne, United Kingdom
`(A.D.B.); and the Department of Pathology, Nuffield Department of
`Pathology and Bacteriology, John Radcliffe Hospital, University of Oxford,
`Oxford, United Kingdom (K.A.F.). Address reprint requests to Dr. Olivieri
`at the Hemoglobinopathy Program, Hospital for Sick Children, 555 Uni-
`versity Ave., Toronto, ON M5G 1X8, Canada.
`
`Volume 339 Number 7
`
`·
`
`417
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on August 23, 2017. For personal use only. No other uses without permission.
`
` Copyright © 1998 Massachusetts Medical Society. All rights reserved.
`
`Apotex Tech.
`Ex. 2011
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`The New England Journal of Medicine
`
`To determine whether the effects of deferiprone
`are sustained during long-term therapy, we measured
`hepatic iron during continued treatment of patients
`in whom body iron had been measured during short-
`term therapy.
` To assess whether long-term therapy
`12
`was associated with progression of hepatic fibrosis, a
`panel of hepatopathologists evaluated the liver-biopsy
`specimens obtained during this trial.
`
`METHODS
`
`Patients
`Of 21 previously studied patients who received deferiprone for
`a mean (±SE) of 3.1±0.3 years,
` 19 continued to receive defer-
`12
`iprone at a dose of 75 mg per kilogram of body weight per day,
`while undergoing repeated biopsies for hepatic iron measurements.
`Long-term effectiveness could be evaluated in 18 patients, who
`had received the drug continuously for 4.6±0.3 years; 1 patient
`had stopped taking deferiprone shortly after the previous analysis,
`but the results of biopsies up to the discontinuation of therapy are
`included in the histologic analysis. Beginning in year 3 of the seven-
`year study, hepatic iron was measured by magnetic susceptometry
`in vivo.
` Because this technique does not provide histologic infor-
`14
`mation, the follow-up period for effectiveness (range, 2 to 7 years)
`was longer than that for histologic analysis (range, 2 to 6 years).
`Patients received regular transfusions. The objective of transfu-
`sion was to maintain the hemoglobin concentration above 9.5 g per
`deciliter. From November 1989 to November 1993, deferiprone
`was synthesized at the University of Toronto and encapsulated by
`NovaPharm Pharmaceuticals (Toronto). After November 1993,
`Apotex (Weston, Ont., Canada) supplied deferiprone tablets. The
`equivalence of the two formulations was not evaluated.
`Body iron was evaluated in tissue obtained at biopsy, as described
`previously, and by magnetic susceptometry (Biomagnetic Technol-
`ogies, San Diego, Calif.).
` The magnetic measurements have been
`14,15
`validated previously.
` We converted the concentration of iron in
`14
`dried samples to a wet weight, assuming a liver water content of 70
`percent; chemical and magnetic values were used interchangeably.
`12
`In the original trial design, each biopsy specimen, obtained pri-
`marily to monitor therapeutic effectiveness, was histologically re-
`viewed, but serial biopsy specimens were not prospectively com-
`pared. Concern about hepatotoxicity
` prompted a retrospective
`13,16
`review of these results and of those from a comparison group of
`20 deferoxamine-treated patients. The comparison group included
`all patients eight years of age or older for whom the results of two
`or more biopsies performed one year apart during continuous de-
`feroxamine treatment were available.
`
`Efficacy Monitoring
`A hepatic iron concentration of less than 80 µmol per gram of
`liver, wet weight, was considered to indicate effective iron-chela-
`tion therapy, and a concentration of 80 µmol or more per gram
`of liver, wet weight, was considered to indicate ineffective thera-
`py.
` These criteria, derived from a long-term trial in deferoxa-
`12
`mine-treated patients,
` were used to evaluate the short-term
`17
`effectiveness of deferiprone in our previous study
` and were ap-
`12
`plied in an identical manner in the present long-term study. Sim-
`ilarly, a serum ferritin concentration of less than 2500 µg per liter
`was considered to indicate effective iron-chelation therapy, and
`higher values ineffective therapy.
`12,18
`
`Histologic Evaluation
`An independent initial review of the biopsy specimens was car-
`ried out by the two study investigators who are hepatopathologists
`and a consultant. Before this review, the 72 biopsy slides were
`randomly arranged and each slide was assigned a unique number.
`Each pathologist graded the findings according to the system
`summarized in Table 1.
` Each was unaware of the patients’ clin-
`19
`
`418
`
`·
`
`August 13, 19 98
`
`H
`
`
` E
`
` S
` 1.
`T
`VALUATING
`OF
`YSTEM
`ABLE
`-B
` S
`
` A
`FOR
`PECIMENS
`IOPSY
`EPATIC
`RCHITECTURAL
`C
`, F
`,
` C
`.*
`HANGES
`IBROSIS
`AND
`IRRHOSIS
`
`
`
`F
`INDING
`
`S
`CORE
`
`†
`
`No fibrosis
`Fibrous expansion of some portal areas, with or
`without short fibrous septa
`Fibrous expansion of most portal areas, with or
`without short fibrous septa
`Fibrous expansion of most portal areas, with
`occasional portal-to-portal bridging
`Fibrous expansion of portal areas, with marked
`portal-to-portal bridging as well as portal-to-
`central bridging
`Marked portal-to-portal bridging as well as portal-
`to-central bridging, with occasional nodules
`(incomplete cirrhosis)
`Probable or definite cirrhosis
`
`0
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`*The criteria were obtained, with modifications, from
`Ishak et al.
` The following additional features were noted but
`19
`not scored: intra-acinar fibrosis, perivenular (“chicken wire”)
`fibrosis, and phlebosclerosis of terminal hepatic venules.
`†The maximal score is 6.
`
`ical status, the date each sample was obtained, and the hepatic
`iron content of each biopsy specimen. After the completion of
`the initial review, the two study investigators conducted a consen-
`sus review, in which all biopsy specimens were examined jointly,
`after standards regarding sample adequacy and definitions of pro-
`gression and regression of fibrosis had been agreed on. The
`results of this evaluation were subsequently reviewed with the
`consultant, and a final decision was made with regard to the ad-
`equacy and stage of each biopsy specimen.
`Clinically significant progression of fibrosis was considered to
`have occurred if there was a change in the histologic score from
`0 (no fibrosis) to 1 or greater, from 1 or 2 to 3 or greater, or from
`3 or 4 to 5 or 6. Changes in the score from 1 to 2 or from 3 to
`4 were not considered clinically significant. Similarly, a change in
`the score from 5 (incomplete cirrhosis) to 6 (probable or definite
`cirrhosis) was not considered clinically significant; hence, patients
`whose initial biopsy specimen showed cirrhosis could not be eval-
`uated for progression of fibrosis. A biopsy specimen was consid-
`ered adequate if two or more portal tracts were present. This was
`considered the absolute minimum necessary for the assessment of
`fibrosis and cirrhosis, since the ability to identify these processes
`varies considerably depending on the size of the biopsy specimen.
`
`Monitoring of Toxicity and Compliance
`Other types of safety monitoring in this study have been de-
`scribed previously.
` Sexually active patients were asked to use reli-
`12
`able methods of contraception. We assessed compliance by mon-
`itoring the frequency with which pill bottles were opened, using
`bottles with microprocessors in the caps.
`12,20
`
`Statistical Analysis
`Data are presented as means ±SE. Medians and ranges are given
`for continuous variables, and proportions are given for dichoto-
`mous variables. Pretreatment variables were compared between
`treatment groups by the Mann–Whitney test for continuous var-
`iables and by the Fisher–Irwin exact test for dichotomous vari-
`ables.
` The Wilcoxon signed-rank test was used to compare pre-
`21
`treatment and post-treatment values for continuous variables and
`to assess whether there was a change in compliance during the last
`two years of deferiprone therapy. The Kaplan–Meier product-limit
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on August 23, 2017. For personal use only. No other uses without permission.
`
` Copyright © 1998 Massachusetts Medical Society. All rights reserved.
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`IRON-CHEL ATION THERAPY WITH DEFERIPRONE FOR THAL ASSEMIA MAJOR
`
`200
`
`150
`
`100
`
`50
`
`0
`
`0
`
`1
`
`5
`4
`3
`2
`Years of Treatment
`
`6
`
`7
`
`Hepatic Iron (µmol/g of liver, wet weight)
`
`Figure 1.
` Initial and Final Hepatic Iron Concentrations in 18
`Patients with Thalassemia Major Treated with Deferiprone.
`The dashed line indicates the value of 80 µmol of iron per gram
`of liver tissue, wet weight, above which there is an increased
`risk of cardiac disease and early death due to iron loading during
` Squares indicate
`long-term treatment with deferoxamine.
`17
`concentrations determined by liver biopsy, and circles concen-
`trations determined by magnetic susceptometry.
`
`
`
`
` C
` 2.
`T
`THE
`OF
`HARACTERISTICS
`ABLE
`
`
`-T
` P
`
`14 D
`THE
`EFERIPRONE
`REATED
`ATIENTS
`AND
`
`12 D
`-T
` P
`
` W
`HOM
`EFEROXAMINE
`REATED
`ATIENTS
`IN
`
` H
` F
` C
` B
` E
`ROGRESSION
`OF
`EPATIC
`IBROSIS
`OULD
`E
`VALUATED
`
`.
`
`P
`
`C
`HARACTERISTIC
`
`Progression of hepatic fibrosis —
`no. (%)
`Age at initial biopsy — yr
`Median
`Range
`Male sex — no. (%)
`Antibody to hepatitis C — no. (%)
`Initial hepatic iron concentration —
`µmol/g of liver, wet weight
`Median
`Range
`Duration of treatment — yr*
`Median
`Range
`Amount of blood transfused —
`ml packed cells/kg/yr*
`Median
`Range
`
`D
`EFERIPRONE
`G
`ROUP
`(N=14)
`
`D
`EFEROXAMINE
`G
`ROUP
`(N=12)
`
`5 (36)
`
`0
`
`P
`V
`ALUE
`
`0.04
`
`18.2
`10.5–23.7
`3 (21)
`6 (43)
`
`13.9
`8.7–31.5
`6 (50)
`5 (42)
`
`0.1
`
`0.2
`1.0
`
`80.9
`24.2–180.1
`
`35.2
`10.8–226.4
`
`0.01
`
`2.3
`1.3–4.0
`
`3.2
`1.3–4.3
`
`77
`56–114
`
`82
`57–109
`
`0.4
`
`0.5
`
`*The period under consideration is the interval between the initial biopsy
`and the final biopsy. Data were available for 13 patients in the deferiprone
`group and 10 in the deferoxamine group.
`
`Volume 339 Number 7
`
`·
`
`419
`
`method was used to estimate the probability that each patient
`would not have progression of fibrosis for a specified period. The
`log-rank test was used to compare differences in the length of time
`to the progression of fibrosis in the treatment groups.
` Because
`22
`the only patients with progression of fibrosis were in the defer-
`iprone group, it was not possible to estimate the risk or odds of
`progression to fibrosis on the basis of the type of chelating ther-
`apy, the dichotomous predictor variable. Thus, multivariate logis-
`tic-regression models were formed to examine the relation be-
`tween the dependent variable, progression of hepatic fibrosis, and
`predictor variables, including the duration of deferiprone therapy,
`age at initial biopsy, sex, the presence of antibody to hepatitis C
`virus, initial hepatic iron concentration, and the amount of blood
`transfused.
` Stepwise analysis and an analysis of all possible sub-
`23
`groups were performed to choose the most parsimonious model
`with statistically significant predictors. All tests were two-tailed; a
`P value of 0.05 was considered to indicate statistical significance.
`The BMDP (BMDP Statistical Software, Los Angeles) and S-PLUS
`(Statistical Sciences, version 3.3 for Windows, Seattle) statistical
`computer packages were used for computations.
`The study was approved by the human subjects committee of
`the Hospital for Sick Children, Toronto, and the Health Protection
`Branch of Health Canada. Written informed consent was obtained
`from each patient or the patients’ parents.
`RESULTS
`Effectiveness of Deferiprone
`Among the 18 patients in whom the effectiveness
`of deferiprone could be evaluated, the mean (±SE)
`hepatic iron concentration decreased from 88.7±12.1
`to 65.5±7.9 µmol per gram of liver, wet weight
`(normal value, about 1.6), after a mean of 4.6±0.3
`years of therapy (range, 2 to 7); this decrease of
`23.2±10.9 µmol of iron per gram of liver, wet
`weight, was not significant (P=0.07). Initial and fi-
`nal hepatic iron concentrations are shown in Figure
`1. In seven patients, hepatic iron concentrations at
`the end of treatment met or exceeded the threshold
`value of 80 µmol per gram of liver, wet weight,
`which is associated with an increased risk of cardiac
` The serum ferritin concen-
`disease and early death.
`17
`tration decreased from 4455±841 µg per liter at the
`beginning of treatment to 2831±491 µg per liter at
`the end of treatment. Expressed logarithmically, this
`decrease was significant (P=0.03). In nine of the
`patients, the serum ferritin concentration exceeded
`2500 µg per liter, the threshold used to distinguish
`effective from ineffective chelation therapy.
`12,18
`Data on compliance were available for all 18 patients
`for the last two full years of treatment. The median
`rate of compliance each year was 98 percent, with
`ranges of 90 to 100 percent for the penultimate year
`and 87 to 100 percent for the final year (P=0.70).
`Histologic Analysis
`Of the 72 biopsy specimens available from the pa-
`tients treated with deferiprone, 17 (24 percent) were
`judged inadequate for evaluation. Histologic changes
`could not be evaluated in five patients: two did not
`have two adequate biopsy specimens that had been
`obtained at least one year apart, and three had cirrho-
`sis at the initial evaluation. Thus, a total of 55 biopsy
`specimens from 14 patients were examined (Table 2).
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on August 23, 2017. For personal use only. No other uses without permission.
`
` Copyright © 1998 Massachusetts Medical Society. All rights reserved.
`
`

`

`The New England Journal of Medicine
`
`Patient 1
`
`Patient 2
`
`Patient 3
`
`Patient 4
`
`Patient 5
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`0123456
`
`200
`150
`100
`50
`0
`
`Fibrosis Score
`
`of liver, wet weight)
`Hepatic Iron (µmol/gH
`
`0
`
`2
`
`4
`
`0
`
`2
`
`0
`4
`2
`0
`4
`Years of Treatment
`
`2
`
`4
`
`0
`
`2
`
`4
`
`Figure 2.
` Changes in Histologic Findings and Hepatic Iron Concentrations in the Five Deferiprone-Treated Patients with
`Progression of Hepatic Fibrosis.
`The worst possible fibrosis score is 6; the staging system is described in Table 1.
`
`Five patients treated with deferiprone had evidence
`of progression of fibrosis. The estimated median
`time to progression was 3.2 years. Figure 2 shows
`the initial and final histologic stages, hepatic iron
`concentrations, and duration of therapy for these
`five patients. In four patients, hepatic iron concen-
`trations stabilized during therapy; in the fifth, the
`iron concentration decreased substantially. Figure 3
`shows representative photomicrographs of liver spec-
`imens from the initial and final biopsies in these pa-
`tients. In one other patient, there was an improve-
`ment in the histologic stage over a period of 3.5 years.
`Table 3 shows that deferiprone-treated patients
`with progression of fibrosis were older than those
`without progression (P=0.03). There were no other
`significant differences (sex, prevalence of hepatitis C
`infection, initial hepatic iron concentration, or the
`duration of therapy or amount of blood transfused
`between the initial biopsy and the final biopsy) be-
`tween the two groups.
`Of the 48 biopsy specimens available from the 20
`patients treated with deferoxamine, 8 (17 percent)
`were judged inadequate for evaluation. Histologic
`changes could not be evaluated in eight patients: six
`did not have two adequate biopsy specimens that
`had been obtained at least one year apart, and two
`had cirrhosis at the initial evaluation. Thus, a total
`of 31 biopsy specimens from 12 patients were exam-
`ined (Table 2). None of the specimens showed evi-
`dence of progression of fibrosis. In one patient,
`there was an improvement in the histologic stage
`over a two-year period.
`
`420
`
`·
`
`August 13, 19 98
`
`After adjustment for initial hepatic iron concentra-
`tions, multivariate logistic-regression analysis showed
`that the estimated odds of progression to fibrosis in-
`creased by a factor of 5.8 (95 percent confidence
`interval, 1.1 to 29.6) with each additional year of
`deferiprone treatment. The deferiprone-treated pa-
`tients had a significantly higher mean initial hepatic
`iron concentration than the deferoxamine-treated
`patients (P=0.01) (Table 2), but no significant dif-
`ferences between the two groups were identified
`with respect to age at initial biopsy, sex, prevalence
`of hepatitis C infection, or the duration of therapy
`or amount of blood transfused between the initial
`biopsy and the final biopsy.
`
`Other Adverse Effects
`Deferiprone therapy was not associated with clin-
`ically significant hematologic changes, as evidenced
`by regular blood counts. No characteristic abnormal-
`ities in liver function were observed, although many
`patients had the small elevations in serum amino-
`transferase concentrations that are commonly found
`during iron overload. In no patient did heart disease
`requiring medication develop during the study.
`DISCUSSION
`These results indicate that deferiprone is not an
`effective means of iron-chelation therapy in patients
`with thalassemia major and may be associated with
`worsening of hepatic fibrosis, even in patients whose
`hepatic iron concentrations have stabilized or de-
`creased. After a mean of 4.6 years of deferiprone
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on August 23, 2017. For personal use only. No other uses without permission.
`
` Copyright © 1998 Massachusetts Medical Society. All rights reserved.
`
`

`

`IRON-CHEL ATION THERAPY WITH DEFERIPRONE FOR THAL ASSEMIA MAJOR
`
` Patient 1 Patient 2 Patient 3 Patient 4 Patient 5
`
`Figure 3. Changes in Histologic Findings in the Five Deferiprone-Treated Patients with Progression of Hepatic Fibrosis.
`The top panels show the initial biopsy specimens, and the bottom panels the last biopsy specimens that could be evaluated (Masson’s
`trichrome, ¬100).
`
`therapy, body iron burden was at concentrations as-
`sociated with a greatly increased risk of cardiac disease
`and early death17 in 7 of 18 patients (39 percent).
`Other investigators have recently reported that he-
`patic iron exceeded this threshold in 58 percent of
`patients who were treated with deferiprone for one
`to four years.24 In our patients, differences in objec-
`tively determined rates of compliance or the rate of
`iron loading could not account for the lack of effec-
`tiveness of deferiprone.
`The results of our review of liver-biopsy specimens
`suggest that extended deferiprone therapy may be
`associated with a worsening of hepatic fibrosis. Fi-
`brosis progressed in 5 of the 14 patients (36 per-
`cent) in whom it could be evaluated, despite the sta-
`bilization of or a marked reduction in hepatic iron
`concentrations in all 5 patients. The estimated median
`time to progression of fibrosis was 3.2 years, and
`after adjustment for initial hepatic iron concentra-
`tions, the estimated odds of progression of fibrosis
`increased by a factor of 5.8 (95 percent confidence
`interval, 1.1 to 29.6) with each additional year of de-
`feriprone treatment.
`The worsening of hepatic fibrosis in deferiprone-
`treated patients is in contrast to the arrest of fibrosis
`regularly observed with deferoxamine therapy. In a
`
`TABLE 3. CHARACTERISTICS OF THE PATIENTS WITH PROGRESSION
`OF HEPATIC FIBROSIS DURING TREATMENT WITH DEFERIPRONE
`AND THOSE WITHOUT PROGRESSION.
`
`CHARACTERISTIC
`
`Age at initial biopsy — yr
`Median
`Range
`Male sex — no. (%)
`Antibody to hepatitis C — no. (%)
`Initial hepatic iron concentration —
`µmol/g of liver, wet weight
`Median
`Range
`Duration of treatment — yr*
`Median
`Range
`Amount of blood transfused —
`ml packed cells/kg/yr*
`Median
`Range
`
`PROGRESSION
`OF HEPATIC
`FIBROSIS
`(N=5)
`
`NO
`PROGRESSION
`OF HEPATIC
`FIBROSIS
`(N=9)
`
`21
`18–24
`2 (40)
`4 (80)
`
`16
`10–22
`1 (11)
`2 (22)
`
`29.6
`24.2–180.1
`
`93.6
`55.4–174.2
`
`2.1
`1.6–4.0
`
`2.4
`1.3–3.4
`
`P
`VALUE
`
`0.03
`
`0.50
`0.09
`
`0.07
`
`0.80
`
`77
`56–89
`
`77
`64–114
`
`0.60
`
`*The period under consideration is the interval between the initial biopsy
`and the final biopsy. Data were available for five patients with progression
`of hepatic fibrosis and eight with no progression.
`
`Volume 339 Number 7
`
`· 421
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on August 23, 2017. For personal use only. No other uses without permission.
`
` Copyright © 1998 Massachusetts Medical Society. All rights reserved.
`
`

`

`The New England Journal of Medicine
`
`seminal study, long-term therapy with deferoxamine
`halted the progression of hepatic fibrosis in patients
`with thalassemia major.25-27 In these patients, pro-
`gression was arrested despite a regimen of deferoxa-
`mine (0.5 g per day intramuscularly, six days per
`week)25 now considered suboptimal because it merely
`stabilizes, rather than reduces, hepatic iron concen-
`trations. Moreover, fibrosis was halted despite a final
`mean hepatic iron concentration (139 µmol per
`gram of liver, wet weight)25 that was more than
`twice that in our deferiprone-treated patients with
`progression of fibrosis (49.5 µmol per gram of liver,
`wet weight). Subsequent studies28 and the results in
`our comparison group of deferoxamine-treated pa-
`tients confirm that modern regimens of parenteral
`deferoxamine arrest fibrosis.
`Our findings virtually recapitulate those in a ger-
`bil model of iron overload in which administration
`of a closely related drug (1,2-diethyl-3-hydroxypyri-
`din-4-one) was associated with initial loss of efficacy,
`worsening of hepatic fibrosis, and cardiac fibrosis.13
`There has been concern that hydroxypyridinones may
`exacerbate iron-related tissue damage.29-31 Deferiprone
`is a bidentate chelator, and three molecules are need-
`ed to occupy the six coordination sites of a single
`atom of iron. In contrast, one molecule of the hexa-
`dentate deferoxamine binds a single atom of iron;
`the chelate (ferrioxamine) is virtually inert biologi-
`cally. At low concentrations of deferiprone relative
`to the concentrations of available iron, partially
`bound forms of iron (bound to only one or two
`molecules of deferiprone) appear in which the unoc-
`cupied coordination sites remain reactive and able to
`catalyze the formation of hydroxyl radical or other
`reactive oxygen species.32 There is increasing evidence
`to suggest that these reactive oxygen species are in-
`volved in the pathogenesis of hepatic fibrosis. Recent-
`ly, the potential cellular toxicity of deferiprone has
`been shown in erythrocytes33 and cultured myocytes.34
`The limitations of our histopathological analysis
`should be emphasized: our analysis was retrospec-
`tive, the number of patients studied was small, and
`the patients treated with deferoxamine do not con-
`stitute a true control population. Furthermore, his-
`tologic assessment was based on relatively few biopsy
`specimens, although this was true in both the defer-
`iprone group and the deferoxamine group. Because
`this study was observational rather than randomized
`and the effect of hepatic fibrosis was not confirmed
`by challenge after discontinuation of the drug, the
`relation between deferiprone and fibrosis cannot be
`considered definite or proved. Nonetheless, we could
`identify no other causes of the accelerated fibrosis.
`The patients with progression of fibrosis did not
`differ significantly from those without progression
`with respect to sex, prevalence of hepatitis C infec-
`tion, initial hepatic iron concentrations, duration of
`therapy, or the rate of iron accumulation. The con-
`
`422 · August 13, 19 98
`
`sensus of the pathologists was that there was no dif-
`ference between groups in the type of inflammatory
`changes. Nonetheless, we cannot rule out the possi-
`bility of an interaction between deferiprone and
`hepatitis C infection. The patients with progression of
`fibrosis were older than those without progression,
`and the likelihood of progression of fibrosis was
`greater in patients with lower hepatic iron concen-
`trations. Because we are unable to identify definite
`risk factors for accelerated fibrosis, we have discon-
`tinued deferiprone therapy in all patients, including
`those who are unable or unwilling to use deferoxa-
`mine in standard regimens.
`Despite their limitations, the results of our analy-
`sis, together with theoretical considerations and find-
`ings in animal studies, indicate that deferiprone may
`worsen hepatic fibrosis. Before it can be considered
`for clinical use, even in patients who are unwilling or
`unable to use deferoxamine in standard regimens,
`prospective clinical trials are mandatory to evaluate
`the possibility of irreversible hepatic damage.
`
`Supported in part by research grants from the Medical Research Council
`of Canada, the Ontario Heart and Stroke Foundation, the Ontario Thal-
`assemia Foundation, the Cooley’s Anemia Foundation, Apotex, and the
`National Institutes of Health (DK49108, HL58203, and HL61219). Dr.
`Olivieri is supported in part by a Scientist Award from the Medical
`Research Council of Canada.
`
`We are indebted to Professor Peter Scheuer, London, for his inde-
`pendent analysis of the liver-biopsy slides; to David Nathan, David
`Weatherall, Michael Baker, Antonio Cao, Helen Chan, John Dick,
`Peter Durie, Brenda Gallie, Marc Giacomelli, John Harris, David
`Kern, F.A. Olivieri, Robert Phillips, Eliot Phillipson, Sergio Piomelli,
`Alvin Zipursky, and Stanley Zlotkin for ongoing encouragement,
`advice, and support; and to Naomi Klein, Maria Muraca, Allyson
`Muroff, and Helen Schinkel for assistance in data management.
`
`REFERENCES
`
`1. Cooley’s Anemia Progress Review Committee. Cooley’s anemia:
`progress in biology and medicine. Bethesda, Md.: Division of Blood Dis-
`eases and Resources, 1995.
`2. Fosburg MT, Nathan DG. Treatment of Cooley’s anemia. Blood 1990;
`76:435-44.
`3. Olivieri NF, Brittenham GM. Iron-chelating therapy and the treatment
`of thalassemia. Blood 1997;89:739-61. [Erratum, Blood 1997;89:2621.]
`4. Bergeron RJ, Brittenham GM, eds. Development of iron chelators for
`clinical use. Boca Raton, Fla.: CRC Press, 1994.
`5. Kontoghiorghes GJ, Aldouri MA, Sheppard LN, Hoffbrand AV. 1,2-
`dimethyl-3-hydroxypyrid-4-one, an orally active chelator for treatment of
`iron overload. Lancet 1987;1:1294-5.
`6. Kontoghiorghes GJ, Bartlett AN, Hoffbrand AV, et al. Long-term
`trial with the oral iron chelator 1,2-dimethyl-3-hydroxypyrid-4-one (L1).
`I. Iron chelation and metabolic studies. Br J Haematol 1990;76:295-
`300.
`7. al-Refaie FN, Wonke B, Hoffbrand AV, Wickens DG, Nortey P, Kon-
`toghiorghes GJ. Efficacy and possible adverse effects of the oral iron che-
`lator 1,2-dimethyl-3-hydroxypyrid-4-one (L1) in thalassemia major. Blood
`1992;80:593-9.
`8. al-Refaie FN, Hershko C, Hoffbrand AV, et al. Results of long-term
`deferiprone (L1) therapy: a report by the International Study Group on
`Oral Iron Chelators. Br J Haematol 1995;91:224-9.
`9. Olivieri NF, Koren G, Hermann C, et al. Comparison of oral iron che-
`lator L1 and desferrioxamine in iron-loaded patients. Lancet 1990;336:
`1275-9.
`10. Tondury P, Kontoghiorghes GJ, Ridolfi-Luthy AR, et al. L1 (1,2-
`dimethyl-3-hydroxypyrid-4-one) for oral iron chelation in patients with
`beta-thalassaemia major. Br J Haematol 1990;76:550-3.
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on August 23, 2017. For personal use only. No other uses without permission.
`
` Copyright © 1998 Massachusetts Medical Society. All rights reserved.
`
`

`

`IRON-CHEL ATION THERAPY WITH DEFERIPRONE FOR THAL ASSEMIA MAJOR
`
`11. Agarwal MB, Gupte SS, Viswanathan C, et al. Long-term assessment
`of efficacy and safety of L1, an oral iron chelator, in transfusion dependent
`thalassaemia: Indian trial. Br J Haematol 1992;82:460-6.
`12. Olivieri NF, Brittenham GM, Matsui D, et al. Iron-chelation therapy
`with oral deferiprone in patients with thalassemia major. N Engl J Med
`1995;332:918-22.
`13. Carthew P, Smith AG, Hid