Deviating From Safety Guidelines During Deferiprone Therapy in Clinical Practice May Not be Associated With Higher Risk of Agranulocytosis
Mohssen Elalfy,
MD,
1* Yasser A. Wali,
MD,
2Mohamad Qari,
MD,
3Ghazi Al Damanhouri,
MD,
4Youssef Al-Tonbary,
MD,
5Dilek Yazman,
MD,
6Zakaria Al Hawsawi,
MD,
7Zeynep Karakas,
MD,
8Yurdanur Kilinc,
MD,
9M. Akif Yesilipek,
MD,
10Mohamed Badr,
MD,
11Usama Elsafy,
MD,
11Mostafa Salama,
MD,
12Yousryeia Abdel Rahman,
MD,
13Shebl Shebl,
MD,
14Anne Stilman,
MD,
15Noemi Toiber Temin,
MD,
15and Fernando Tricta,
MD15INTRODUCTION
Deferiprone (Ferriprox1, ApoPharma, Inc., Canada) is an orally active iron chelator that has been shown to reduce iron overload in transfusion-dependent patients [1]. Compared to the other iron chelators, deferoxamine (DFO), and deferasirox (DFX), defer- iprone has been found to have superior efficacy in removing excess cardiac iron, in effecting rapid improvement in cardiac function, and in reducing iron-induced morbidity and mortality [2,3].
However, a factor limiting its widespread use is the occurrence of agranulocytosis, defined as two consecutive absolute neutrophil counts (ANC) less than 0.5109/L, which is seen in 1–2% of deferiprone recipients [1]. The underlying cause of deferiprone- associated agranulocytosis remains obscure; its characteristics suggest an idiosyncratic response [1,4], and its incidence is not dose-related within the recommended doses of up to 100 mg/kg/
day. Agranulocytosis normally resolves upon discontinuation of deferiprone. To avoid delay in its detection, the deferiprone product label recommends weekly monitoring of neutrophil count so as to ensure immediate interruption of therapy if needed.
Out of concern that episodes of neutropenia, defined as ANC between 0.5109/L and 1.5109/L, might progress to agranulo- cytosis during continued deferiprone use, the product label recommends that therapy be interrupted at the first sign of neutropenia. Despite evidence that such progression does not necessarily occur, weekly monitoring and treatment interruption upon detection of neutropenia or signs of infection have been mandated in clinical trials [5]. In clinical practice, however, monitoring is less rigorous. Thus, there is a need to know how often ANC is monitored under real-life conditions, and if deferiprone is in fact discontinued when neutropenia or signs of infection appear.
The objectives of this post-marketing surveillance program were to examine how deferiprone is prescribed and monitored in clinical practice, and to assess its risks and benefits based on the characterization of its use.
METHODS
This was an observational, open-label, prospective, multi- center, non-interventional, post-marketing drug surveillance pro- gram designed to examine how deferiprone is prescribed and to assess how the safety of deferiprone therapy is monitored in clinical practice. Patients were enrolled in the program for up to approximately 1 year. Fifteen treatment sites in four Middle East countries took part: seven sites in Egypt, four in Turkey, three in Saudi Arabia, and one in Oman. The program was approved by local ethics boards, and informed consent was obtained from the patients or their legal representatives prior to enrollment.
There were no restrictions on age, health status, or primary diagnosis. The only exclusion criterion was that patients could not Background.A risk associated with the iron chelator deferiprone
is the development of neutropenia or agranulocytosis. Accordingly, the product label recommends weekly blood monitoring and immediate interruption of treatment upon detection of an absolute neutrophil count (ANC)<1.5109/L, out of concern that continued therapy might lead to a more severe drop. However, it is uncertain how these recommendations are followed under real-life conditions and, if they are not followed, whether continuation of therapy results in increased incidence of agranulocytosis.Procedure. This non- interventional surveillance program assessed the monitoring of deferiprone therapy in clinical practice. A total of 294 patients with transfusion-dependent anemias received deferiprone, as monother- apy or with another chelator, for up to 1 year. The participating
physicians were not given any instructions about treatment and monitoring beyond being referred to the information in the package insert.Results.ANC monitoring was conducted at an average interval of 54 weeks, and deferiprone was not always interrupted upon detection of neutropenia. One patient (0.3%) experienced agranulo- cytosis, and nine others (3%) experienced a total of 11 episodes of neutropenia. All neutropenia episodes resolved; median time to resolution was similar whether or not treatment was interrupted; and no case of neutropenia progressed to agranulocytosis.Conclusions.
These data indicate that less frequent ANC monitoring and continuation of deferiprone therapy during neutropenia are not associated with prolonged neutropenia or with progression to agranulocytosis. Pediatr Blood Cancer #2013 Wiley Periodicals, Inc.
Key words: agranulocytosis; deferiprone; hemoglobinopathies; neutropenia; red blood cell disorders; thalassemia
1Ain Shams University, Cairo, Egypt;2Child Health, Sultan Qaboos University Hospital, Muscat, Oman;3Department of Hematology, King Abdulaziz University Hospital, Jeddah, Saudi Arabia;4King Abdulaziz University Hospital, Jeddah, Saudi Arabia; 5Children’s Hospital, Mansoura University, Mansoura, Egypt;6Department of Hematology, Nicosia, North Cyprus, Burhan Nalbantoglu State Hospital, Nicosia, Cyprus; 7Ministry of Health, Madinah, Saudi Arabia; 8Istanbul University, School of Medicine, Istanbul, Turkey; 9Faculty of Medicine, Department of Pediatric Hematology, Cukurova University, Adana, Turkey;10Pediatric Hematology, Akdeniz University School of Medicine, Antalya, Turkey;11Zagazig University Thalassemia Associ- ation, Zagazig, Egypt;12Paediatric Haematology, Alexandria Univer- sity, Alexandria, Egypt;13Assuit University Hospital, Assuit, Egypt;
14Pediatric Hematology Unit, Tanta University, Egypt;15ApoPharma Inc., Toronto, Canada
Conflict of interest: Nothing to declare.
Correspondence to: Moshen Elalfy, Thalassemia Center, Children Hospital, Ain Shams University, 7 Sobhi Nasr, 5th Dist. Heliopolis, Cairo, Egypt. E-mail: [email protected]
Received 1 August 2013; Accepted 6 December 2013 C 2013 Wiley Periodicals, Inc.
DOI 10.1002/pbc.24920
Published online in Wiley Online Library (wileyonlinelibrary.com).
have been taking deferiprone for more than 1 month prior to enrollment; the reason for this was to increase the likelihood of detecting episodes of agranulocytosis, which is known to occur most frequently during the first 6 months of deferiprone therapy.
The treating physicians determined the dose of deferiprone and whether it was prescribed as monotherapy or in combination with another chelator, and could change the treatment regimen at any time based on patient needs. Physicians were not given any instructions about treatment and monitoring beyond being referred to the information in the package insert. At enrollment, some patients remained on the chelation regimen they were currently following while others were assigned a different regimen.
The treating physicians were requested to collect and transmit the following data within 2 months of availability: start dates, stop dates, dose, and frequency of deferiprone and other chelator(s), if applicable; dates of blood transfusions and amount of blood transfused; dates and results of hematology and blood chemistry testing; dates and results of assessments of iron load (serum ferritin, liver iron concentration, and cardiac iron concentration); and dates and descriptions of adverse events and of concurrent medications.
Data were collected up until completion of approximately 1 year of therapy or until termination from the program, if this occurred earlier than 1 year. Any data inconsistencies were identified and sent back to the site for resolution.
Occurrences of neutropenia were classified as mild if the ANC level was between 1.0 and 1.5109/L, moderate if it was between
0.5 and 1.0109/L, and severe (i.e., agranulocytosis) if it was less than 0.5109/L. All patients were included in the data analysis. No imputation was performed on any missing data; all analyses were based on reported data. Statistical analyses were performed using SAS version 9.1.3. Fisher’s exact test was used in comparing two proportions. For continuous variables, the appropriate t-test was used. A two-sideda-value of 0.05 was used as the significance level in all statistical tests.
RESULTS
Patient Population
A total of 294 patients were enrolled in the program: 150 in Egypt, 60 in Saudi Arabia, 54 in Turkey, and 30 in Oman.
Demographics are shown in Table I. The overall mean age was 12.210.1 years (range: 1–52 years), with 93 of the patients aged between 1 and 5 years, 83 aged between 6 and 11 years, 35 aged between 12 and 15 years, and 83 aged16 years. Just under half (46.6%) were female. The most common primary diagnosis was thalassemia major, in 261 patients, followed by thalassemia intermedia in 17, sickle cell disease in 9, and other anemias (spherocytosis, red cell aplasia/Diamond Blackfan anemia, immune hemolytic anemia, or sideroblastic anemia) in 7. Most patients (230) had previously received chelation therapy, while 64 were naı¨ve to chelation. Of the non-naı¨ve patients, 113 had received monotherapy DFO, 51 had received monotherapy DFX, 34 had
TABLE I. Demographics
Chelation therapy at baseline
Mono DFP (n¼247)
Combination DFP & DFO (n¼35)
Alternating DFP & DFO (n¼6)
Alternating DFP & DFX (n¼6)
Total (n¼294) Age (years)
Mean (SD) 11.59.5 15.611.8 3.32.3 28.89.3 12.210.1
Min, max 1, 43 2, 52 1, 7 11, 36 1, 52
n (%) n (%) n (%) n (%) n (%)
Sex
Female 115 (46.6) 16 (45.7) 3 (50.0) 3 (50.0) 137 (46.6)
Male 132 (53.4) 19 (54.3) 3 (50.0) 3 (50.0) 157 (53.4)
Primary diagnosis
Thalassemia major 216 (87.4) 34 (97.1) 6 (100.0) 5 (83.3) 261 (88.8)
Thalassemia intermedia 15 (6.1) 1 (2.9) 0 (0.0) 1 (16.7) 17 (5.8)
Sickle cell disease 9 (3.6) 0 (0.0) 0 (0.0) 0 (0.0) 9 (3.1)
Other 7 (2.8) 0 (0.0) 0 (0.0) 0 (0.0) 7 (2.4)
Naı¨ve to chelation
Yes 59 (23.9) 4 (11.4) 1 (16.7) 0 (0.0) 64 (21.8)
No 188 (76.1) 31 (88.6) 5 (83.3) 6 (100.0) 230 (78.2)
Splenectomy status
Yes 76 (30.8) 15 (42.9) 0 (0.0) 4 (66.7) 95 (32.3)
No 171 (69.2) 20 (57.1) 6 (100.0) 2 (33.3) 199 (67.7)
Hepatitis B statusy n¼174 n¼32 n¼4 n¼6 n¼216
Positive 2 (1.15) 0 (0.00) 0 (0.00) 0 (0.00) 2
Negative 172 (98.85) 32 (100.00) 4 (100.00) 6 (100.00) 214
Hepatitis C statusy n¼179 n¼32 n¼4 n¼6 n¼221
Positive 34 (18.99) 11 (34.38) 0 (0.00 2 (33.33) 47
Negative 145 (81.01) 21 (65.63) 4 (100.00) 4 (66.67) 174
Neutrophil count at baseline (1.5109/L)
Mean (SD) 5.62 (5.91) 4.71 (2.33) 3.49 (1.8) 7.88 (3.47) 552 (5.53)
Min, max 1.22, 75.24 2.01, 11.98 1.8, 6 4.3, 14.2 1.22, 75.24
DFP, deferiprone; DFO, deferoxamine; DFX, deferasirox.Spherocytosis, red cell aplasia, and sideroblastic anemia.yInformation on hepatitis status was not available for all patients.
received alternating DFO and DFX, and 7 had received combination DFO and DFX. Seventeen patients had received deferiprone within 30 days prior to enrollment in the program, which did not exclude them from participation as they had not been taking it for more than 1 month. (There were no reports of neutropenia during this period of exposure.) With respect to baseline factors that might affect the clinical outcome, two patients were positive for hepatitis B, 47 (16%) were positive for hepatitis C, and 95 (32.3%) had been splenectomized. There was a wide range in baseline values of ANC.
Patients were either continued on their existing regimen or prescribed a different one, using the clinician’s own criteria according to the needs of the patient. At the time of enrollment, 247 patients were receiving deferiprone monotherapy, at dosages ranging from 20 to 100 mg/kg/day; 35 were receiving combination therapy (i.e., simultaneous use) of deferiprone and DFO; six were receiving alternating therapy of deferiprone and DFO (e.g., deferiprone alone for 4 days a week and DFO alone for the other 3 days); and six were receiving alternating therapy of deferiprone and DFX. The treating physicians were free to switch patients at any time during the program from one regimen to another. Of the patients who started the program on a regimen of monotherapy deferiprone, 27 ended it on combination DPFþDFO and three on monotherapy DFX; of those who started on combination deferiproneþDFO, one ended on monotherapy deferiprone; and of those who started on alternating deferiprone/DFO, one ended on monotherapy DFX. The six patients who started on alternating deferiprone/DFX all ended on the same regimen.
Compliance rated as “good” (defined as taking >80% of the prescribed doses during the 12-month period) was higher for deferiprone (70.8%) than for either DFO (40.1%) or DFX (40.0%);
however, for DFX there were too few patients (6) for the comparison to be meaningful.
Disposition of patients is shown in Table II. The majority of participants (90.8%) completed the program. Of the 27 (9.2%) who withdrew early, 14 (4.8%) did so due to adverse events, 9 (3.1%) withdrew voluntarily, and 4 (1.4%) were lost to follow-up. The adverse events leading to withdrawal in the five patients in the monotherapy deferiprone group were pyrexia and chills, arthralgia, vomiting, nausea and vomiting, and stomach disturbances; in the seven patients in the combination deferiproneþDFO group, they were nausea and headache, vomiting, and hypoaesthesia, vomiting and abdominal pain, leukemia, arthralgia, and asthenia; and in the two patients in the alternating deferiprone/DFX group, they were gastritis and pyrexia. No patients in the alternating deferiprone/
DFO group withdrew due to adverse events. There was one death, in
the patient with leukemia. The leukemia was not considered related to chelation treatment.
Frequency of ANC Monitoring
Despite the recommendation on the deferiprone product label that ANC be monitored weekly, the average interval between blood tests for all sites was 54 weeks, with a range from 4 days to 6 months (Fig. 1). There was considerable site variability in the frequency of monitoring. At the site in Oman, the mean was 3 weeks; at the three sites in Saudi Arabia, it ranged from 3 to 4 weeks; at the four sites in Turkey, it ranged from 4 to 7 weeks; and at the seven sites in Egypt, it ranged from 3 to 8 weeks at four sites and from 12 to 15 weeks at the other three.
Safety Findings
While the frequency of ANC monitoring and occurrence of episodes of neutropenia and agranulocytosis were of greatest interest in this program, data on all adverse events were collected.
Seventy patients (23.8%) reported at least one adverse event considered to be at least possibly related to deferiprone treatment.
The most common adverse events related to deferiprone treatment were vomiting in 26 patients (8.8%), nausea in 18 (6.1%), arthralgia in 12 (4.1%), increased alanine aminotransferase in 9 (3.1%), and neutropenia in 8 (2.7%). No other adverse event considered related to deferiprone was seen in more than 2% of patients.
There was one death, in a 52-year-old female with thalassemia intermedia who was being treated with combination deferiprone þDFO. The patient was diagnosed with acute myeloid leukemia
Fig. 1. Frequency of monitoring.
TABLE II. Disposition
Chelation therapy at baseline
Mono DFP (n¼247); n (%)
Combination DFP & DFO (n¼35); n (%)
Alternating DFP & DFO (n¼6); n (%)
Alternating DFP & DFX (n¼6); n (%)
Total (n¼294); n (%) Status
Completed 233 (94.3) 26 (74.3) 5 (83.3) 3 (50.0) 267 (90.8)
Withdrawn 14 (5.7) 9 (25.7) 1 (16.7) 3 (50.0) 27 (9.2)
Reason for withdrawal
Adverse event 5 (2.0) 7 (20.0) 0 (0.0) 2 (33.3) 14 (4.8)
Voluntary withdrawal 5 (2.0) 2 (5.7) 1 (16.7) 1 (16.7) 9 (3.1)
Lost to follow-up 4 (1.6) 0 (0.0) 0 (0.0) 0 (0.0) 4 (1.4)
DFP; deferiprone; DFO, deferoxamine; DFX, deferasirox.
(AML) about 4 months after enrollment in the program, and died 4 days later. The death was due to the AML, and was not considered related to the chelation treatment.
A total of 11 episodes of neutropenia were detected in nine patients (3.1%), with two individuals experiencing two episodes each. (In one case, the episodes were 17 days apart; in the other, they were 78 days apart. Both patients had had at least two neutrophil counts above 1.5109/L in between the episodes.) All but one of the 11 episodes were considered to be related to deferiprone treatment. With respect to baseline characteristics of these 9 patients, the primary diagnosis was thalassemia major for 8 and thalassemia intermedia for the ninth; one patient was splenectom- ized and eight were not; two were positive for hepatitis C, nine were negative, and hepatitis C status was unknown for the other three;
and mean baseline value of ANC was 2.6109/L, which was lower than the mean of 5.61109/L for the 285 patients who did not develop neutropenia, but not significantly different. The treatment regimen at the start of the program was monotherapy deferiprone for eight of the patients and combination deferiproneþDFO for one.
The mean time between ANC monitoring for these nine patients was 2.90.6 weeks (range: 2.2–3.9 weeks). Treatment was interrupted upon detection of neutropenia in seven cases and was continued in the other four. Resolution (defined as two consecutive ANC levels above 1.5109/L) occurred in all cases, with a median
time to resolution of 35 days (range: 8–167 days). For the seven cases in which treatment was interrupted, median time to resolution was 35 days (range: 10–92 days); for the four in which it was continued, it was 38 days (range: 8–167 days). The difference in these times was not significant (P¼0.7768). None of the patients were treated with granulocyte colony stimulating factor (G-CSF). Overall, the frequency of monitoring post onset of neutropenia ranged from 1 to 91 days, which may have affected the data for the time to resolution (i.e., in the case of longer intervals, resolution may have occurred in the interim but gone undetected).
There was one episode of agranulocytosis, in a 17-year-old female with thalassemia intermedia who was receiving deferiprone monotherapy at a dose of 50 mg/kg/day. The mean time between monitoring visits for this patient was 1.9 weeks. The agranulocyto- sis was detected 3 weeks after the start of treatment; the last ANC monitoring had been done 2 weeks earlier, at which time there had been no sign of neutropenia. Deferiprone therapy was interrupted, and the patient was treated with G-CSF and other medications. She did not require hospitalization. The event resolved (two consecutive ANC levels above 1.5109/L) after 12 days, and therapy with deferiprone was re-initiated upon recovery. The patient discon- tinued deferiprone 57 days after rechallenge due to an episode of fever and chills, but that was not associated with decreased neutrophil count.
Fig. 2. Absolute neutrophil count in selected patients.
Figure 2 presents four graphical depictions of ANC over time, with each panel representative of a different scenario seen in the program. Panel A depicts the ANCs of a patient who had 1 episode of neutropenia for which treatment was interrupted; Panel B, those of a patient who had one episode and did not have treatment interrupted; Panel C, those of a patient who had two episodes for which treatment was interrupted the first time but not the second;
and Panel D, those of the patient who had agranulocytosis, recovered, and later withdrew for other reasons. With the exception of the second episode of neutropenia shown in Panel C, the drop in ANC in all cases occurred shortly after the start of treatment. In the case of the patient with two episodes of neutropenia, the episodes were far apart in time: that is, the second episode did not occur immediately after resumption of treatment. ANC increased following all episodes, regardless of whether treatment was stopped or not.
With respect to efficacy findings, mean serum ferritin of the 282 patients who had a baseline assessment and at least one follow-up assessment declined from 2,8582,481mg/L at baseline to 2,4422,014mg/L at the last assessment, for a mean decrease of 4161,451mg/L (P<0.0001). With respect to cardiac iron and liver iron levels, only 14 patients contributed data on these measures and the techniques varied among the participating centers, so no valid conclusions could be drawn.
DISCUSSION
Realistically, the recommendations on a product label for minimization of risk may sometimes be logistically difficult to follow due to patient location or other factors. This active surveillance program is the first to look at the following: (1) how often neutrophil count is monitored during deferiprone therapy in clinical practice in the regions where this study was carried out; (2) whether treating physicians are interrupting deferiprone therapy upon detection of ANC below 1.5109/L, as recommended in the product label; and (3) what are the safety consequences of infrequent monitoring and continuation of deferiprone treatment during neutropenia. Findings showed that the majority of patients in this program had their ANC monitored less frequently—in many cases far less frequently—than the weekly interval recommended in the product label; and that deferiprone treatment was not consistently interrupted upon detection of neutropenia. Despite this, only one case of agranulocytosis occurred, and there were no episodes of infection secondary to agranulocytosis. The episode of agranulocytosis occurred soon after initiation of therapy, which is consistent with previous reports; it was in a patient who was receiving deferiprone at a dose of 50 mg/kg/day, which is consistent with the available data that the event is not dose-dependent within the recommended dosage of up to 100 mg/kg/day; and it resolved within 12 days, which is consistent with the median time to resolution of 10 days seen in clinical trials [6]. Of the patients with neutropenia, two had concurrent infections: one experienced bronchitis on the same day that the neutropenia was diagnosed, and the other experienced acute tonsillitis 1 day before the neutropenia ended. The single fatality that occurred during the program was considered unrelated to chelation therapy.
The reported incidence of neutropenia in 3.1% of patients is significantly lower than the rate of 6.7% seen in the combined safety database (N¼642) of 11 ApoPharma-sponsored clinical trials of deferiprone (P¼0.0302). It must be noted, however, that because of
the frequently long intervals between blood tests, the actual rate may have been higher, with transient episodes going undetected because the neutrophil count could have risen above the threshold value by the time of the next ANC measurement. That is, if there had been more frequent monitoring, the study might have detected more occurrences of transient neutropenia. Another observation was that the median time between the diagnosis of neutropenia and its resolution was 35 days (range: 8–167 days), which is longer than the median of 10 days observed in clinical trials. Again, it is possible that this difference was mainly due to the longer interval between ANC monitoring: that is, resolution may have occurred earlier but was not detected until the patient was next tested. The fact that the time to resolution of neutropenia did not differ for events for which deferiprone therapy was interrupted and those for which it was not supports the likelihood that episodes of neutropenia occur in thalassemia patients irrespective of defer- iprone therapy.
The incidence of neutropenia during deferiprone therapy also needs to be considered in light of increased knowledge of the rates of neutropenia in normal populations and in patients with thalassemia in particular. While neutropenia has been defined as an ANC of <1.5109/L, population data suggest that there is considerable demographic diversity, and that using the same threshold for all individuals may be inappropriate. According to Hsieh et al. [7], benign reductions in neutrophil count are more common at certain ages and in certain ethnic groups. In a population-based cross-sectional study involving 25,222 presum- ably healthy participants in the United States whose blood cells count was measured only once, it was determined that relative to whites, blacks had lower leukocyte and neutrophil counts (although similar lymphocyte counts) while Mexican-Americans had higher counts on all three measures. Using the criterion of ANC
<1.5109/L, the prevalence of neutropenia was found to be 4.5% among black participants, 0.8% among white participants, and 0.4% among Mexican-American participants, and was higher among males and children younger than 5 years of age. This variability suggests that a diagnosis of neutropenia should take age, sex, and ethnicity into account.
Furthermore, the rate of neutropenia may be higher in thalassemia patients than in the general population, likely due to hypersplenism. In clinical trials conducted by ApoPharma, while the percentage of thalassemia patients experiencing episodes of neutropenia was slightly higher in deferiprone-treated groups (7.3%, N¼683) than in DFO-treated groups (4.2%, N¼118), the difference was not statistically significant (P¼0.4111), and may have been due to a longer follow-up of deferiprone patients (a mean of 2.1 years compared to 1.1 years for DFO patients). Another issue is that ANC monitoring was done weekly for deferiprone-treated patients but only at the time of blood transfusions for DFO-treated patients. Hence, it is possible that some cases of transient neutropenia were missed in DFO patients because they had resolved before the next blood test, whereas most cases of transient neutropenia in deferiprone patients were detected.
Regardless of the relationship between deferiprone and neutropenia, the reality is that weekly ANC monitoring is sometimes not feasible. Given that most cases of agranulocytosis develop soon after the commencement of deferiprone therapy, and that rigid adherence to the recommended schedule of blood collection is impractical for some patients, one author reports that it is his center’s practice to do weekly counts for the first 2 months
only and thereafter do them monthly, along with instructing patients to go immediately to the nearest hospital in case of fever [8]. Such practice appears to be common among the centers participating in this program.
The important finding is that in the current program, continuation of deferiprone therapy in patients with detected (and quite likely with undetected) events of neutropenia was not associated with progression to agranulocytosis. This finding is consistent with that of a clinical trial that demonstrated that episodes of mild neutropenia did not progress to agranulocytosis when deferiprone treatment was continued [9]. The results of this surveillance program support the hypothesis that events of mild or moderate neutropenia may commonly occur in thalassemia patients irrespective of deferiprone therapy, and are likely transient events that do not lead to agranulocytosis. These findings raise the question of whether there is a clinical benefit to interrupting deferiprone therapy at the first sign of neutropenia without signs of infection.
In conclusion, real-life clinical conditions often do not match those mandated in the closely managed environment of clinical trials. At the participating treatment centers in the Middle East, monitoring of ANC often was conducted at considerably longer intervals than the 1 week recommended in the product label, and deferiprone therapy was not always interrupted upon detection of neutropenia. However, these circumstances were not associated with prolonged duration of neutropenia, progression of neutropenia to agranulocytosis, or any other safety concern. In light of these
results, the benefit/risk ratio of stopping deferiprone when ANC falls below 1.5109/L needs to be re-evaluated.
ACKNOWLEDGMENTS
Drs. Elalfy, Wali, Qari, Al Damanhouri, Al-Tonbary, Yazman, Al Hawsawi, Karakas, Kilinc, Yesilipek, Badr, Elsafy, Salama, Rahman, and Shebl contributed the data for this study; A. Stilman contributed to the writing; N. Toiber Temin analyzed the data and created the figures; and F. Tricta contributed to the design of the study and to the writing.
REFERENCES
1. Galanello R, Campus S. Deferiprone chelation therapy for thalassemia major. Acta Haematol 2009;122:155–164.
2. Pennell DJ, Berdoukas V, Karagiorga M, et al. Randomized controlled trial of deferiprone or deferoxamine in beta-thalassemia major patients with asymptomatic myocardial siderosis. Blood 2006;107:3738–3744.
3. Borgna-Pignatti C, Cappellini MD, De Stefano P, et al. Cardiac morbidity and mortality in deferoxamine- or deferiprone-treated patients with thalassemia major. Blood 2006;107:3733–3737.
4. Masera N, Tavecchia L, Longoni DV, et al. Agranulocytosis due to deferiprone: A case report with cytomorphological and functional bone marrow examination. Blood Transfus 2011;4:1–4.
5. Cohen AR, Galanello R, Piga A, et al. Safety profile of the oral iron chelator deferiprone: A multicentre study. Br J Haematol 2000;108:305–312.
6. Food and Drug Administration, Oncologic Drugs Advisory Committee (ODAC) Briefing Document NDA # 21-825 Ferriprox1(deferiprone). http://www.fda.gov/downloads/AdvisoryCommittees/Com- mitteesMeetingMaterials/drugs/OncologicDrugsAdvisoryCommittee/ucm271538.pdf
7. Hsieh MM, Everhart JE, Byrd-Holt DD, et al. Prevalence of neutropenia in the U.S. population: Age, sex, smoking status, and ethnic differences. Ann Intern Med 2007;146:486–492.
8. Wali Y, Shidhani AA, Daar S. Agranulocytosis in beta thalassemia major patients treated with oral iron chelating agent (deferiprone). Oman Med J 2008;23:275–277.
9. ElAlfy MS, Sari TT, Lee CL, et al. The safety, tolerability, and efficacy of a liquid formulation of deferiprone in young children with transfusional iron overload. J Pediatr Hematol Oncol 2010;32:601–605.