Int J Hematol (2014) 100:254–259 DOI 10.1007/s12185-014-1624-9
ORIGINAL ARTICLE
Elevated total iron-binding capacity as a predictor of response to deferasirox therapy in the setting of chronic iron overload Junichi Watanabe • Ken Sato • Toshikatsu Horiuchi • Shoichiro Kato • Reina Hikota • Takaaki Maekawa • Takeshi Yamamura • Ayako Kobayashi Yukiko Osawa • Shinichi Kobayashi • Fumihiko Kimura
•
Received: 25 February 2014 / Revised: 16 June 2014 / Accepted: 16 June 2014 / Published online: 2 July 2014 Ó The Japanese Society of Hematology 2014
Abstract It is difficult to predict the efficacy of deferasirox (DFX) as its pharmacokinetics varies among patients. The area under the curve (AUC) is reportedly useful for determining adequate DFX dosage; however, serum concentration measurements are often challenging. Effective DFX dosage is thus defined by assessing the efficacy of this agent in clinical practice. To analyze a predictive response marker to DFX therapy for use in adjusting the effective dosage during the early treatment phase, we retrospectively evaluated 39 DFX-treated patients. We defined response as a [40 % decrease in serum ferritin concentration from the pretreatment level. A maximum elevation of the total iron-binding capacity (TIBC) correlated with response in a multivariate analysis of iron metabolic markers (R2 = 0.37, p \ 0.001). A receiver operating characteristic curve analysis revealed that TIBC elevation had an AUC of 0.85 (p \ 0.001) and the optimal cut-off value of TIBC elevation was 150 lg/dl. TIBC elevation of [150 lg/dl is a favorable predictor of effective ferritin reduction in DFX therapy (hazard ratio 29.6, 95 % confidence interval 4.8–183.6; p \ 0.001). DFX therapy with TIBC monitoring may enable the determination of the minimum effective DFX dosage. Keywords Total iron-binding capacity (TIBC) Deferasirox Chronic iron overload
J. Watanabe (&) K. Sato T. Horiuchi S. Kato R. Hikota T. Maekawa T. Yamamura A. Kobayashi Y. Osawa S. Kobayashi F. Kimura Division of Hematology, Department of Internal Medicine, National Defense Medical College, 3-2 Namiki, Tokorozawa, Saitama 359-8513, Japan e-mail: [email protected]
123
Introduction Chronic iron overload is a serious complication that can cause organ failure in blood transfusion-dependent patients with hematological diseases such as aplastic anemia and hematological malignancies including myelodysplastic syndromes (MDS) [1, 2]. Iron chelation therapy for chronic iron overload improves both survival and quality of life [1–3]. Some clinical studies have demonstrated a benefit from the oral iron chelator deferasirox (DFX) in patients suffering from iron overload [4–12]. However, the effective DFX dosage needed to improve the iron overload differs for each patient because of differences in red blood cell (RBC) transfusion. Moreover, the pharmacokinetics of DFX differs among patients, and the area under the curve (AUC) of DFX significantly correlates with the response to DFX [13]. The AUC measurements for DFX could be useful in determining the effective DFX dosage; however, this analysis is difficult to perform in clinical practice. Actually, we need to adjust the DFX dosage according to changes in the patients’ ferritin levels. In the multicenter Evaluation of Patients’ Iron Chelation with Exjade (EPIC) study, 20.4 % patients required a decrease in or cessation of the DFX dosage because of the adverse events [4]. Therefore, an early determination of the minimum effective DFX dosage may enable clinicians to treat patients effectively. We experienced some patients who exhibited remarkable total iron-binding capacity (TIBC) increases prior to apparent ferritin level reductions during DFX therapy. We retrospectively analyzed the relationships between iron metabolic markers and ferritin reduction to identify a simple marker that could facilitate early determination of the effective DFX dosage.
TIBC elevation prior to DFX response
Design and methods This study retrospectively reviewed 24 males and 15 females (n = 39) who had received frequent RBC transfusions and began receiving DFX therapy between August 2008 and November 2012 at the National Defense Medical College Hospital. Thirteen patients had been diagnosed with MDS, 4 with aplastic anemia, 3 with myelodysplastic/ myeloproliferative neoplasms, 16 with acute leukemia, 2 with malignant lymphoma, and 1 with pure red cell aplasia. All patients had serum ferritin concentrations [1000 lg/l when DFX therapy was initiated. An initial DFX dose (10 or 20 mg/kg) was administered to these patients and was gradually increased with respect to efficacy and safety. The patients’ blood samples were analyzed every 2 or 4 weeks to determine the serum iron, unsaturated iron binding capacity (UIBC), TIBC, and ferritin levels. The QuickAuto Neo Fe and QuickAuto Neo UIBC assays (Shino-Test Corp., Tokyo, Japan) were used to measure the serum iron and UIBC levels, respectively. These assays use 2-nitroso5-(N-propyl-N-sulfopropylamino)-phenol as the chromogen and react with samples at 37 °C. A patient who achieved a serum ferritin level reduction of [40 % from the baseline level during DFX chelation therapy was defined as a ‘‘responder’’ in accordance with the EPIC study, in which chelation-naive patients in the MDS group exhibited a median decrease of 35 % in their serum ferritin levels [7]. Regarding the other markers of iron metabolism, the changes in marker levels from the baseline (before DFX therapy initiation) were calculated. Statistical analyses The patient characteristic analysis, logistic regression analysis, and receiver operating characteristic (ROC) curve analysis were performed with JMP10 software (SAS Institute Inc., Cary, NC, USA). Gray tests and the Fine– Gray competing risks model were performed using the Stata software (StataCorp LP, College Station, TX, USA). Statistical significance was defined as a p value of \0.05.
Results The median follow-up period was 24 weeks (range 4–24 weeks). The median age at the DFX therapy initiation was 60 years (range 18–86 years), and the median baseline serum ferritin concentration was 3,034 lg/l (range 1,075–10,560 lg/l). Twenty-three patients responded to DFX. There were no significant differences between the responders and non-responders with regard to baseline characteristics such as age, body weight, duration of DFX therapy, baseline serum ferritin levels, organ function, and
255
iron markers according to the Wilcoxon rank sums test or underlying diseases according to Fisher’s exact test and the Chi-square test (Table 1). Two patients required therapy discontinuation or dosage reduction because of the adverse events without achieving a [40 % ferritin reduction; 1 patient stopped DFX use because of diarrhea, whereas the other patient required a temporarily decreased DFX dosage because of diarrhea but returned to the increased dosage after exhibiting improvement. To determine a predictive response marker, we initially performed logistic regression analyses with DFX response as the outcome. In addition to iron metabolic markers, the maximum DFX dose, number of RBC transfusion units, duration of DFX use, and DFX dosage per unit of RBC transfusion during DFX therapy were included as variables. The univariate analyses indicated that the maximum TIBC elevation (R2 = 0.37, p \ 0.001), maximum UIBC elevation (R2 = 0.26, p \ 0.001), minimum transferrin saturation (R2 = 0.25, p \ 0.001), maximum iron elevation (R2 = 0.11, p = 0.021), and DFX dosage per unit of RBC transfusion (R2 = 0.27, p = 0.003) during DFX therapy were significant variables. Multivariate analyses were performed using 2 variable combinations. First, the number of RBC transfusion units, DFX dosage, duration of DFX use, and maximum TIBC elevation were analyzed as covariates. Second, the DFX dosage per unit of RBC transfusion was included instead of the number of RBC transfusion units and DFX dosage. Stepwise multivariate analyses revealed that the maximum TIBC elevation was primarily related to responders in both analyses (R2 = 0.37, p \ 0.001; Table 2). Next, we performed the ROC curve analysis to determine the TIBC elevation cutoff value that would most effectively predict a [40 % serum ferritin reduction. TIBC elevation had an AUC of 0.85 (p \ 0.001; Fig. 1a). The optimal cut-off value of TIBC elevation was 150 lg/dl, and this value had a sensitivity, specificity, positive-predictive value, and negativepredictive value of 0.80, 0.93, 0.94, and 0.78, respectively. Furthermore, we compared the cumulative incidence of a [40 % ferritin reduction in the 2 groups with and without TIBC elevations [150 lg/dl over the baseline level during DFX therapy while considering the mortality without a [40 % ferritin reduction as a competing risk. We defined an achievement of a [150 lg/dl TIBC increase as a timevarying covariate. The DFX therapy response incidence was significantly higher in patients with a TIBC elevation [150 lg/dl (Gray’s test, p \ 0.001; Fig. 1b). The changes between the ferritin levels measured before and after DFX for each patient are shown in Fig. 2. The Fine–Gray competing risks model confirmed that a TIBC elevation [150 lg/dl is a favorable predictive factor for effective ferritin reduction in response to DFX therapy (hazard ratio 29.6, 95 % confidence interval 4.8–183.6; p \ 0.001;
123
256 Table 1 Baseline characteristics of patients
J. Watanabe et al.
Characteristics
Responder (n = 23)
Non-responder (n = 16)
p value
Normal range
Age (years)
60
64
0.85
Body weight (kg)
54.1
53.6
1.00
Duration of taking DFX (weeks)
8
8
0.13
Hb (g/dl)
7.6
6.95
0.36
M 13.0–17.0, F 12.0–16.0
Serum ferritin (lg/l)
3430
2725
0.20
M 23–250, F 3.4–120
Transferrin saturation (%)
87.7
82.2
0.79
TIBC (lg/dl)
253
226
0.46
M 235–385, F 220–420
Fe (lg/dl)
179
170
0.82
M 80- 160, F 70-160
Creatinine (mg/dl)
0.76
0.76
0.71
ALT (IU/L)
30
40
0.47
M 0.61–1.13, F 0.44–0.78 M 5–35, F 5–35
10
9
MDS
6
7
Pure red cell aplasia
1
0
Aplastic anemia
2
2
IMF
1
0
13
7
Baseline
Underlying disease
0.43
Bone marrow failure Data are expressed as median except disease DFX deferasirox, Hb hemoglobin, TIBC total ironbinding capacity, ALT alanine aminotransferase, MDS myelodysplastic syndrome, IMF idiopathic myelofibrosis, AML acute myeloid leukemia, ALL acute lymphoblastic leukemia, MPN myeloproliferative neoplasms, M male, F female
Hematological malignancy AML
10
4
ALL
1
1
Malignant lymphoma MDS/MPN
1
1
1
1
Discussion
Table 2 Regression analysis of response Univariate Unit of RBC transfusion during this study (units) Maximum dose of DFX (mg) Minimum transferrin saturation (%) Duration of taking DFX (weeks)
Multivariate
0.253 0.685 \0.001 0.109
Maximum elevation of TIBC (lg/dl)
\0.001
Maximum elevation of UIBC (lg/dl)
\0.001
Maximum elevation of iron (lg/dl)
0.021
DFX/RBC/day (mg/unit/day)
0.003
\0.001
RBC red blood cell, DFX deferasirox, TIBC total iron-binding capacity, UIBC unsaturated iron-binding capacity
Table 3). A TIBC elevation [150 lg/dl was observed at a median of 5 weeks earlier than the achievement of a [40 % serum ferritin reduction. Representative cases are shown, including 2 response cases that achieved TIBC elevations[150 lg/dl and 1 non-response case that did not achieve (Fig. 3).
123
Transferrin is an iron-binding transport glycoprotein with a molecular weight of 76.5 kDa and a biological half-life of 7–9 days. Low hepatocyte iron levels regulate the transcriptional level of transferrin [14, 15]. TIBC is calculated as the sum of UIBC and serum iron concentration. However, measurements of UIBC and serum iron are difficult during DFX therapy [16]. The measured levels of UIBC and serum iron have been reported to exceed the actual values during DFX therapy because of the presence of DFX in the serum [16]. UIBC is calculated by measuring the unbound iron after adding excessive iron to a serum sample. Iron-free DFX can bind this extrinsic iron and thus increase the measured UIBC value. Similarly, DFX-binding iron is undistinguishable from transferrin-bound iron and is measured as serum iron in an iron-quantitating assay [16]. Accordingly, the presence of DFX increased the apparent TIBC value, as it represented the sum of UIBC and serum iron concentration. Taken together, TIBC elevation could be a useful surrogate marker for the serum DFX concentration. Our clinical observation suggested that
TIBC elevation prior to DFX response
257
Fig. 1 ROC curve analysis for the association between TIBC elevation and a [40 % ferritin reduction. The optimum cut-off point was defined as the closest point on the curve to the point (X, Y) = (0, 1), where X denotes 1 - specificity and Y denotes sensitivity. The ROC curve reveals an optimal TIBC elevation cut-off point of
Fig. 2 Changes in the serum ferritin levels of patients who did or did not achieve a TIBC elevation of [150 lg/dl (left and right panels, respectively). Solid lines indicate patients who achieved a ferritin reduction [40 % above the baseline level during chelation therapy (responders). Dashed lines indicate the non-responders. All but 1 patient with a TIBC elevation of [150 lg/dl were responders. The excluded patient could not continue DFX therapy because of diarrhea
Table 3 Multivariate analysis for the DFX response by Fine–Gray proportional model HR
95 % CI
p value
Age (years)
1.014
0.99–1.04
0.331
Female
0.186
0.04–0.90
0.036
Body weight (kg)
1.062
0.96–1.18
0.253
Mean DFX dosage Elevation of TIBC [ 150 lg/dl
0.998 29.62
0.99–1.00 4.78–183.6
DFX deferasirox, TIBC total iron-binding capacity
0.104 \0.001
150 lg/dl (a). The cumulative response incidence revealed a statistically significant difference between the groups with and without TIBC elevations of [150 lg/dl (solid line and dashed line, respectively; b)
a TIBC elevation of [150 lg/dl indicates a sufficient serum DFX level for ferritin reduction. Another possibility is that DFX can induce transferrin production. Ikuta et al. [16] reported that the direct transferrin concentration measurement was not elevated in the context of DFX therapy. This finding indicated that the TIBC elevation observed during DFX therapy reflected the DFX concentration rather than the transferrin production. We think that an efficient approach to DFX therapy comprises the initial administration of a small dosage (e.g., 10 mg/kg/day), followed by a gradual increase with TIBC monitoring until the TIBC increases to [150 lg/dl above the baseline, as shown in Fig. 3a. In our analysis, the hazard ratio for effective ferritin reduction was significantly lower among DFX-treated women according to the Fine–Gray competing risks model. Women have a 17.5 % lower clearance with DFX relative to men [17]. Although we did not define the DFX dosage because of the retrospective study design, this difference in clearance observed among women might have an influence on the efficacy of DFX. Our study has several limitations. First, these data should be confirmed by prospective data because our study was a retrospective analysis. We analyzed various blood test time points and therefore needed regular intervals between DFX use and blood tests to assess the DFX pharmacokinetics and pharmacodynamics [18]. Second, TIBC elevation values differ among UIBC and iron assay systems because each system uses a different chromogen. Reportedly, the apparent increases in serum iron and TIBC depend on the assay system and particularly the chromogen [16]. Although we used the most widely used chromogen, a re-evaluation is necessary to determine the cut-off values for other chromogens. We are planning a prospective study to validate the association
123
258
J. Watanabe et al.
Fig. 3 The DFX dosage, RBC transfusion, TIBC elevation (gray column), 150 lg/dl of TIBC elevation (dashed line), and serum ferritin levels (black line) for the representative cases in this study. a A 76-year-old female with MDS 5q syndrome was treated with lenalidomide and regular RBC transfusion. We increased the DFX dosage gradually every 2 weeks and maintained the dosage after observing a TIBC elevation of [150 lg/dl. The patient achieved a ferritin reduction of [40 % from the baseline level after 8 weeks and achieved a TIBC elevation of [150 lg/dl. b A 79-year-old male with PRCA was treated with cyclosporine and regular RBC transfusion. He initiated treatment with 20 mg/kg of DFX and maintained this dosage.
He quickly exhibited a TIBC elevation of [500 lg/dl after 2 weeks but a slight serum ferritin reduction of only 6 % at that time. Within 8 weeks, he achieved sufficient ferritin reduction and became RBC transfusion-independent. c The third case was a non-responder. A 73-year-old male with relapsed AML was treated with azacitidine and regular RBC transfusion. We initiated DFX treatment at a 10 mg/kg dose. His serum ferritin level decreased slightly and concomitantly with a slight TIBC increase, but his TIBC did not increase of more than 150 lg/dl. His TIBC decreased when his condition worsened, and his serum ferritin level rebounded
between TIBC elevation and other factors such as transferrin, DFX concentration, non-transferrin bound iron, liver iron concentration, and serum ferritin reduction. In conclusion, our study revealed that TIBC elevation occurs before a favorable response to DFX therapy. TIBC monitoring during DFX therapy may provide indications for DFX dosage adjustments in individual patient and may help clinicians to more effectively treat patients with iron overload using DFX.
6. Vichinsky E, Onyekwere O, Porter J, Swerdlow P, Eckman J, Lane P, et al. A randomized comparison of deferasirox versus deferoxamine for the therapy of transfusional iron overload in sickle cell disease. Br J Haematol. 2006;136:501–8. 7. Gattermann N, Finelli C, Porta MD, Fenaux P, Ganser A, Guerci-Bresler A, et al. Deferasirox in iron-overloaded patients with transfusion-dependent myelodysplastic syndromes: Results from the large 1-year EPIC study. Leuk Res. 2010;34:1143–50. 8. Taher A, Cappellini MD, Vinchinsky E, Galanello R, Piga A, Lawniczek T, et al. Efficacy and safety of deferasirox dosage of [30 mg/kg per d in patients with transfusion-dependent anemia and iron overload. Br J Haematol. 2009;147:752–9. 9. Lee JW, Yoon SS, Shen ZX, Ganser A, Hsu HC, Habr D, et al. Iron chelating therapy with deferasirox in patients with aplastic anemia: a subgroup analysis of 116 patients from the EPIC trial. Blood. 2010;116:2448–54. 10. Porter J, Galanello R, Saglio G, Neufeld EJ, Vichinsky E, Cappellini MD, et al. Relative response of patients with myelodysplastic syndromes and other transfusion dependent anaemias to deferasirox (ICL670): a 1-yr prospective study. Eur J Haematol. 2007;80:168–76. 11. Porter JB, Elalfy MS, Taher AT, Aydinok Y, Chan LL, Lee SH, et al. Efficacy and safety of deferasirox at low and high iron burdens: results from the EPIC magnetic resonance imaging substudy. Ann Hematol. 2013;92:211–9. 12. Metzgeroth G, Dinter D, Schultheis B, Dorn-Beineke A, Lutz K, Leismann O, et al. Deferasirox in MDS patients with transfusion-caused iron overload-a phase-2 study. Ann Hematol. 2009;88:301–10. 13. Chirnomas D, Smith AL, Braunstein J, Finkelstein Y, Pereira L, Bergmann AK, et al. Deferasirox pharmacokinetics in patients with adequate versus inadequate response. Blood. 2009;114:4009–13.
Conflict of interest
The authors declare no conflicts of interest.
References 1. Kushner JP, Porter JP, Olivieri NF. Secondary iron overload. Hematol Am Soc Hematol Educ Program. 2001;1:47–61. 2. Kikuchi S, Kobune M, Iyama S, Sato T, Murase K, Kawano Y, et al. Prognostic significance of serum ferritin level at diagnosis in myelodysplastic syndrome. Int J Hematol. 2012;95:527–34. 3. Cohen AR. New advances in iron chelation therapy. Hematol Am Soc Hematol Educ Program. 2006;1:42–7. 4. Cappellini MD, Porter J, El-Beshlawy A, Li CK, Seymour JF, Elalfy M, et al. Tailoring iron chelation by iron intake and serum ferritin: the prospective EPIC study of deferasirox in 1744 patients with transfusion-dependent anemias. Haematologica. 2010;95:557–66. 5. Cappellini MD, Cohen A, Piga A, Bejaoui M, Perrotta S, Agaoglu L, et al. A phase 3 study of deferasirox (ICL670), a once-daily oral iron chelator, in patients with B-thalassemia. Blood. 2006;107:3455–62.
123
TIBC elevation prior to DFX response 14. Kohgo Y, Ikuta K, Ohtake T, Torimoto Y, Kato J. Body iron metabolism and pathophysiology pf iron overload. Int J Hematol. 2008;88:7–15. 15. Crichton RR, Charloteauc-Wauters M. Iron transport and storage. Eur J Biochem. 1987;164:485–506. 16. Ikuta K, Ito S, Tanaka H, Sasaki K, Torimoto Y, Kohgo Y. Interference of deferasirox with assays for serum iron and serum unsaturated iron binding capacity during iron chelating therapy. Clin Chim Acta. 2011;412:2261–6.
259 17. Exjade. Highlight of prescribing information. https://www. pharma.us.novartis.com/product/pi/pdf/exjade.pdf 18. Miyazawa K, Ohyashiki K, Urabe A, Hata T, Nakao S, Ozawa K, et al. A safety, pharmacokinetic and pharmacodynamics investigation of deferasirox (Exjade, ICL670) in patients with transfusion-dependent anemias and iron overload: a phase I study in Japan. Int J Hematol. 2008;88:73–81.
123
ORIGINAL ARTICLE
Elevated total iron-binding capacity as a predictor of response to deferasirox therapy in the setting of chronic iron overload Junichi Watanabe • Ken Sato • Toshikatsu Horiuchi • Shoichiro Kato • Reina Hikota • Takaaki Maekawa • Takeshi Yamamura • Ayako Kobayashi Yukiko Osawa • Shinichi Kobayashi • Fumihiko Kimura
•
Received: 25 February 2014 / Revised: 16 June 2014 / Accepted: 16 June 2014 / Published online: 2 July 2014 Ó The Japanese Society of Hematology 2014
Abstract It is difficult to predict the efficacy of deferasirox (DFX) as its pharmacokinetics varies among patients. The area under the curve (AUC) is reportedly useful for determining adequate DFX dosage; however, serum concentration measurements are often challenging. Effective DFX dosage is thus defined by assessing the efficacy of this agent in clinical practice. To analyze a predictive response marker to DFX therapy for use in adjusting the effective dosage during the early treatment phase, we retrospectively evaluated 39 DFX-treated patients. We defined response as a [40 % decrease in serum ferritin concentration from the pretreatment level. A maximum elevation of the total iron-binding capacity (TIBC) correlated with response in a multivariate analysis of iron metabolic markers (R2 = 0.37, p \ 0.001). A receiver operating characteristic curve analysis revealed that TIBC elevation had an AUC of 0.85 (p \ 0.001) and the optimal cut-off value of TIBC elevation was 150 lg/dl. TIBC elevation of [150 lg/dl is a favorable predictor of effective ferritin reduction in DFX therapy (hazard ratio 29.6, 95 % confidence interval 4.8–183.6; p \ 0.001). DFX therapy with TIBC monitoring may enable the determination of the minimum effective DFX dosage. Keywords Total iron-binding capacity (TIBC) Deferasirox Chronic iron overload
J. Watanabe (&) K. Sato T. Horiuchi S. Kato R. Hikota T. Maekawa T. Yamamura A. Kobayashi Y. Osawa S. Kobayashi F. Kimura Division of Hematology, Department of Internal Medicine, National Defense Medical College, 3-2 Namiki, Tokorozawa, Saitama 359-8513, Japan e-mail: [email protected]
123
Introduction Chronic iron overload is a serious complication that can cause organ failure in blood transfusion-dependent patients with hematological diseases such as aplastic anemia and hematological malignancies including myelodysplastic syndromes (MDS) [1, 2]. Iron chelation therapy for chronic iron overload improves both survival and quality of life [1–3]. Some clinical studies have demonstrated a benefit from the oral iron chelator deferasirox (DFX) in patients suffering from iron overload [4–12]. However, the effective DFX dosage needed to improve the iron overload differs for each patient because of differences in red blood cell (RBC) transfusion. Moreover, the pharmacokinetics of DFX differs among patients, and the area under the curve (AUC) of DFX significantly correlates with the response to DFX [13]. The AUC measurements for DFX could be useful in determining the effective DFX dosage; however, this analysis is difficult to perform in clinical practice. Actually, we need to adjust the DFX dosage according to changes in the patients’ ferritin levels. In the multicenter Evaluation of Patients’ Iron Chelation with Exjade (EPIC) study, 20.4 % patients required a decrease in or cessation of the DFX dosage because of the adverse events [4]. Therefore, an early determination of the minimum effective DFX dosage may enable clinicians to treat patients effectively. We experienced some patients who exhibited remarkable total iron-binding capacity (TIBC) increases prior to apparent ferritin level reductions during DFX therapy. We retrospectively analyzed the relationships between iron metabolic markers and ferritin reduction to identify a simple marker that could facilitate early determination of the effective DFX dosage.
TIBC elevation prior to DFX response
Design and methods This study retrospectively reviewed 24 males and 15 females (n = 39) who had received frequent RBC transfusions and began receiving DFX therapy between August 2008 and November 2012 at the National Defense Medical College Hospital. Thirteen patients had been diagnosed with MDS, 4 with aplastic anemia, 3 with myelodysplastic/ myeloproliferative neoplasms, 16 with acute leukemia, 2 with malignant lymphoma, and 1 with pure red cell aplasia. All patients had serum ferritin concentrations [1000 lg/l when DFX therapy was initiated. An initial DFX dose (10 or 20 mg/kg) was administered to these patients and was gradually increased with respect to efficacy and safety. The patients’ blood samples were analyzed every 2 or 4 weeks to determine the serum iron, unsaturated iron binding capacity (UIBC), TIBC, and ferritin levels. The QuickAuto Neo Fe and QuickAuto Neo UIBC assays (Shino-Test Corp., Tokyo, Japan) were used to measure the serum iron and UIBC levels, respectively. These assays use 2-nitroso5-(N-propyl-N-sulfopropylamino)-phenol as the chromogen and react with samples at 37 °C. A patient who achieved a serum ferritin level reduction of [40 % from the baseline level during DFX chelation therapy was defined as a ‘‘responder’’ in accordance with the EPIC study, in which chelation-naive patients in the MDS group exhibited a median decrease of 35 % in their serum ferritin levels [7]. Regarding the other markers of iron metabolism, the changes in marker levels from the baseline (before DFX therapy initiation) were calculated. Statistical analyses The patient characteristic analysis, logistic regression analysis, and receiver operating characteristic (ROC) curve analysis were performed with JMP10 software (SAS Institute Inc., Cary, NC, USA). Gray tests and the Fine– Gray competing risks model were performed using the Stata software (StataCorp LP, College Station, TX, USA). Statistical significance was defined as a p value of \0.05.
Results The median follow-up period was 24 weeks (range 4–24 weeks). The median age at the DFX therapy initiation was 60 years (range 18–86 years), and the median baseline serum ferritin concentration was 3,034 lg/l (range 1,075–10,560 lg/l). Twenty-three patients responded to DFX. There were no significant differences between the responders and non-responders with regard to baseline characteristics such as age, body weight, duration of DFX therapy, baseline serum ferritin levels, organ function, and
255
iron markers according to the Wilcoxon rank sums test or underlying diseases according to Fisher’s exact test and the Chi-square test (Table 1). Two patients required therapy discontinuation or dosage reduction because of the adverse events without achieving a [40 % ferritin reduction; 1 patient stopped DFX use because of diarrhea, whereas the other patient required a temporarily decreased DFX dosage because of diarrhea but returned to the increased dosage after exhibiting improvement. To determine a predictive response marker, we initially performed logistic regression analyses with DFX response as the outcome. In addition to iron metabolic markers, the maximum DFX dose, number of RBC transfusion units, duration of DFX use, and DFX dosage per unit of RBC transfusion during DFX therapy were included as variables. The univariate analyses indicated that the maximum TIBC elevation (R2 = 0.37, p \ 0.001), maximum UIBC elevation (R2 = 0.26, p \ 0.001), minimum transferrin saturation (R2 = 0.25, p \ 0.001), maximum iron elevation (R2 = 0.11, p = 0.021), and DFX dosage per unit of RBC transfusion (R2 = 0.27, p = 0.003) during DFX therapy were significant variables. Multivariate analyses were performed using 2 variable combinations. First, the number of RBC transfusion units, DFX dosage, duration of DFX use, and maximum TIBC elevation were analyzed as covariates. Second, the DFX dosage per unit of RBC transfusion was included instead of the number of RBC transfusion units and DFX dosage. Stepwise multivariate analyses revealed that the maximum TIBC elevation was primarily related to responders in both analyses (R2 = 0.37, p \ 0.001; Table 2). Next, we performed the ROC curve analysis to determine the TIBC elevation cutoff value that would most effectively predict a [40 % serum ferritin reduction. TIBC elevation had an AUC of 0.85 (p \ 0.001; Fig. 1a). The optimal cut-off value of TIBC elevation was 150 lg/dl, and this value had a sensitivity, specificity, positive-predictive value, and negativepredictive value of 0.80, 0.93, 0.94, and 0.78, respectively. Furthermore, we compared the cumulative incidence of a [40 % ferritin reduction in the 2 groups with and without TIBC elevations [150 lg/dl over the baseline level during DFX therapy while considering the mortality without a [40 % ferritin reduction as a competing risk. We defined an achievement of a [150 lg/dl TIBC increase as a timevarying covariate. The DFX therapy response incidence was significantly higher in patients with a TIBC elevation [150 lg/dl (Gray’s test, p \ 0.001; Fig. 1b). The changes between the ferritin levels measured before and after DFX for each patient are shown in Fig. 2. The Fine–Gray competing risks model confirmed that a TIBC elevation [150 lg/dl is a favorable predictive factor for effective ferritin reduction in response to DFX therapy (hazard ratio 29.6, 95 % confidence interval 4.8–183.6; p \ 0.001;
123
256 Table 1 Baseline characteristics of patients
J. Watanabe et al.
Characteristics
Responder (n = 23)
Non-responder (n = 16)
p value
Normal range
Age (years)
60
64
0.85
Body weight (kg)
54.1
53.6
1.00
Duration of taking DFX (weeks)
8
8
0.13
Hb (g/dl)
7.6
6.95
0.36
M 13.0–17.0, F 12.0–16.0
Serum ferritin (lg/l)
3430
2725
0.20
M 23–250, F 3.4–120
Transferrin saturation (%)
87.7
82.2
0.79
TIBC (lg/dl)
253
226
0.46
M 235–385, F 220–420
Fe (lg/dl)
179
170
0.82
M 80- 160, F 70-160
Creatinine (mg/dl)
0.76
0.76
0.71
ALT (IU/L)
30
40
0.47
M 0.61–1.13, F 0.44–0.78 M 5–35, F 5–35
10
9
MDS
6
7
Pure red cell aplasia
1
0
Aplastic anemia
2
2
IMF
1
0
13
7
Baseline
Underlying disease
0.43
Bone marrow failure Data are expressed as median except disease DFX deferasirox, Hb hemoglobin, TIBC total ironbinding capacity, ALT alanine aminotransferase, MDS myelodysplastic syndrome, IMF idiopathic myelofibrosis, AML acute myeloid leukemia, ALL acute lymphoblastic leukemia, MPN myeloproliferative neoplasms, M male, F female
Hematological malignancy AML
10
4
ALL
1
1
Malignant lymphoma MDS/MPN
1
1
1
1
Discussion
Table 2 Regression analysis of response Univariate Unit of RBC transfusion during this study (units) Maximum dose of DFX (mg) Minimum transferrin saturation (%) Duration of taking DFX (weeks)
Multivariate
0.253 0.685 \0.001 0.109
Maximum elevation of TIBC (lg/dl)
\0.001
Maximum elevation of UIBC (lg/dl)
\0.001
Maximum elevation of iron (lg/dl)
0.021
DFX/RBC/day (mg/unit/day)
0.003
\0.001
RBC red blood cell, DFX deferasirox, TIBC total iron-binding capacity, UIBC unsaturated iron-binding capacity
Table 3). A TIBC elevation [150 lg/dl was observed at a median of 5 weeks earlier than the achievement of a [40 % serum ferritin reduction. Representative cases are shown, including 2 response cases that achieved TIBC elevations[150 lg/dl and 1 non-response case that did not achieve (Fig. 3).
123
Transferrin is an iron-binding transport glycoprotein with a molecular weight of 76.5 kDa and a biological half-life of 7–9 days. Low hepatocyte iron levels regulate the transcriptional level of transferrin [14, 15]. TIBC is calculated as the sum of UIBC and serum iron concentration. However, measurements of UIBC and serum iron are difficult during DFX therapy [16]. The measured levels of UIBC and serum iron have been reported to exceed the actual values during DFX therapy because of the presence of DFX in the serum [16]. UIBC is calculated by measuring the unbound iron after adding excessive iron to a serum sample. Iron-free DFX can bind this extrinsic iron and thus increase the measured UIBC value. Similarly, DFX-binding iron is undistinguishable from transferrin-bound iron and is measured as serum iron in an iron-quantitating assay [16]. Accordingly, the presence of DFX increased the apparent TIBC value, as it represented the sum of UIBC and serum iron concentration. Taken together, TIBC elevation could be a useful surrogate marker for the serum DFX concentration. Our clinical observation suggested that
TIBC elevation prior to DFX response
257
Fig. 1 ROC curve analysis for the association between TIBC elevation and a [40 % ferritin reduction. The optimum cut-off point was defined as the closest point on the curve to the point (X, Y) = (0, 1), where X denotes 1 - specificity and Y denotes sensitivity. The ROC curve reveals an optimal TIBC elevation cut-off point of
Fig. 2 Changes in the serum ferritin levels of patients who did or did not achieve a TIBC elevation of [150 lg/dl (left and right panels, respectively). Solid lines indicate patients who achieved a ferritin reduction [40 % above the baseline level during chelation therapy (responders). Dashed lines indicate the non-responders. All but 1 patient with a TIBC elevation of [150 lg/dl were responders. The excluded patient could not continue DFX therapy because of diarrhea
Table 3 Multivariate analysis for the DFX response by Fine–Gray proportional model HR
95 % CI
p value
Age (years)
1.014
0.99–1.04
0.331
Female
0.186
0.04–0.90
0.036
Body weight (kg)
1.062
0.96–1.18
0.253
Mean DFX dosage Elevation of TIBC [ 150 lg/dl
0.998 29.62
0.99–1.00 4.78–183.6
DFX deferasirox, TIBC total iron-binding capacity
0.104 \0.001
150 lg/dl (a). The cumulative response incidence revealed a statistically significant difference between the groups with and without TIBC elevations of [150 lg/dl (solid line and dashed line, respectively; b)
a TIBC elevation of [150 lg/dl indicates a sufficient serum DFX level for ferritin reduction. Another possibility is that DFX can induce transferrin production. Ikuta et al. [16] reported that the direct transferrin concentration measurement was not elevated in the context of DFX therapy. This finding indicated that the TIBC elevation observed during DFX therapy reflected the DFX concentration rather than the transferrin production. We think that an efficient approach to DFX therapy comprises the initial administration of a small dosage (e.g., 10 mg/kg/day), followed by a gradual increase with TIBC monitoring until the TIBC increases to [150 lg/dl above the baseline, as shown in Fig. 3a. In our analysis, the hazard ratio for effective ferritin reduction was significantly lower among DFX-treated women according to the Fine–Gray competing risks model. Women have a 17.5 % lower clearance with DFX relative to men [17]. Although we did not define the DFX dosage because of the retrospective study design, this difference in clearance observed among women might have an influence on the efficacy of DFX. Our study has several limitations. First, these data should be confirmed by prospective data because our study was a retrospective analysis. We analyzed various blood test time points and therefore needed regular intervals between DFX use and blood tests to assess the DFX pharmacokinetics and pharmacodynamics [18]. Second, TIBC elevation values differ among UIBC and iron assay systems because each system uses a different chromogen. Reportedly, the apparent increases in serum iron and TIBC depend on the assay system and particularly the chromogen [16]. Although we used the most widely used chromogen, a re-evaluation is necessary to determine the cut-off values for other chromogens. We are planning a prospective study to validate the association
123
258
J. Watanabe et al.
Fig. 3 The DFX dosage, RBC transfusion, TIBC elevation (gray column), 150 lg/dl of TIBC elevation (dashed line), and serum ferritin levels (black line) for the representative cases in this study. a A 76-year-old female with MDS 5q syndrome was treated with lenalidomide and regular RBC transfusion. We increased the DFX dosage gradually every 2 weeks and maintained the dosage after observing a TIBC elevation of [150 lg/dl. The patient achieved a ferritin reduction of [40 % from the baseline level after 8 weeks and achieved a TIBC elevation of [150 lg/dl. b A 79-year-old male with PRCA was treated with cyclosporine and regular RBC transfusion. He initiated treatment with 20 mg/kg of DFX and maintained this dosage.
He quickly exhibited a TIBC elevation of [500 lg/dl after 2 weeks but a slight serum ferritin reduction of only 6 % at that time. Within 8 weeks, he achieved sufficient ferritin reduction and became RBC transfusion-independent. c The third case was a non-responder. A 73-year-old male with relapsed AML was treated with azacitidine and regular RBC transfusion. We initiated DFX treatment at a 10 mg/kg dose. His serum ferritin level decreased slightly and concomitantly with a slight TIBC increase, but his TIBC did not increase of more than 150 lg/dl. His TIBC decreased when his condition worsened, and his serum ferritin level rebounded
between TIBC elevation and other factors such as transferrin, DFX concentration, non-transferrin bound iron, liver iron concentration, and serum ferritin reduction. In conclusion, our study revealed that TIBC elevation occurs before a favorable response to DFX therapy. TIBC monitoring during DFX therapy may provide indications for DFX dosage adjustments in individual patient and may help clinicians to more effectively treat patients with iron overload using DFX.
6. Vichinsky E, Onyekwere O, Porter J, Swerdlow P, Eckman J, Lane P, et al. A randomized comparison of deferasirox versus deferoxamine for the therapy of transfusional iron overload in sickle cell disease. Br J Haematol. 2006;136:501–8. 7. Gattermann N, Finelli C, Porta MD, Fenaux P, Ganser A, Guerci-Bresler A, et al. Deferasirox in iron-overloaded patients with transfusion-dependent myelodysplastic syndromes: Results from the large 1-year EPIC study. Leuk Res. 2010;34:1143–50. 8. Taher A, Cappellini MD, Vinchinsky E, Galanello R, Piga A, Lawniczek T, et al. Efficacy and safety of deferasirox dosage of [30 mg/kg per d in patients with transfusion-dependent anemia and iron overload. Br J Haematol. 2009;147:752–9. 9. Lee JW, Yoon SS, Shen ZX, Ganser A, Hsu HC, Habr D, et al. Iron chelating therapy with deferasirox in patients with aplastic anemia: a subgroup analysis of 116 patients from the EPIC trial. Blood. 2010;116:2448–54. 10. Porter J, Galanello R, Saglio G, Neufeld EJ, Vichinsky E, Cappellini MD, et al. Relative response of patients with myelodysplastic syndromes and other transfusion dependent anaemias to deferasirox (ICL670): a 1-yr prospective study. Eur J Haematol. 2007;80:168–76. 11. Porter JB, Elalfy MS, Taher AT, Aydinok Y, Chan LL, Lee SH, et al. Efficacy and safety of deferasirox at low and high iron burdens: results from the EPIC magnetic resonance imaging substudy. Ann Hematol. 2013;92:211–9. 12. Metzgeroth G, Dinter D, Schultheis B, Dorn-Beineke A, Lutz K, Leismann O, et al. Deferasirox in MDS patients with transfusion-caused iron overload-a phase-2 study. Ann Hematol. 2009;88:301–10. 13. Chirnomas D, Smith AL, Braunstein J, Finkelstein Y, Pereira L, Bergmann AK, et al. Deferasirox pharmacokinetics in patients with adequate versus inadequate response. Blood. 2009;114:4009–13.
Conflict of interest
The authors declare no conflicts of interest.
References 1. Kushner JP, Porter JP, Olivieri NF. Secondary iron overload. Hematol Am Soc Hematol Educ Program. 2001;1:47–61. 2. Kikuchi S, Kobune M, Iyama S, Sato T, Murase K, Kawano Y, et al. Prognostic significance of serum ferritin level at diagnosis in myelodysplastic syndrome. Int J Hematol. 2012;95:527–34. 3. Cohen AR. New advances in iron chelation therapy. Hematol Am Soc Hematol Educ Program. 2006;1:42–7. 4. Cappellini MD, Porter J, El-Beshlawy A, Li CK, Seymour JF, Elalfy M, et al. Tailoring iron chelation by iron intake and serum ferritin: the prospective EPIC study of deferasirox in 1744 patients with transfusion-dependent anemias. Haematologica. 2010;95:557–66. 5. Cappellini MD, Cohen A, Piga A, Bejaoui M, Perrotta S, Agaoglu L, et al. A phase 3 study of deferasirox (ICL670), a once-daily oral iron chelator, in patients with B-thalassemia. Blood. 2006;107:3455–62.
123
TIBC elevation prior to DFX response 14. Kohgo Y, Ikuta K, Ohtake T, Torimoto Y, Kato J. Body iron metabolism and pathophysiology pf iron overload. Int J Hematol. 2008;88:7–15. 15. Crichton RR, Charloteauc-Wauters M. Iron transport and storage. Eur J Biochem. 1987;164:485–506. 16. Ikuta K, Ito S, Tanaka H, Sasaki K, Torimoto Y, Kohgo Y. Interference of deferasirox with assays for serum iron and serum unsaturated iron binding capacity during iron chelating therapy. Clin Chim Acta. 2011;412:2261–6.
259 17. Exjade. Highlight of prescribing information. https://www. pharma.us.novartis.com/product/pi/pdf/exjade.pdf 18. Miyazawa K, Ohyashiki K, Urabe A, Hata T, Nakao S, Ozawa K, et al. A safety, pharmacokinetic and pharmacodynamics investigation of deferasirox (Exjade, ICL670) in patients with transfusion-dependent anemias and iron overload: a phase I study in Japan. Int J Hematol. 2008;88:73–81.
123
