Volume 15, Issue 2 , Pages 195-204, February 2009
Influence of Pretransplantation Serum Ferritin on Nonrelapse Mortality after Myeloablative and Nonmyeloablative Allogeneic Hematopoietic Stem Cell Transplantation
Article Outline
Abstract
Iron overload might be an important contributor to nonrelapse mortality (NRM) in hematopoietic stem cell transplantation (HSCT). We studied 264 patients undergoing allogeneic HSCT for hematologic malignancies between 1996 and 2006, using pretransplantation serum ferritin as a surrogate marker of iron overload. At 5 years, patients in the high ferritin group (≥599 ng/mL) had a lower overall survival (OS; 33.0% versus 63.5%; P < .001) and a higher NRM (34.9% versus 13.7%; P < .001) than those in the low ferritin group (<599 ng/mL). Multivariate analyses showed that high pretransplantation serum ferritin was a significant risk factor for worse survival (relative risk [RR] = 1.68; P = .05) and increased NRM (RR = 2.47; P = .01). There was no significant difference in the cumulative incidence of relapse, and acute and chronic graft-versus-host disease (aGVHD, cGVHD) between the 2 groups. Patients in the high ferritin group were more likely to die of infection (P < .010) and organ failure (P < .019). Similar results were observed after dividing the patients according to the intensity of conditioning regimens. These findings emphasize the prognostic impact of pretransplantation serum ferritin in HSCT recipients.
Key words: Iron overload, Nonrelapse mortality, Hematopoietic stem cell transplantation, Ferritin, Hematologic malignancy, Hematopoietic cell transplantation-specific comorbidity index
Introduction
Iron overload is common in patients undergoing allogeneic hematopoietic stem cell transplantation (HSCT) for hematologic disorders 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15. Several studies have reported the association between iron overload and transplant complications such as chronic liver disease 4, 5, 6, 7, 13, 14, 16, 17, 18, sinusoidal obstruction syndrome 13, 14, 19, infection 11, 12, 13, 14, 20, 21, 22, 23, 24, 25, and idiopathic pneumonia syndrome [13]. The adverse impact of iron overload on survival of patients undergoing allogeneic HSCT for thalassemia is well established [26]. However, very few studies have assessed the role of iron overload in the outcome of allogeneic HSCT for other hematologic disorders. Previous studies showed that an elevated pretransplantation serum ferritin level, a surrogate marker of iron overload, was associated with lower overall survival (OS) and increased nonrelapse mortality (NRM) in patients undergoing myeloablative allogeneic HSCT for hematologic malignancies 9, 10. In general, the main causes of NRM after allogeneic HSCT are infection and graft-versus-host disease (GVHD). However, little is known about the causes of NRM, and the incidence of acute and chronic GVHD (aGVHD, cGVHD) in these patients.
Nonmyeloablative HSCT represents an effective strategy to reduce the toxicity of transplantation, as was shown in patients with older age or organ dysfunction with hematologic malignancies [27]. However, it is not clear whether the adverse impact of iron overload is present in nonmyeloablative HSCT.
The aims of this study were: (1) to determine the impact of pretransplantation serum ferritin on transplant outcome in patients undergoing allogeneic HSCT for hematologic malignancies; (2) to identify the causes of NRM in patients with elevated ferritin values; (3) to clarify the incidence of aGVHD and cGVHD in patients with high ferritin values, and (4) to examine whether this association of pretransplantation serum ferritin and transplant outcome is present in nonmyeloablative HSCT.
Methods
Patients
We retrospectively studied 272 consecutive adult patients (≥16 years old) with hematologic malignancies who underwent allogeneic HSCT at the University of Tokyo Hospital, Japan, between June 1996 and October 2006. Two hundred sixty-four patients were included in the present study, as 8 patients were excluded because of the lack of data.
Myeloablative conditioning regimens included either total body irradiation (TBI)-based regimens or non-TBI-based regimens. TBI-based regimens mainly consisted of cyclophosphamide (Cy; 60 mg/kg/day for 2 days) plus fractionated TBI (12 Gy), whereas non-TBI-based regimens busulfan (Bu; 1 mg/kg every 6 hours for 4 days) plus Cy (60 mg/kg/day for 2 days). Patients were offered nonmyeloablative conditioning regimens as they were ineligible for conventional HSCT because of (1) age ≥56 years (n = 24); (2) presence of preexisting significant medical problem (severe cardiac dysfunction [n = 3], serious respiratory failure [n = 2], invasive pulmonary aspergillosis [n = 1]); and (3) history of high-dose HSCT (n = 10; 4 patients were also old, and 2 patients also had significant medical problems). Nonmyeloablative conditioning regimens consisted of fludarabine (Flu)-based regimens with or without low-dose TBI (4 Gy).
Prophylaxis for GVHD was performed with calcineurin inhibitors (cyclosporine [CsA] or tacrolimus) with or without short-term methotrexate (sMTX) in most patients. In vivo T cell depletion using alemtuzumab was performed in 30 patients, concomitant with CsA and sMTX.
Human leukocyte antigen (HLA)-matching for donor selection was based on serologic typing for HLA-A and -B antigens, and molecular typing for HLA-DRB1 antigen.
Disease morphology was determined according to the French-American-British classification. Standard-risk diseases were defined as acute leukemia in the first or second complete remission, chronic myelogenous leukemia (CML) in the first or second chronic phase, chemosensitive lymphoma, and myelodysplastic syndrome (MDS) in refractory anemia (RA) or refractory anemia with ringed sideroblasts (RARS). All the other conditions were classified as high-risk diseases.
Prophylaxis against bacterial infection was performed with tosufloxacin, fungal infection with fluconazole, herpes simplex virus (HSV) infection with acyclovir, and Pneumocystis jiroveci infection with sulfamethoxazole/trimethoprim.
Pretransplantation comorbidities were assessed retrospectively by comprehensive review of medical records and computer database system according to the hematopoietic cell transplantation-specific comorbidity index (HCT-CI) classification (44 missing data of pulmonary diffusing capacity for carbon monoxide and 3 missing data of whole pulmonary function tests) [28]. Each patient was assigned a comorbidity score and was stratified into low-risk (score 0), intermediate-risk (score 1-2), and high-risk (score ≥3) groups based on their total score.
Information concerning Karnofsky performance status was collected by a comprehensive review of medical records and assessed as per the Karnofsky scale. Karnofsky performance status scores were categorized as low (≤80%) or high (>80%) in accordance with previous reports [29].
Cause of death information was obtained from detailed review of the patients' medical charts and database that routinely record the primary cause of death as assigned by the treating physician at the time of patient death in a uniform manner.
This study was performed in accordance with the Helsinki Declaration and approved by the Ethics Committee of the University of Tokyo Hospital. All patients provided written informed consent.
Assessment of Iron Overload
We used pretransplantation serum ferritin as a surrogate marker of iron overload at the time of transplantation, because it is inexpensive, correlated with directly measured liver iron content [30], and reliable, with extensive clinical validation in monitoring iron status [31]. Serum ferritin was routinely measured as a part of routine pretransplant workup before the beginning of the conditioning regimen by using fluorescein enzyme immunoassay at our hospital. Because ferritin reflects acute inflammatory reaction besides iron overload, we included pretransplantation serum CRP and albumin, which were measured at the same time as serum ferritin and available in all of the studied patients, in the multivariate models for adjusting the influence of inflammation.
Statistical Methods
The duration of follow-up was the time from transplant to the last assessment for survivors. Variables related to the patients, the underlying diseases, the transplantation procedures, and the causes of NRM were compared between the groups with the Fisher's exact test for categoric variables and the Mann-Whitney U-test for continuous variables [32]. Probabilities of OS were calculated with the Kaplan-Meier method; the log-rank test was used for univariate comparisons [33]. For analyses of OS, death from any cause was considered an event, and data on patients who were alive at the last follow-up were censored. Probabilities of NRM, relapse, and aGVHD and cGVHD were calculated with the use of the cumulative-incidence-function method [34]. For NRM, relapse was the competing event; for relapse, NRM was the competing event; for GVHD, death without GVHD was the competing event. Data on patients who were alive without an event were censored at the last follow-up.
The association of pretransplantation serum ferritin and the transplant outcomes was evaluated in multivariate analyses, with the use of Cox proportional-hazards regression to adjust for OS and Fine and Gray's proportional-hazards model for subdistribution of a competing risk for NRM [35]. The variables considered in multivariate analyses were pretransplantation serum ferritin, CRP, and albumin, age and sex of recipients, type of hematologic malignancies, disease status at transplantation, interval between diagnosis and transplantation, type of grafts, HLA compatibility, conditioning regimen, prophylaxis against GVHD, year of transplantation, history of prior transplantation, donor-recipient sex mismatch, HCT-CI, and KPS. Pretransplantation serum ferritin was coded according to whether it exceeds the median values. Pretransplantation serum CRP and albumin were coded according to whether they were normal or not. The relative risk estimates with their 95% confidence intervals and respective P values were reported from these analyses. All P values are 2-sided, with a type I error rate fixed at 0.05. Statistical analyses were performed with R 2.6.1 software (The R Foundation for Statistical Computing, 2007).
Results
Patients
We reviewed the records of 264 adult patients who underwent allogeneic HSCT for hematologic malignancies at our institution. The median value of pretransplantation serum ferritin was 599 ng/mL (range: 5-8128 ng/mL). There was a strong relationship between pretransplantation serum ferritin and OS (Figure 1). The 5-year OS for patients with pretransplantation ferritin in the first quartile (<182 ng/mL) was 62.9% (95% confidence interval [CI], 51.6%-76.8%); in the second quartile (182-599 ng/mL), 64.2% (95% CI, 52.2%-79.0%); in the third quartile (599-1178.5 ng/mL), 40.3% (95% CI, 29.4%-55.4%); and in the fourth quartile (>1178.5 ng/mL), 23.7% (95% CI, 13.7%-40.9%). Compared with patients in the first quartile, those in the third and fourth quartile had a significantly inferior OS (P < .001, each), but those in the second quartile did not (P = .856). The patients, thus, were divided into 2 groups (ie, high ferritin group and low ferritin group) according to the median value of pretransplantation ferritin (ie, 599 ng/mL).
Figure 1
OS after allogeneic HSCT for hematologic malignancies in patients stratified by pretransplantation serum ferritin. *In the first quartile, ferritin level was <182 ng/mL, the second quartile, 182-599 ng/mL, the third quartile, 599-1178.5 ng/mL, and the fourth quartile, >1178.5 ng/mL. Compared with patients in the first quartile, those in the third and fourth quartile had significantly inferior survival (P < .001, each), but those in the second quartile did not (P = .856).
Their clinical characteristics are shown in Table 1. One hundred two patients received grafts from HLA-matched related donors, 77 patients from matched unrelated donors, and 85 patients from mismatched related (n = 45) or unrelated (n = 40) donors. Compared with patients in the low ferritin group, patients in the high ferritin group were more likely to be male, have acute myelogenous leukemia (AML) and MDS, have high-risk diseases at transplantation, have histories of prior transplantations, have high HCT-CI scores, and have low KPS. Patients in the high ferritin group were less likely to have CML and non-Hodgkin lymphoma (NHL), and have received CsA and sMTX as GVHD prophylaxis than those in the low ferritin group. Median CRP and albumin were 0.30 mg/dL (range: 0.02-17.20 mg/dL) and 4.0 g/dL (range: 2.3-5.2 g/dL), respectively. The percentage of patients with elevated serum CRP (>0.30 mg/dL) and lower albumin (<3.9 g/dL) in the high ferritin group was significantly higher than those in the low ferritin group.
Table 1. Clinical Characteristics of 264 Patients Undergoing Allogeneic Hematopoietic Stem Cell Transplantation for Hematologic Malignancies According to Pretransplantation Serum Ferritin
| Variable | Total (N = 264) | Low ferritin < 599 ng/mL (N = 132) | High ferritin ≥599 ng/mL (N = 132) | P Value∗ |
|---|---|---|---|---|
| Age, years | .082 | |||
 Median | 40 | 38 | 41 | |
| Range | 16-66 | 16-62 | 18-66 | |
| Sex | .003 | |||
| Male | 170 (64) | 73 (55) | 97 (73) | |
| Female | 94 (36) | 59 (45) | 35 (27) | |
| Disease classification | <.001 | |||
| Acute myelogenous leukemia | 78 (30) | 16 (12) | 62 (47) | |
| Acute lymphoblactic leukemia | 62 (23) | 31 (23) | 31 (23) | |
| Chronic myelogenous leukemia | 51 (19) | 45 (34) | 6 (5) | |
| Myelodysplastic syndrome | 34 (13) | 14 (11) | 20 (15) | |
| RA/RARS | 9 (3) | 2 (2) | 7 (5) | |
| RAEB/RAEBT | 21 (8) | 12 (9) | 9 (7) | |
| CoL | 4 (2) | 0 (0) | 4 (3) | |
| Non-Hodgkin lymphoma | 33 (13) | 24 (18) | 9 (7) | |
| Others | 6 (2) | 2 (2) | 4 (3) | |
| Disease risk at transplantation† | <.001 | |||
| Standard | 146 (55) | 88 (67) | 58 (44) | |
| High | 118 (45) | 44 (33) | 74 (56) | |
| Interval between diagnosis and transplantation, months | 1.000 | |||
| Median | 10.6 | 10.9 | 10.1 | |
| Range | 1.8-202.3 | 2.4-173.4 | 1.8-202.3 | |
| Graft source | .510 | |||
| Bone marrow | 168 (64) | 89 (67) | 79 (60) | |
| Peripheral blood | 85 (32) | 39 (30) | 46 (35) | |
| Others | 11 (4) | 4 (3) | 7 (5) | |
| Donor type | .348 | |||
| Matched related donor | 102 (39) | 53 (40) | 49 (37) | |
| Matched unrelated donor | 77 (29) | 42 (32) | 35 (27) | |
| Mismatched donor | 85 (32) | 37 (28) | 48 (36) | |
| Conditioning regimen | .099 | |||
| Myeloablative | 230 (87) | 121 (92) | 109 (82) | |
| TBI-based regimen | 190 (72) | 97 (73) | 93 (70) | |
| Non-TBI-based regimen | 40 (15) | 24 (18) | 16 (12) | |
| Nonmyeloablative | 34 (13) | 11 (8) | 23 (18) | |
| TBI-based regimen | 12 (5) | 3 (2) | 9 (7) | |
| Non-TBI-based regimen | 22 (8) | 8 (6) | 14 (11) | |
| GVHD prophylaxis | .028 | |||
| Cyclosporine alone | 11 (4) | 2 (2) | 9 (7) | |
| Cyclosporine and methotrexate | 201 (76) | 110 (83) | 91 (69) | |
| Tacrolimus and methotrexate | 24 (9) | 9 (7) | 15 (11) | |
| Alemzutumab | 28 (11) | 11 (8) | 17 (13) | |
| Transplant year | .122 | |||
| 1995-1998 | 43 (16) | 25 (19) | 18 (14) | |
| 1999-2002 | 113 (43) | 61 (46) | 52 (39) | |
| 2003-2006 | 108 (41) | 46 (35) | 62 (47) | |
| Prior transplantation | .005 | |||
| 243 (92) | 128 (97) | 115 (87) | ||
| >1 | 21 (8) | 4 (3) | 17 (13) | |
| Sex pair (Donor-Recipient) | .396 | |||
| Female-male | 67 (25) | 30 (23) | 37 (28) | |
| Others | 197 (75) | 102 (77) | 95 (72) | |
| Hematopoietic cell transplantation specific comorbidity index | <.001 | |||
| Low (0) | 84 (32) | 54 (41) | 30 (23) | |
| Intermediate (1-2) | 96 (36) | 51 (39) | 45 (34) | |
| High (≥) | 84 (32) | 27 (20) | 57 (43) | |
| Karnofsky performance status | .001 | |||
| >80 | 168 (64) | 97 (73) | 71 (54) | |
| ≤80 | 96 (36) | 35 (27) | 61 (46) | |
| C-reactive protein | <.001 | |||
| 03 | 170 (64) | 105 (80) | 65 (49) | |
| >0.3 | 94 (36) | 27 (20) | 67 (51) | |
| Albumin | <.001 | |||
| ≥3.9 | 178 (67) | 104 (79) | 74 (56) | |
| <3.9 | 86 (33) | 28 (21) | 58 (44) | |
∗The Fisher's exact test was used for categoric variables and the Mann-Whitney U-test for continuous variables. |
†Standard-risk diseases were defined as acute leukemia in the first or second complete remission, chronic myelogenous leukemiin the first or second chronic phase, chemosensitive lymphoma, and myelodysplastic syndrome in refractory anemia or refractory anemia with ringed sideroblasts. All the other conditions were classified as high-risk diseases. |
OS, NRM, and Relapse
Median follow-up period for survivors after HSCT was 48.9 months (range: 3.0-136.7 months). At 5 years, patients in the high ferritin group had a significantly inferior OS than those in the low ferritin group (33.0% versus 63.5%; P < .001) (Figure 2A). Multivariate analyses of risk factors among all patients showed that high pretransplantation serum ferritin (relative risk [RR] = 1.68; 95% CI, 1.01-2.78; P = .05), high-risk disease (RR = 3.25; 95% CI, 1.96-5.39; P < .01), the use of alemtuzumab (RR = 3.41; 95% CI, 1.49-7.82; P ≤ .01), and high-risk HCT-CI (RR = 1.75; 95% CI, 1.02-2.99; P = .04) significantly predicted worse OS (Table 2).
Figure 2
OS (A), NRM (B), and relapse (C) rates after allogeneic HSCT for hematologic malignancies in patients divided by the median value of the pretransplantation serum ferritin level. Patients in the high ferritin group had significantly lower OS and higher NRM than those in the low ferritin group (P < .001, each). The relapse rate was similar between the 2 groups (P = .137).
Table 2. Factors Associated with Overall Survival and Nonrelapse Mortality within 264 Patients Undergoing Allogeneic Hematopoietic Stem Cell Transplantation for Hematologic Malignancies: Multivariate Analysis
| Overall Survival | Nonrelapse Mortality | |||
|---|---|---|---|---|
| Variable | RR (95%CI) | P Value∗ | RR (95%CI) | P Value∗ |
| Ferritin | ||||
| Low (<599 ng/mL) | 1.00 | 1.00 | ||
| High (≥599 ng/mL) | 1.68 (1.01-2.78) | .05 | 2.47 (1.21-5.07) | .01 |
| Age | ||||
| <50 years | 1.00 | 1.00 | ||
| ≥50 years | 0.97 (0.55-1.69) | .90 | 1.35 (0.56-3.23) | .51 |
| Sex | ||||
| Male | 1.00 | 1.00 | ||
| Female | 1.30 (0.78-2.14) | .31 | 1.30 (0.54-3.10) | .56 |
| Disease classification | ||||
| Acute myelogenous leukemia | 1.00 | 1.00 | ||
| Acute lymphoblastic leukemia | 1.36 (0.78-2.36) | .28 | 0.90 (0.38-2.18) | .82 |
| Chronic myelogenous leukemia | 0.90 (0.46-1.78) | .77 | 1.84 (0.70-4.84) | .22 |
| Myelodysplastic syndrome | 0.56 (0.29-1.06) | .07 | 1.97 (0.77-5.01) | .15 |
| Non-Hodgkin lymphoma | 1.28 (0.62-2.62) | .50 | 2.22 (0.91-5.42) | .08 |
| Others | 1.35 (0.34-5.39) | .67 | 1.45 (0.11-19.04) | .78 |
| Disease risk at transplantation | ||||
| Standard | 1 00 | 1 00 | ||
| High | 3.25 (1.96-5.39) | <.01 | 1.22 (0.61-2.44) | .58 |
| Interval between diagnosis and transplantation | ||||
| ≤10.6 months | 1.00 | 1.00 | ||
| >10.6 months | 1.28 (0.83-1.98) | .27 | 1.47 (0.75-2.80) | .26 |
| Graft source | ||||
| Bone marrow | 1.00 | 1.00 | ||
| Peripheral blood | 0.97 (0.51-1.84) | .93 | 0.52 (0.18-1.50) | .23 |
| Others | 1.73 (0.71-4.23) | .23 | 1.88 (0.79-4.46) | .15 |
| Donor type | ||||
| Matched related donor | 1.00 | 1.00 | ||
| Alternative donor | 0.65 (0.37-1.14) | .13 | 0.46 (0.20-1.05) | .06 |
| Conditioning regimen | ||||
| Myeloablative | 1.00 | 1.00 | ||
| Nonmyeloablative | 1.13 (0.52-2.46) | .75 | 0.90 (0.29-2.82) | .86 |
| Total body irradiation | ||||
| Non-TBI-based regimen | 1.00 | 1.00 | ||
| TBI-based regimen | 1.06 (0.60-1.85) | .84 | 1.68 (0.70-4.04) | .25 |
| GVHD prophylaxis | ||||
| Cyclosporine-based | 1.00 | 1.00 | ||
| Tacrolimus-based | 1.02 (0.51-2.02) | .96 | 1.21 (0.37-3.92) | .75 |
| Alemzutumab | 3.41 (1.49-7.82) | <.01 | 4.70 (1.19-18.58) | .03 |
| Transplant year | ||||
| 1995-1998 | 1.00 | 1.00 | ||
| 1999-2002 | 1.20 (0.66-2.19) | .55 | 1.10 (0.45-2.67) | .84 |
| 2003-2006 | 1.10 (0.54-2.22) | .80 | 0.76 (0.28-2.09) | .59 |
| Prior transplantation | ||||
| 0 | 1.00 | 1.00 | ||
| ≥1 | 1.61 (0.85-3.06) | .15 | 1.62 (0.53-4.96) | .40 |
| Sex pair (Donor-Recipient) | ||||
| Female-male | 1.00 | 1.00 | ||
| Others | 1.03 (0.65-1.63) | .91 | 0.96 (0.47-1.97) | .92 |
| Hematopoietic cell transplantation specific comorbidity index | ||||
| Low (0) | 1.00 | 1.00 | ||
| Intermediate (1-2) | 1.71 (1.00-2.91) | .05 | 4.04 (1.57-10.37) | <.01 |
| High (≥3) | 1.75 (1.02-2.99) | .04 | 4.34 (1.64-11.52) | <.01 |
| Karnofsky performance status | ||||
| >80 | 1.00 | 1.00 | ||
| ≤80 | 0.89 (0.56-1.40) | .60 | 0.86 (0.43-1.71) | .66 |
| C-reactive protein | ||||
| ≤0.3 | 1.00 | 1.00 | ||
| >0.3 | 1.45 (0.94-2.23) | .10 | 1.88 (0.90-3.90) | .09 |
| Albumin | ||||
| ≥3.9 | 1.00 | 1.00 | ||
| <3.9 | 1.43 (0.90-2.28) | .13 | 1.36 (0.71-2.61) | .35 |
∗Cox proportional-hazards regression model was used for overall survival, and Fine and Gray's proportional-hazards model for subdistribution of a competing risk was used for continuous nonrelapse mortality. |
At 5 years, patients in the high ferritin group had a significantly higher NRM than those in the low ferritin group (34.9% versus 13.7%; P < .001) (Figure 2B). There was no statistical difference in relapse rate between the 2 groups of patients (34.4% versus 28.3%; P = .137) (Figure 2C). In multivariate analyses, high pretransplantation serum ferritin (RR = 2.47; 95% CI, 1.21-5.07; P = .01), the use of alemtuzumab (RR = 4.70; 95% CI, 1.19-18.58; P = .03), and intermediate-risk (RR = 4.04; 95% CI, 1.57-10.37; P < .01) and high-risk HCT-CI (RR = 4.34; 95% CI, 1.64-11.52; P < .01) were the significant risk factors associated with increased NRM (Table 2).
aGVHD and cGVHD
The cumulative incidence of grade II, III, and IV aGVHD was similar between the 2 groups of patients (37.1% versus 42.4% for the patients in the high and low ferritin group; P = .422). There was no statistical difference in the incidence of extensive cGVHD between the 2 groups of patients (43.9% versus 40.4%; P = .703).
Causes of NRM
A total of 58 patients died from nonrelapse causes, 16 of 132 patients in the low ferritin group (12%) and 42 of 132 patients in the high ferritin group (32%). Table 3 lists the causes of NRM according to pretransplantation serum ferritin. Patients in the high ferritin group were more likely to die of infection (P = .010) and organ failure (P = .019), compared with those in the low ferritin group.
Table 3. Causes of Nonrelapse Death in 264 Patients Undergoing Allogeneic Hematopoietic Stem Cell Transplantation for Hematologic Malignancies According to Pretransplantation Serum Ferritin
| Variable | Total (N = 264) | Low ferritin <599 ng/mL (N = 132) | High ferritin ≥599 ng/mL (N = 132) | P Value† |
|---|---|---|---|---|
| No. of Patients (%)∗ | ||||
| Nonrelapse mortality | 58 (22) | 16 (12) | 42 (32) | |
| GVHD | 14 (5) | 5 (4) | 9 (7) | .411 |
| Infection‡ | 25 (9) | 6 (5) | 19 (14) | .010 |
| Hemorrhage | 2 (1) | 1 (1) | 1 (1) | 1.000 |
| Interstitial pneumonia | 5 (2) | 2 (2) | 3 (2) | .663 |
| Organ failure§ | 10 (4) | 1 (1) | 9 (7) | .019 |
| Secondary malignancy | 1 (0) | 1 (1) | 0 (0) | 1.000 |
| Unknown | 1 (0) | 0 (0) | 1 (1) | 1.000 |
∗Percentages of patients in each group are shown in the parentheses. Because of rounding, the percentages of respective nonrelapse causes of death may not sum to the percentage of overall nonrelapse mortality. |
†The Fisher's exact test was used. |
‡A comparable proportion of patients who died of infect, had grade II, III, and IV acute GVHD/or extensive chronic GVHD; 5 out of 6 in the low ferritin group and 11 out of 19 in the high ferritin group. Infections in the low ferritin group were bacterial (n = 1), fungal (n = 3), and viral (n = 2). Infections in the high ferritin group were bacterial (n = 10), fungal (n = 6), and viral (n = 3). |
§Organ failure was assigned when the specific cause of organ failure, including infection, GVHD, hemorrhage, or veno-occlusive disease, is not determined despite the appropriate wsorerep. Multiorgan failure was assigned when two or more organ failure was attributable to death. Organ failure in the low ferritin group included heart failure (n = 1). Organ failure in the high ferritin group included heart failure (n = 2) renal failure (n = 3), hepatic failure (n = 2), and multiorgan failure (n = 2). |
Outcomes after Nonmyeloablative Compared with Myeloablative Conditioning
As the decreased OS was attributable to the increased NRM in patients with elevated ferritin values who underwent allogeneic HSCT for hematologic malignancies, we aimed to investigate whether this relationship is still present in nonmyeloablative HSCT. After myeloablative HSCT, patients in the high ferritin group had an inferior OS (38.5% versus 62.6%; P < .001), and a higher NRM (30.8% versus 13.8%; P < .001) at 5 years, compared with those in the low ferritin group. However, there was no significant difference in relapse rate (33.2% and 28.8%; P = .302) between the 2 groups. Similarly, after nonmyeloablative HSCT, patients in the high ferritin group had an inferior OS (7.0% versus 80.8%; P = .002) (Figure 3A), and a higher NRM (54.3% versus 9.1%; P < .001) (Figure 3B) at 5 years, compared with those in the low ferritin group. However, there was no significant difference in relapse rates (39.1% versus 22.7%; P = .137) (Figure 3C) between the 2 groups. Thus, patients with elevated ferritin values had decreased OS because of increased NRM after allogeneic HSCT, regardless of the intensity of conditioning regimens.
Figure 3
OS (A), NRM (B), and relapse (C) rates after allogeneic HSCT following nonmyeloablative conditioning in patients divided by the median value of the pretransplantation serum ferritin level. Patients in the high ferritin group had significantly lower OS (P = .002) and higher NRM (P < .001) than those in the low ferritin group. The relapse rate was similar between the 2 groups (P = .137).
Discussion
The current analysis extended previous reports on the role of pretransplantation serum ferritin in allogeneic HSCT. In this study, we showed that an elevated pretransplantation serum ferritin level was associated with decreased OS and increased NRM in patients undergoing allogeneic HSCT for hematologic malignancies, consistent with previous observations 9, 10. Furthermore, to our knowledge, this is the first report to clarify the causes of NRM and the incidence of aGVHD and cGVHD in patients with high ferritin values. Moreover, we showed that pretransplantation serum ferritin affected adversely the transplant outcome as well in patients who received nonmyeloablative conditioning regimens.
Decreased OS in patients with high ferritin values was attributable to increased NRM. Among the various causes that might explain NRM, infection and organ failure were significantly more frequent in patients in the high ferritin group than those in the low ferritin group. These findings are consistent with the previous reports on the association between iron overload and transplant related complications, such as infection 11, 12, 13, 14, 20, 21, 22, 23, 24, 25 and chronic liver disease 4, 5, 6, 7, 13, 14, 16, 17, 18 after HSCT. One explanation for these results is that free radical iron, in a state of iron excess, can lead to a pro-oxidant state by generating free radicals, which could subsequently cause tissue injury and increase the risk of transplant related complications [14]. Indeed, a previous study showed the presence of catalytic iron that increased oxidative damage to proteins in the serum of children with acute lymphoblastic leukemia (ALL) receiving high-dose MTX [36]. Furthermore, iron is an essential cofactor for the growth of a number of opportunistic bacteria and fungi and free iron can also increase the susceptibility of patients to infections through impairment of cellular immunity and inhibition of chemotaxis and phagocytosis [14].
We used the median value of pretransplantation serum ferritin level (ie, 599 ng/mL) as the cutoff value for dividing the patients into the 2 groups. Guidelines on iron-chelation therapy for heavily transfused patients such as thalassemia, aplastic anemia (AA), and MDS recommend that the patients with serum ferritin levels higher than 1000-2500 ng/mL might benefit from the treatment for iron overload 37, 38, 39. We showed that patients with moderately elevated serum ferritin levels (599-1178.5 ng/mL) had inferior survival, compared with patients with normal (5-182 ng/mL) and mildly elevated serum ferritin levels (182-599 ng/mL) (Figure 1). This suggests that pretransplantation serum ferritin has adverse effects at lower level for HSCT recipients than transfused patients.
In patients undergoing allogeneic HSCT, the main cause of NRM in patients with high ferritin values was infection. This cause of death is quite different from that in chronically transfused patients such as thalassemia, in which iron overload associated cardiomyopathy and liver fibrosis were among the leading causes of death before the introduction of iron-chelation therapy [31]. Possible explanation for this difference is that patients undergoing allogeneic HSCT for hematologic malignancies were more susceptible to infections than those receiving chronic transfusions because of prolonged neutropenia and breaks in the mucocutaneous barrier because of the HSCT preparative regimens, cell-mediated and humoral immunity defects, and impaired functioning of the reticuloendothelial system [40]. Considering that the incidence of severe aGVHD and cGVHD in the high ferritin group was comparable with those in the low ferritin group, elevated ferritin value might be a risk factor for infection, independent of aGVHD and cGVHD. Thus, these results indicated the significance of the prophylaxis and surveillance regimen for infectious diseases to prevent iron overload related complications in patients with high ferritin values undergoing allogeneic HSCT.
Allogeneic HSCT following nonmyeloablative conditioning has been performed for hemoglobinopathies and reduced the transplant related complications in heavily transfused patients [41]. We, however, showed that the association of pretransplantation serum ferritin and NRM was present as well in patients undergoing nonmyeloablative HSCT. As patients with elevated ferritin value had decreased OS and increased NRM after allogeneic HSCT, regardless of the intensity of conditioning regimens, intervention for iron overload, including phlebotomy and iron-chelation therapy, in the pre- and posttransplantation setting may be important to reduce the morbidity and mortality in these patients, as well as thalassemia patients 42, 43, and its feasibility and safety need to be investigated in further study.
This study had a number of limitations. (1) This study was a retrospective, single-institution study. In particular, it is difficult to assign, retrospectively, the causes of death in the posttransplantation setting. Additional prospective multicenter studies are required to confirm the true association between iron overload and transplant outcomes. (2) Ferritin is an acute phase reactant and is not the best indicator of iron overload. However, ferritin is a useful noninvasive surrogate, because it is inexpensive, widely available, and reliable with extensive clinical validation in monitoring iron status [31]. Furthermore, ferritin has a strong association with hepatic iron concentration directly measured by liver biopsy [30]. Moreover, several guidelines on iron overload recommend serum ferritin as a diagnostic tool of iron overload 37, 38, 39. Although the influence of inflammation could be adjusted by the inclusion of CRP and/or albumin in the multivariate analyses, further studies using more reliable methods for assessment of body iron stores such as liver biopsy and magnetic resonance imaging would be helpful 15, 18. (3) The patients included in this study were a heterogeneous population who had various backgrounds and diseases, and underwent various transplant procedures, although these factors were adjusted in the multivariate analyses. (4) As the number of patients who received nonmyeloablative conditioning was limited, we could not perform a multivariate analysis of the nonmyeloablative subgroup alone. Further studies with larger number of nonmyeloablative transplants would be necessary to confirm these results.
In conclusion, our findings indicate that patients with an elevated pretransplantation serum ferritin level had inferior survival because of increased NRM, mainly from infection and organ failure, in allogeneic HSCT. Moreover, this association was observed as well in patients who received nonmyeloablative HSCT. These results emphasize the potential value of the treatment of iron overload in the pre- and posttransplantation setting. Different strategy from that for chronically transfused patients might be needed for pretransplant patients, because early intervention for iron overload and judicious prophylaxis and surveillance regimens for infections may be more important in allogeneic HSCT. Further study to confirm these findings would be helpful.
Acknowledgments
Financial disclosure: The authors have nothing to disclose.
References
- Sequential changes in serum iron and ferritin in patients undergoing high-dose chemotherapy and radiation with autologous bone marrow transplantation: possible implications for treatment related toxicity. Free Radic Biol Med. 1995;18:383–389
- Nontransferrin-bound iron in serum of patients receiving bone marrow transplants. Free Radic Biol Med. 1997;22:1159–1163
- . Non-transferrin-bound iron during allogeneic stem cell transplantation. Br J Haematol. 2001;113:836–838
- Iron overload and liver dysfunction after allogeneic or autologous bone marrow transplantation. Bone Marrow Transplant. 1996;17:63–66
- . Iron overload in bone marrow transplant recipients. Bone Marrow Transplant. 1998;22:167–173
- . Role of non-transferrin bound iron in iron overload and liver dysfunction in long term survivors of acute leukaemia and bone marrow transplantation. J Clin Pathol. 1996;49:853–856
- Liver dysfunction in allogeneic bone marrow transplantation recipients. Transplantation. 1996;62:56–61
- . Autografting as a risk factor for persisting iron overload in long-term survivors of acute myeloid leukaemia. Bone Marrow Transplant. 2003;32:909–913
- Iron overload might increase transplant-related mortality in haematopoietic stem cell transplantation. Bone Marrow Transplant. 2002;29:987–989
- Prognostic impact of elevated pretransplantation serum ferritin in patients undergoing myeloablative stem cell transplantation. Blood. 2007;109:4586–4588
- Frequent severe liver iron overload after stem cell transplantation and its possible association with invasive aspergillosis. Bone Marrow Transplant. 2004;34:505–509
- Early clinical impact of iron overload in stem cell transplantation. A prospective study. Ann Hematol. 2007;86:443–447
- . Rust and corrosion in hematopoietic stem cell transplantation: the problem of iron and oxidative stress. Bone Marrow Transplant. 2004;34:561–571
- . Iron overload in hematopoietic cell transplantation. Bone Marrow Transplant. 2008;41:997–1003
- Quantification by magnetic resonance imaging and liver consequences of post-transfusional iron overload alone in long term survivors after allogeneic hematopoietic stem cell transplantation (HSCT). Haematologica. 2007;92:850–853
- . Iron overload manifesting as apparent exacerbation of hepatic graft-versus-host disease after allogeneic hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2006;12:506–510
- Long-term liver dysfunction after allogeneic bone marrow transplantation: clinical features and course in 61 patients. Bone Marrow Transplant. 2000;26:649–655
- . Abnormal liver function tests following bone marrow transplantation: aetiology and role of liver biopsy. Eur J Gastroenterol Hepatol. 2004;16:157–162
- BMT: serum ferritin as risk factor for veno-occlusive disease of the liver. Prospective cohort study. Hematology. 2000;4:505–512
- Mucormycosis in allogeneic bone marrow transplant recipients: report of five cases and review of the role of iron overload in the pathogenesis. Bone Marrow Transplant. 1999;24:307–312
- . Listeria septicemia complicating bone marrow transplantation for Diamond-Blackfan syndrome. Pediatr Hematol Oncol. 1995;12:295–299
- . Apotransferrin administration prevents growth of Staphylococcus epidermidis in serum of stem cell transplant patients by binding of free iron. FEMS Immunol Med Microbiol. 2003;37:45–51
- Iron overload is a major risk factor for severe infection after autologous stem cell transplantation: a study of 367 myeloma patients. Bone Marrow Transplant. 2006;37:857–864
- Increased bone marrow iron stores is an independent risk factor for invasive aspergillosis in patients with high-risk hematologic malignancies and recipients of allogeneic hematopoietic stem cell transplantation. Cancer. 2007;110:1303–1306
- Mucormycosis after bone marrow transplantation: report of four cases in thalassemia and review of the literature. Bone Marrow Transplant. 1996;17:409–414
- Bone marrow transplantation in patients with thalassemia. N Engl J Med. 1990;322:417–421
- Engraftment of allogeneic hematopoietic progenitor cells with purine analog-containing chemotherapy: harnessing graft-versus-leukemia without myeloablative therapy. Blood. 1997;89:4531–4536
- Hematopoietic cell transplantation (HCT)-specific comorbidity index: a new tool for risk assessment before allogeneic HCT. Blood. 2005;106:2912–2919
- Hematopoietic cell transplantation-comorbidity index and Karnofsky performance status are independent predictors of morbidity and mortality after allogeneic nonmyeloablative hematopoietic cell transplantation. Cancer. 2008;112:1992–2001
- Long-term safety and effectiveness of iron-chelation therapy with deferiprone for thalassemia major. N Engl J Med. 1998;339:417–423
- Survival in medically treated patients with homozygous beta-thalassemia. N Engl J Med. 1994;331:574–578
- . Statistical Methods in Medical Research. 4th ed.. Malden, MA: Blackwell Science; 2002;
- . Survival Analysis: Techniques for Censored and Truncated Data. 2nd ed.. New York: Springer; 2003;
- . Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18:695–706
- . A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509
- Presence of iron catalytic for free radical reactions in patients undergoing chemotherapy: implications for therapeutic management. Cancer Lett. 1995;94:219–226
- Myelodysplastic syndromes clinical practice guidelines in oncology. J Natl Compr Cancer Netw. 2006;4:58–77
- Italian Society of Hematology practice guidelines for the management of iron overload in thalassemia major and related disorders. Haematologica. 2008;93:741–752
- Guidelines for the diagnosis and management of acquired aplastic anaemia. Br J Haematol. 2003;123:782–801
- . Summary of the guidelines for preventing opportunistic infections among hematopoietic stem cell transplant recipients. Clin Infect Dis. 2001;33:139–144
- Results of minimally toxic nonmyeloablative transplantation in patients with sickle cell anemia and beta-thalassemia. Biol Blood Marrow Transplant. 2003;9:519–528
- Intravenous chelation therapy during transplantation for thalassemia. Haematologica. 1995;80:300–304
- Phlebotomy to reduce iron overload in patients cured of thalassemia by bone marrow transplantation. Italian Cooperative Group for Phlebotomy Treatment of Transplanted Thalassemia Patients. Blood. 1997;90:994–998
Financial disclosure: See Acknowledgments on page 203.
PII: S1083-8791(08)00506-5
doi:10.1016/j.bbmt.2008.11.012
© 2009 American Society for Blood and Marrow Transplantation. Published by Elsevier Inc. All rights reserved.
Volume 15, Issue 2 , Pages 195-204, February 2009
