|Home | About | Journals | Submit | Contact Us | Français|
We retrospectively compared clinical outcomes in 1593 T-repleted URD marrow transplant recipients with AML, MDS and CML who received myeloablative conditioning regimens of either busulfan and cyclophosphamide (BuCy), standard-dose Cy/TBI (1,000-1,260 cGy) or high-dose Cy/TBI (1,320-1,500 cGy). Subjects were drawn from patients transplanted between 1991 and 1999 facilitated by the National Marrow Donor Program (NMDP). Patients who received high-dose Cy/TBI regimens were slightly younger, more likely to receive a mismatched transplant and to have intermediate or advanced disease compared to patients in the BuCy or standard-dose TBI group. Neutrophil recovery was significantly higher in the standard dose CY/TBI group compared to the high-dose Cy/TBI or BuCy group. Patients who received the high-dose Cy/TBI regimen had an increased risk of developing grade III-IV aGVHD when compared to the control group who received BuCy (p=0.011). Overall survival (OS), disease free survival (DFS), transplant-related mortality (TRM) and relapse were not significantly different between any of the regimens. We conclude that BuCy, standard-dose and high dose Cy/TBI regimens have equivalent efficacy profiles for OS, DFS, TRM and relapse risk in patients undergoing T-replete URD marrow transplantation for AML, CML and MDS.
The traditional preparative regimens used in unrelated donor hematopoietic stem cell transplantation (HSCT) are myeloablative. The justifications for high-intensity regimens include advanced disease in many patients and the higher probability of graft failure compared to patients receiving related donor HSCT.(1) Cyclophosphamide combined with total body irradiation (Cy/TBI) is the most common myeloablative regimen used in unrelated donor (URD) transplantation. The combination of the two agents provides the most effective immunosuppression allowing both engraftment and tolerance required for URD while also maximizing anti-tumor activity. In the early 1980's, busulfan was introduced in combination with cyclophosphamide (BuCy) as an alternative to Cy/TBI for myeloablation in sibling donor transplantation. The goals of this new preparative regimen were to reduce toxicity, improve outcome and provide an alternative to patients receiving prior radiation who would not be suitable candidates for TBI.(2, 3) Controversy exists as to the relative merits of BuCy and Cy/TBI regimens. Several randomized studies, as well as retrospective registry data in sibling donor transplants comparing the two types of preparative regimens report conflicting results concerning outcomes and toxicities.(4-7) Although TBI-based regimens are widely used in URD transplantation, emerging data from several single institution trials report excellent outcome of busulfan-based regimens in leukemia patients.(3, 8, 9) However, there are no reports comparing outcomes between TBI-based and the busulfan-based regimens in URD transplants. The objective of this report is to compare the clinical outcomes and toxicities of patients who received Cy/TBI regimens with those receiving BuCy regimens for URD transplantation in patients with acute myeloid leukemia, chronic myeloid leukemia and myelodysplasia.
The CIBMTR is a working group of more than 500 transplant centers worldwide that voluntarily contribute data on allogeneic and autologous transplants. Detailed demographic, disease and transplant characteristics and outcome data are collected on a sample of registered patients including all unrelated donor transplants facilitated by the NMDP in the U.S. Observational studies conducted by the CIBMTR are done with a waiver of informed consent and in compliance with HIPAA regulations as determined by the Institutional Review Board and the Privacy Officer of the Medical College of Wisconsin.
The study population included unrelated bone marrow transplant patients reported to the NMDP between 1991 and 1999, who met the following criteria: a) recipients of a T-cell replete bone marrow transplant; b) patients with the diagnosis of AML, CML and MDS; c) patients who received a BuCy or Cy/TBI preparative regimen with TBI dose ranging from 10 to 15Gy. Patients were excluded from this study if a) they received reduced-intensity preparative regimens; or b) TBI doses in preparative regimens that were lower than 10Gy or over 15Gy; or c) they received a T-depleted marrow graft or ATG was given for in vivo T-cell depletion. There were 2,089 patients selected according to the above criteria. Surviving recipients were then retrospectively contacted and provided informed consent for participation in the NMDP research program. The NMDP institutional review board waived consent for patients who died before soliciting consent. To address the potential bias introduced by the exclusion of non consenting surviving patients, a corrective action plan (CAP) modeling process which randomly excluded the same percentage of deceased patients using a weighted randomized scheme was used to adjust for over sampling of dead patients in the consented cohort.(10) After the CAP model was applied, a total of 1,804 (86.4%) patients remained. We further excluded 211 patients from the study because of missing key variables that could not be retrieved from the database or from the centers. Our final population, which included 1,593 cases with AML, CML and MDS, was used in the univariate and multivariate analysis; these eligible cases came from 82 NMDP reporting centers.
The effect of transplant center on overall survival was tested using a score test for homogeneity and was found to be statistically significant. No significant interactions between center and the other main effects were found using a significance level of 1% (p<0.01). All multivariable analyses were stratified on centers to adjust for the center effect when determining whether outcomes were different depending upon conditioning regimens.
Allele typing for HLA A, B, C and DRB1 was available for most donors and recipients in our final population. Using previously defined matching criteria, well matched was defined as no known disparity at HLA-A,-B,-C and DRB1, partially matched as one known locus or a likely locus mismatch with their donors and mismatched as ≥2 locus disparity.(11)
Outcomes for recipients were reported to the NMDP Coordinating Center by the transplant centers on standardized NMDP case report forms submitted at the time of transplantation (baseline), at 100 days, 6 months, and annually thereafter. Acute GVHD was reported according to the consensus criteria for each organ stage.(12) The overall follow-up was 95%; patient follow-up at 1 year was 100%, at 3 years 99.33 % and at 5 years 98.50%. The occurrence of organ toxicities, relevant clinical and laboratory data were captured on the case report forms, including specific organ toxicities such as veno-occlusive disease (VOD) of the liver, and interstitial pneumonitis (IPN).
The primary endpoints studied were transplant-related mortality (TRM), relapse, disease-free survival (DFS), overall survival (OS), primary neutrophil and platelet engraftment. TRM was defined as death during a continuous complete remission. Relapse was defined as clinical or hematologic recurrence. Due to the timing of the study period, our definition of relapse for CML patients was based on hematologic relapse. For analysis of DFS, failures were relapse or death from any cause; patients alive and in complete remission were censored at time of last follow-up. For analysis of overall survival, failure was death from any cause; surviving patients were censored at the date of last contact. Neutrophil and platelet engraftment were assessed in patients who survived at least 21 days post transplant. Time to neutrophil engraftment was defined as the time to achieve a sustained absolute neutrophil count (ANC) of ≥ 500 cells/μL for three consecutive days. Time to platelet engraftment was defined as time to achieve a platelet count of 20,000/μL, evaluable at 7 days from the last platelet transfusion.
Disease stage was defined as follows: early diseases were AML in CR1, CML-CP1 and MDS of RA and RARS subtype; intermediate diseases were AML CR >1 or 1st relapse, CML-2nd chronic phase or accelerated phase; advanced diseases were AML with primary induction failure (PIF) or ≥2nd relapse, CML-blast phase, MDS of RAEB-1 or RAEB-2 subtype and CMML. The growth factor variable was defined as whether a growth factor was initiated on day -1 to day 7 post transplant to promote engraftment.
Patient-, disease- and transplant-related variables were compared between patients who received BuCy, Cy/TBI standard dose and Cy/TBI high dose conditioning groups using the chi-square statistic for categorical and the Kruskal-Wallis test for continuous variables. The BuCy group included those patients who received busulfan (usually 16 mg/kg) and cyclophosphamide (usually 120-200 mg/kg). The BuCy-2 and BuCy-4 regimen could not be distinguished with the available data as it did not specify the exact cyclophosphamide dose in all cases. Dose adjustment by weight is not reported, nor is the use of busulfan levels to direct dosing. In addition the data set did not capture whether any of the patients received IV busulfan. However as the approval for IV busulfan did not occur until March of 1999, its influence was negligible. Because of large variations of the dose of radiation therapy used, a proportional hazards regression model using a forward selection was used on several TBI categories to see if a breakpoint which had an impact on survival could be defined. This step was repeated for various TBI categories by 100's by 50's and by 25's cGy. By this method, only 1000-1200 cGy entered into the survival model. This was used as the breakpoint for the dichotomous choice of standard-dose (1,000-1200 cGy) and high-dose (1320-1500cGy) TBI.
Univariate probabilities of disease-free survival and overall survival were calculated using the Kaplan-Meier estimator.(13) The log-rank test was used for comparing survival curves. Probabilities of TRM, relapse, neutrophil engraftment, platelet engraftment, acute GVHD chronic GVHD, veno-occlusive disease (VOD) and interstitial pneumonitis (IPN) were calculated using cumulative incidence estimates. Apart from TRM, the cumulative incidence calculated for the aforementioned outcomes treated death as a competing risk.(14) Relapse was treated as a competing risk for TRM. The cumulative incidence for chronic GVHD was only considered in patients who survived for at least 80 days.
In the multivariate analysis, proportional hazards regression models were used to investigate the effects of different conditioning regimens (BuCy, standard-dose Cy/TBI, and high-dose Cy/TBI) on clinical endpoints while controlling for effects of the other potentially confounding factors. Main effects were considered significant if the corresponding p-values were smaller than 5%, interactions and proportional hazard assumptions were tested at 1%. The main effect being tested (BuCy, standard-dose Cy/TBI, and high-dose Cy/TBI) was forced into each model in a stepwise selection procedure, regardless of its statistical significance for outcome. The following variables were considered as possible confounders and included in the regression model if they demonstrated a statistically significant association with the primary outcome of interest: HLA match status, disease and disease stage, recipient and donor age, gender and CMV status, Karnofsky score at transplant, interval from diagnosis to transplant and year of transplantation. All computations were performed using the statistical package SAS version 9.1.
Patient demographics and characteristics are shown in Table 1. A total of 1,593 patients were evaluated in this study; 318 received BuCy, 420 received standard-dose Cy/TBI and 855 received high-dose Cy/TBI. The median follow up of survivors was 97 months (range, 12-168). The median age of the entire group was 37 years (range, 1-58). Patients who received high-dose Cy/TBI regimens were slightly younger than patients who received standard-dose Cy/TBI or BuCy regimens (median age 36 vs. 38 vs. 40 years, respectively) (p < 0.001). The use of TBI or BuCy based regimens was associated with the disease, the stage of disease and the HLA matching criteria. Of the 318 patients who received a BuCy regimen, 18% had MDS. In comparison, only 5% and 7% of patients receiving standard dose or high dose Cy/TBI respectively had MDS (p<0.001). Thirty-three percent of the patients who received high-dose Cy/TBI had AML, compared to 23% and 15% of patients who received BuCy or standard-dose Cy/TBI with AML. A greater proportion of patients with intermediate and advanced stage disease received high dose Cy/TBI as their preparative regimen. Fifty-five percent of patients who received high-dose Cy/TBI had intermediate or advanced stage disease. In contrast, 28% of patients who received standard-dose Cy/TBI and 34% of patients who received Bu/Cy regimens had intermediate or advanced stage disease (p<0.001). In addition, 37% of patients receiving the high dose Cy/TBI had an HLA mismatched transplant compared to 24% of patients receiving either BuCy or standard dose Cy/TBI (p<0.001). There was no significant association between the preparative regimens used and the year of infusion, donor age, recipient/donor CMV status, use of growth factors, or interval from diagnosis to transplantation.
The cumulative incidence of neutrophil and platelet engraftment is shown in Table 2. By day +28 post BMT, the cumulative incidence of neutrophil engraftment was 88% (95% CI, 84-91) in the standard dose Cy/TBI group, compared to 81% (95% CI, 78-84) in the high dose Cy/TBI and 80% (95% CI, 75-84) in the BuCy patients (p=0.001). The multivariate analysis also confirmed an advantage for neutrophil engraftment in the standard dose Cy/TBI compared to either the BuCy or high-dose Cy/TBI group (p <0.001) (Table 3).
The cumulative incidence of platelet engraftment was also higher in the standard-dose Cy/TBI group (Table 2). On day +30, 49% (95% CI, 43-54) of patients in the standard-dose Cy/TBI group had evidence of platelet engraftment (>20,000/μL) compared to 29% (95% CI, 26-33) in the high-dose Cy/TBI group and 38% (95% CI, 33-44) in the BuCy group (p<0.001). This advantage in platelet engraftment in the standard dose Cy/TBI group was also significant at day 100 (p=0.002) and at 1 year (p=0.002).
The cumulative incidence of grade II-IV aGVHD (Table 2) shows a statistically significant difference between the regimens however in the multivariate analysis there was no difference in the incidence of aGVHD between the three groups (p=0.96) (Table 3). The cumulative incidence of grade III-IV aGVHD (Table 2) shows no significant difference between the regimens however, the multivariate analysis showed that patients who received the high dose Cy/TBI regimen had an increased risk of developing grade III-IV aGVHD (p=0.011). At 6 months, the cumulative incidence of cGVHD was higher in patients receiving either standard dose TBI or high dose TBI regimens 35% (95% CI, 31-40) and 28% (95% CI, 25-31) respectively compared to patients receiving busulfan regimens whose incidence of cGVHD was 23% (95% CI, 18-28) (p=0.001) (Table 2). At 1 year there continued to be a lower incidence of cGVHD in the busulfan group compared to either TBI group but these differences did not achieve statistical significance (p=0.07).
At five years the adjusted overall survival for the BuCy, standard dose Cy/TBI and high dose Cy/TBI was 35% (95% CI, 30-40), 32% (95% CI, 28-36), and 33% (95% CI, 30-36) respectively (p=0.779) (Figure 1). In the multivariate analysis for overall survival there was no difference found in the overall survival regardless of the preparative regimen used (p=0.236) (Table 3). In the multivariate analysis there was also no difference in the DFS between the groups (p=0.464).
The cumulative incidence of TRM and relapse is also shown in Table 2. At 1 year, the cumulative incidence of TRM was similar among all three groups; 48% (95% CI, 43-54) in the BuCy group, 43% (95% CI, 39-48) in the standard dose Cy/TBI group and 47% (95% CI, 43-50) in the high dose Cy/TBI group (p=0.37). In the multivariate analysis there was no difference in the relative risk of transplant related mortality between the three groups (p=0.384) (Table 3). In the univariate analysis the cumulative incidence of relapse was significantly higher at 1 year in the high dose Cy/TBI group compared to the BuCy group or the standard dose Cy/TBI group (p<0.001). However in the multivariate analysis there was no difference in the relative risk of relapse between the three groups (p=0.155) (Table 3).
The incidence of veno-occlusive disease (VOD) using Seattle criteria(15) at day 100 was higher in patients who received the BuCy regimen - 21% (95% CI, 16-26) compared to patients who received either standard dose Cy/TBI 13% (95% CI, 10-16) or high dose Cy/TBI 15% (95% CI, 13-18) (p=0.02). The development of IPN at day 100 was similar between the three groups: 20% (95% CI, 16-25) in the BuCy group, 22% (95% CI, 18-26) in the standard dose Cy/TBI group and 21% (95% CI, 19-24) in the high dose Cy/TBI group (p=0.91). These outcomes were not assessed in multivariate analysis.
Historically, total body irradiation has been the main modality to provide both tumoricidal and immunosuppressive effects to facilitate engraftment of donor cells.(16, 17) Since the initial introduction of the busulfan and cyclophosphamide combination (BuCy-4) as a preparative regimen for leukemias in the mid-1980s, there have been increasing reports of its efficacy.(18) Modification of this regimen with the reduction of cyclophosphamide to 60 mg/kg/day for 2 days (BuCy-2) was found to be equally efficacious with apparent reduction of toxicity.(19, 20) Although there were no prospective trials comparing BuCy-2 and BuCy-4, many centers adopted BuCy-2 as standard for this combination. In the present study, we compared the outcome of 1593 patients with myeloid disorders, AML, CML, and myelodysplasia who received URD transplantation facilitated by the NMDP. There was no statistical differences noted in OS, DFS, TRM and relapse between patients who received BuCy or the TBI based regimens.
Presently, five prospective randomized trials have compared TBI-based and BuCy regimens in recipients of sibling donor transplantation.(4-6, 21, 22) Most included patients with various stages of acute and chronic myeloid leukemia. In the two studies of patients in chronic phase CML, the treatment-related mortality, overall and disease-free survivals of patients who received TBI and BuCy were similar.(6, 22) The remaining three studies included patients with early and advanced stages of disease.(4, 5, 21) Two studies showed a significant survival advantage for patients who received TBI regimens; one in AML in CR1 (4), and one in advanced disease.(5) In the French multicenter randomized trial comparing BuCy-2 with a TBI-based regimen in matched sibling transplantation for patients with AML in first remission, there was a significantly higher relapse rate in patients who received the BuCy-2 regimen, resulting in poorer overall and disease-free survival.(4) There was also a higher relapse in patients who received BuCy for AML in first remission in a retrospective review from the IBMTR.(23) In this study in spite of the higher relapse rate the survival was equivalent. The higher relapse in both these studies which included sibling transplants may not be directly comparable to our study which only focused on unrelated transplants. In support of our study findings, a published meta-analysis suggested equivalent survival between the regimens.(24) In addition, a combined long-term follow up of 4 of the previously published randomized trials, patients with CML and AML undergoing HLA-identical sibling transplantation had equivalent survival whether they received Cy/TBI or BuCy regimen.(25)
The risk of developing grade III-IV aGVHD was higher in patients who received the high-dose TBI based regimens compared to patients who received either the standard-dose TBI or BuCy regimen. This higher risk of severe acute GVHD in patients who received higher doses of TBI may be related to the increased intensity of the preparative regimen. Preparative regimens that cause more tissue damage (higher intensity regimens) may enhance the risk of GVHD.(26, 27) It is possible that different preparative regimens, e.g. busulfan-based regimens, may attenuate tissue injury, thus lessening the risk of aGVHD. In the univariate analysis, the cumulative incidence of cGVHD was higher for Cy/TBI regimens at 6 months but only showed a trend at 1 year and 2 years with the incidence of 39% - 48%, which was lower than ~ 70% observed in the prospective clinical trial comparing tacrolimus and cyclosporine in URD transplantation.(28) The difference of cGVHD incidence is most likely due to lower reporting sensitivity of registry data in patient with limited chronic GVHD compared to prospective data captured in real-time.
Neutrophil engraftment at day 28 and 60 was higher in patients who received standard-dose Cy/TBI compared to patients who received either the BuCy or high-dose Cy/TBI regimens. This was statistically significant in both the univariate and multivariate analysis. Previous studies from the NMDP with CML patients indicated that the rate of neutrophil engraftment was improved in patients who received TBI-based regimens.(11) The reason was thought to be due to greater immunosuppressive property of TBI-based regimens compared to busulfan-containing regimens. However the patients who received the high-dose TBI-based regimens in our study had a lower incidence of neutrophil engraftment compared to patients who received standard-dose TBI. Several differences in the characteristics of patients who received the high-dose Cy/TBI may account for the differences seen in our current study. Patients receiving the high-dose Cy/TBI regimens had a greater likelihood of receiving an HLA-mismatched transplant and had advanced disease at the time of transplantation; both of these characteristics could potentially impede the recovery of neutrophils. Furthermore, other significant covariates that influenced neutrophil recovery in the multivariate analysis were HLA match status, karnofsky score and total nucleated cell dose, the latter was higher in patients receiving BuCy regimen. Attainment of the desired cell dose in marrow harvest from unrelated donor cannot consistently be achieved due to weight discrepancy and technical skill of the operators. Peripheral blood stem cell harvest is probably less operator dependent and weight discrepancy is less of an issue, thus the issue of neutrophil engraftment might be more consistent. This is being addressed by the just completed BMTCTN trial comparing marrow and peripheral blood stem cell transplantation. It is also possible that the use of targeted blood levels of busulfan and intravenous formulation of busulfan might improve tumor cytoreduction in patients with advanced disease and enhance engraftment.(29-31)
In this study, the relative risk of IPN was similar in all three groups. This finding is similar to previously reported studies on the risk of IPN after transplant.(32) The pathogenetic mechanism of IPN in the allogeneic transplant setting might not be directly related toxicity of any particular components of the conditioning regimens. Immunologic mechanism might have played a larger role in the development of IPN, thus overriding the impact of the preparative regimen on this complication.(33) The risk of VOD was higher in patients who received BuCy, compared to either Cy/TBI regimen in our study. The association of the preparative regimen and the risks of VOD were inconclusive in 3 randomized studies. One reported significantly higher incidence of VOD in patients who received busulfan-based regimen (5, 32); the other 2 studies did not find a difference in the risk of VOD between those who received TBI-based regimen and busulfan-based regimen.(6, 7, 22) The risk of VOD might be minimized by strategies such as the use of ursodiol as a protective agent,(34) therapeutic monitoring of busulfan and using pharmacokinetic parameters(35, 36) or using a targeted steady state plasma level of busulfan at 800-900 mg/mL.(30) More recently, the introduction of the intravenous formulation of busulfan may also decrease the risk of hepatic veno-occlusive disease.(37)
Despite our best effort to delineate the interaction and influence of confounding factors in this large cohort of patients, some practical questions remain unanswered. For instance, these findings do not take into account several changes in the practice of unrelated transplantation and therefore caution must be used in generalization of these results. Certainly, current improvement of clinical practice that includes new microbial agents, antibodies to perturb immune response, GVHD prophylactic regimens, selection of donor based on allelic typing and stem cell source has improved the outcome of unrelated transplants. In addition, the number of PBSC transplants has increased dramatically and now represents the major source of stem cells in patients undergoing unrelated transplantation. It is unclear whether the same observation in marrow transplant recipient as in this study would be true in the recipients of a PBSC transplant. This study does not delineate the role of IV busulfan which is increasingly used in current practice. Intravenous preparations of busulfan can potentially decreased the risk of VOD by avoiding the first pass effects on the liver. Intravenous busulfan was approved by the FDA for use in 1999 and therefore not available for vast majority of patients in this study. Finally, transplantation in CML was the most common myeloid neoplasm in this study, but it is an uncommon necessity since the introduction of tyrosine kinase inhibitors as frontline therapy for this disease.
In this large cohort analysis of CIBMTR registry data, we conclude that Cy/TBI and BuCy regimens result in similar clinical outcome in URD transplant recipient with myeloid malignancies. Whether one regimen is superior in some subsets of URD transplant recipients is uncertain and will need further prospective study. Other prognostic variables might have larger influence on survival than the preparative regimens described in this report.
We would also like to acknowledge Esteban M. Abella, MD, Asad Bashey, MD, PhD, Scott I. Bearman, MD, Arkadiusz Dudek, MD, PhD, FACP, Stephanie Elkins, MD, Nancy A. Kernan, MD, and James Wade, MD for their contributions to this manuscript.
The CIBMTR is supported by Public Health Service Grant/Cooperative Agreement U24-CA76518 from the National Cancer Institute (NCI), the National Heart, Lung and Blood Institute (NHLBI) and the National Institute of Allergy and Infectious Diseases (NIAID); a Grant/Cooperative Agreement 5U01HL069294 from NHLBI and NCI; a contract HHSH234200637015C with Health Resources and Services Administration (HRSA/DHHS); two Grants N00014-06-1-0704 and N00014-08-1-0058 from the Office of Naval Research; and grants from AABB; Aetna; American Society for Blood and Marrow Transplantation; Amgen, Inc.; Anonymous donation to the Medical College of Wisconsin; Association of Medical Microbiology and Infectious Disease Canada; Astellas Pharma US, Inc.; Baxter International, Inc.; Bayer HealthCare Pharmaceuticals; BloodCenter of Wisconsin; Blue Cross and Blue Shield Association; Bone Marrow Foundation; Canadian Blood and Marrow Transplant Group; Celgene Corporation; CellGenix, GmbH; Centers for Disease Control and Prevention; ClinImmune Labs; CTI Clinical Trial and Consulting Services; Cubist Pharmaceuticals; Cylex Inc.; CytoTherm; DOR BioPharma, Inc.; Dynal Biotech, an Invitrogen Company; Enzon Pharmaceuticals, Inc.; European Group for Blood and Marrow Transplantation; Gambro BCT, Inc.; Gamida Cell, Ltd.; Genzyme Corporation; Histogenetics, Inc.; HKS Medical Information Systems; Hospira, Inc.; Infectious Diseases Society of America; Kiadis Pharma; Kirin Brewery Co., Ltd.; Merck & Company; The Medical College of Wisconsin; MGI Pharma, Inc.; Michigan Community Blood Centers; Millennium Pharmaceuticals, Inc.; Miller Pharmacal Group; Milliman USA, Inc.; Miltenyi Biotec, Inc.; National Marrow Donor Program; Nature Publishing Group; New York Blood Center; Novartis Oncology; Oncology Nursing Society; Osiris Therapeutics, Inc.; Otsuka Pharmaceutical Development & Commercialization, Inc.; Pall Life Sciences; PDL BioPharma, Inc; Pfizer Inc; Pharmion Corporation; Saladax Biomedical, Inc.; Schering Plough Corporation; Society for Healthcare Epidemiology of America; StemCyte, Inc.; StemSoft Software, Inc.; Sysmex; Teva Pharmaceutical Industries; The Marrow Foundation; THERAKOS, Inc.; Vidacare Corporation; Vion Pharmaceuticals, Inc.; ViraCor Laboratories; ViroPharma, Inc.; and Wellpoint, Inc. The views expressed in this article do not reflect the official policy or position of the National Institute of Health, the Department of the Navy, the Department of Defense, or any other agency of the U.S. Government.
AUTHOR CONTRIBUTIONSJPU and JDR designed the study. MH, CK, SG prepared the data file. M-AA and ST performed statistical analysis. JPU, M-AA, ST, MH, SG, CA, KSB, BJB, MB, KWC, EC, SMD, JF, GAH, CK, PLM, VR, OR, DJW and JDR participated in the interpretation of data and preparation of manuscript.