Cytopenias after day 28 in allogeneic hematopoietic cell transplantation: impact of recipient/donor factors, transplant conditions and myelotoxic drugs

Abstract

Background Secondary cytopenias are serious complications following hematopoietic cell transplantation. Etiologies include myelotoxic agents, viral infections, and possibly transplant-related factors such as the intensity of the conditioning regimen and the source of stem cells.Design and Methods We retrospectively analyzed data from 2162 hematopoietic cell transplant recipients to examine the effect of these factors on overall cytopenias occurring after 28 days in hematopoietic cell transplantation.Results Advanced age of the patient, recipient cytomegalovirus seropositivity, unrelated donor status, human leukocyte antigen mismatch and lower doses of transplanted CD34⁺ cells (≤ 6.4×10⁶/kg) significantly increased the risk of cytopenias after day 28. Non-myeloablative hematopoietic cell transplantation had protective effects on anemia and thrombocytopenia after day 28 (adjusted odds ratio 0.76, probability value of 0.05 and adjusted odds ratio 0.31, probability value of <0.0001, respectively) but not on overall or ganciclovir-related neutropenia. This lack of protection appeared to be due to the use of mycophenolate mofetil in the majority of recipients of non-myeloablative hematopoietic cell transplants. Peripheral blood stem cells did not confer protection from cytopenias when compared to bone marrow.Conclusions Elderly patients appear to be more prone to cumulative toxicities of post-transplant drug regimens, but non-myeloablative conditioning, optimized human leukocyte antigen matching, and higher doses of CD34⁺ cell infusions may reduce the risk of cytopenia after day 28.

Introduction

Secondary cytopenia is a common complication after hematopoietic cell transplantation (HCT). Causes include viral infection, septicemia, graft-versus-host disease (GVHD), and myelotoxic drugs.1–5 Of the commonly used drugs with myelotoxic potential, ganciclovir is particularly prone to cause neutropenia, which occurs in up to 40% of allograft recipients and may increase the risk of invasive bacterial and fungal infections.1,6 The underlying mechanism of ganciclovir-related neutropenia is a dose-dependent inhibition of DNA-polymerase in hematopoietic progenitor cells.7 We previously reported that ganciclovir-related neutropenia is associated with low marrow cellularity, hyperbilirubinemia, and elevated serum creatinine levels after myeloablative conditioning (M-HCT).1 However, it is not known how non-myeloablative conditioning (NM-HCT) influences the incidences of secondary cytopenias in general and ganciclovir-related neutropenia in particular.

Less toxic non-myeloablative conditioning regimens that can be successfully used in elderly patients and/or patients with comorbidities have been developed.8–13 Non-myeloablative conditioning does not eradicate host hematopoiesis and allows relatively prompt hematopoietic recovery within 28 days after transplantation.14,15 NM-HCT may, therefore, be associated with a lower incidence of cytopenias, including ganciclovir-related neutropenia. In addition, the increased use of hematopoietic growth factors for secondary neutropenia at moderate levels in recent years may also be associated with a lower risk of profound levels of neutropenia.

The purpose of this study was to examine risk factors for the occurrence of cytopenias 28 days after HCT as a surrogate for secondary neutropenia overall, and ganciclovir-related neutropenia in particular.

Design and Methods

Study population

This retrospective study population consisted of 2162 consecutive patients who underwent HCT between 1998 and 2006 at the Fred Hutchinson Cancer Research Center (FHCRC) (Seattle, WA, USA). The retrospective analysis was approved by the Institutional Review Board of the FHCRC. Informed consent was obtained from all the patients before HCT. We compared events between 534 patients undergoing NM-HCT and 1628 contemporaneous patients undergoing M-HCT who served as a comparison group (Table 1). Clinical and laboratory data were extracted from the computerized database and from patients’ charts.

The most common regimens for NM-HCT were fludarabine (30 mg/m/day for 3 consecutive days) together with low-dose total body irradiation (2 Gy, day 0), or low-dose total body irradiation (2 Gy, day 0) alone. In contrast, many different types of conditioning regimens were used for M-HCT. The most common regimen for M-HCT consisted of cyclophosphamide (60 mg/kg/day for 2 consecutive days) followed by total body irradiation (12 Gy or 13.2 Gy) or busulfan (4 mg/kg/day for 4 consecutive days) followed by cyclophosphamide (60 mg/kg/day for 2 consecutive days). The NM-HCT group included more elderly patients, almost exclusive use of peripheral blood stem cells as the source of stem cells, and higher doses of transplanted CD34 cells than those in the M-HCT group (Table 1).

In terms of GVHD prophylaxis, M-HCT patients most commonly received a combination of a calcineurin inhibitor (either cyclosporine or tacrolimus) and short-term methotrexate (15 mg/m intravenously on day 1, and 10 mg/m on days 3, 6, and 11). All NM-HCT patients received post-grafting immunosuppressants including mycophenolate mofetil (MMF) and a calcineurin inhibitor, cyclosporine or tacrolimus (Table 1)

Table 1.Characteristics of the patients for the neutropenia/transfusion analysis (allogeneic transplant from 1998 to 2006, hematologic malignancy n=2162).

MMF was administered at a dose of 15 mg/kg orally twice a day from day 0 to day 27 and discontinued for the human leukocyte antigen (HLA) matched-related NM-HCT patients; while for the unrelated NM-HCT patients MMF was given at a dose of 15 mg/kg orally two or three times a day from day 0 to day 40, with tapering to day 96. For the single HLA-antigen and combined HLA-antigen and allele-mismatched NM-HCT patients, 15 mg/kg MMF was given three times a day and then tapered at day 100 over 2 months.8,9,12,16,17

Transfusion requirements

Red blood cells were routinely transfused when the hematocrit fell below 26%. In patients with severe uremia, other causes of platelet dysfunction, active bleeding, or thrombocytopenia refractory to platelet transfusion, the hematocrit was maintained at 30% or above. A hematocrit of at least 30% was also maintained in patients with a history of cardiac or peripheral vascular disease, and in patients over 65 years old.

A platelet threshold for transfusion of 1.0×10/L was used for clinically stable, afebrile patients without evidence of hemorrhage, infection, or uncontrolled GVHD. Transfusions at higher platelet levels were given to patients receiving anti-coagulant medications and patients with abnormal coagulation times, platelet dysfunction such as uremia, or other bleeding diatheses. Invasive procedures, anticoagulation management, prevention of blood clots and management of central venous catheter-associated thrombosis required maintenance of higher platelet levels.

Infection surveillance, prophylaxis and pre-emptive therapy against cytomegalovirus

Cytomegalovirus (CMV) surveillance with polymerase chain reaction (PCR) analysis or the pp65 antigenemia assay was performed on a weekly basis until day 100 as previously described.18,19 After day 100, surveillance and pre-emptive therapy were recommended for CMV intermediate- and high-risk patients on a weekly or biweekly basis until day 365.

Pre-emptive ganciclovir treatment was started when CMV pp65 antigenemia/PCR became positive during the first 100 days after HCT. Ganciclovir (5 mg/kg IV twice daily) for 7 to 14 days was administered as induction therapy followed by a half-dose of ganciclovir (5 mg/kg IV daily) or valganciclovir 900 mg once a day orally as maintenance therapy until negative surveillance or day 100.20 All doses were adjusted based on the patients’ renal function according to the manufacturers’ recommendation. After day 100, pre-emptive therapy was recommended for patients with CMV pp65 antigenemia or more than 1000 copies/mL (assessed by PCR) as previously described.18,19

Figure 1.Cumulative incidence of neutropenia after day 28. The probabilities of neutropenia after day 28 (absolute neutrophil count < 500/L) according to (A) non-myeloablative (NM-HCT) vs. myeloablative hematopoietic stem cell transplantation (M-HCT), (B) patients’ age (≤ 40 vs. > 40 years), (C) MMF use (yes vs. no) and (D) CD34⁺ cell dose (≤ 6.4×10⁶/kg vs. > 6.4×10⁶/kg) are illustrated. The cumulative incidence of development of neutropenia was higher for NM-HCT than it was for M-HCT. However, in a multivariate analysis NM-HCT was not a significant risk factor for development of neutropenia.

Pneumocystis jirovecii prophylaxis consisted of trimethoprim sulfamethoxazole as the primary agent and dapsone as the secondary agent. Identical doses of both drugs were used for all patients, regardless of conditioning regimen.21 However, recipients of non-myeloablative conditioning regimens started prophylaxis at day 28 after HCT while recipients of myeloablative transplantation received pretransplant dosing which was then resumed after neutrophil engraftment.

Definitions of cytopenias after day 28

We evaluated neutropenia, anemia and thrombocytopenia after day 28, and ganciclovir-related neutropenia. Ganciclovir-related neutropenia was defined as non-relapse–related neutropenia (absolute neutrophil count < 500/μL and < 200/μL) after the start of pre-emptive therapy for pp65 antigenemia/PCR positivity in patients with an absolute neutrophil count greater than 1000/μL at the time of CMV infection. Neutropenia after day 28 was defined as absolute neutrophil counts less than 500/μL and less than 200/μL occurring any time between day 28 post-HCT and day 120 among relapse-free patients. We used transfusion support after day 28 as a surrogate marker of anemia and thrombocytopenia. Significant anemia and thrombocytopenia beyond day 28 were both defined as the upper 25 percentile of transfusion support after day 28, up to the first of day 80, death or relapse. Specifically, we defined patients who received more than 0.8 units of red blood cell transfusions per week as cases with anemia after day 28; similarly, we defined patients who were given more than 1.6 units of platelets per week up to day 80 as cases with thrombocytopenia.

Statistical analysis

The characteristics of NM-HCT and M-HCT patients were summarized using frequency counts and percentages for categorical variables and medians and ranges for continuous variables. The cumulative incidence of neutropenia after day 28 was estimated by previously described methods, with death or relapse treated as a competing risk. Univariate and multivariate Cox regression models were used to estimate hazard ratios and 95% confidence intervals (95% CI) for risk factors associated with neutropenia after day 28 and ganciclovir-related neutropenia as defined above. Univariate and multivariate logistic regression models were used to estimate odds ratios for risk factors associated with anemia and thrombocytopenia. Cox regression was used to perform a landmark analysis among patients alive and disease-free at day 100, to evaluate the impact of prior cytopenias and other risk factors on subsequent non-relapse mortality, defined as any death without prior relapse. Covariates included recipient/donor age and sex, recipient/donor race, donor CMV serostatus, sex mismatch, HLA disparity, donor relationship, intensity of conditioning, stem cell source, T-cell-depleted conditioning, year of transplantation, disease risk, GVHD prophylaxis, acute GVHD and chronic GVHD. Acute and chronic GVHD and other post-transplant factors were analyzed as time-dependent variables. Variables with a significance level of less than 0.05 in the univariate models were candidates for the multivariate models. All P values are two-sided and unadjusted for multiple comparisons.

Results

Risk factors for cytopenias after day 28

Among the 1818 patients with neutrophil engraftment at day 28, 711 (39%) had at least one form of cytopenia after day 28: 103 (6%) had neutropenia only, 128 (7%) had anemia only, 102 (6%) had thrombocytopenia only and 123 (7%) had all three cytopenias. Neutropenia after day 28 was significantly more frequent in NM-HCT than in M-HCT (23% and 13%, respectively) (Figure 1A).

In univariate analysis, the risk factors for neutropenia (< 500/μL) after day 28 included the patients’ age (> 40 years), recipient CMV seropositivity, patients at higher risk of CMV, unrelated donor status, receipt of NM-HCT, use of MMF, lower CD34 cell dose (≤ 6.4×10/kg) in the graft, chronic GVHD, high bilirubin level (> 6 mg/dL), and elevated creatinine (> 2 mg/dL) (Table 2). In a multivariate model, we identified patients’ age, CMV seropositivity, unrelated donor status, HLA mismatched donor, MMF use and lower CD34 cell dose (≤ 6.4×10/kg) as significant risk factors for neutropenia after day 28 in both HCT with bone marrow and peripheral blood stem cells (Table 3) (Figure 1B-D). Analysis of a lower threshold for defining neutropenia (< 200/μL) did not reveal additional risk factors (data not shown).

Results of univariate analyses for anemia and thrombocytopenia after day 28 are presented in Table 4. ABO-mismatched donors, patients’ age (> 40 years), female donor, patients at higher risk of CMV, unrelated donor, HLA-mismatched donor, bone marrow as the stem cell source and lower CD34 cell dose (≤ 6.4×10/kg) were risk factors for anemia after day 28. Risk factors for thrombocytopenia after day 28 included ABO-mismatched donor, unrelated donor, HLA-mismatched donor, bone marrow as the stem cell source and CD34 cell dose (≤ 6.4×10/kg) (Table 4). In a multivariate model, ABO-mismatched donor, patients’ age (> 40 years), CMV infection, unrelated donor, HLA-mismatched donor and lower CD34 cell dose (≤ 6.4×10/kg) were identified as common risk factors for anemia and thrombocytopenia after day 28 in both HCT with bone marrow and peripheral blood stem cells (Table 5). CMV serostatus was no longer significant when active post-transplant CMV infection was included in the model.

NM-HCT was significantly associated with a lower incidence of anemia and thrombocytopenia after day 28 in both univariate and multivariate models (Tables 4 and 5).

Risk factor for ganciclovir-related neutropenia

The cumulative incidence of ganciclovir-related neutropenia was similar for NM-HCT and M-HCT recipients (26% and 22%, respectively). A univariate model for ganciclovir-related neutropenia is shown in Table 2. In the univariate model, we found patients’ age (> 40 years), unrelated donor status, MMF use, lower CD34 cell dose (≤ 6.4×10/kg) and high bilirubin level (> 6 mg/dL) to be significant risk factors for ganciclovir-related neutropenia (Table 2). All factors except MMF use and high bilirubin levels remained statistically significant risk factors for ganciclovir-related neutropenia (Table 3). Analysis of a lower threshold for neutropenia (< 200/μL) did not reveal additional risk factors (data not shown).

The impact of cytopenias after day 28 on non-relapse mortality

In a multivariate analysis of non-relapse mortality, we included acute GVHD, age of the patients and donors, sex of the patients and donors, patients’ CMV status, donor relation, HLA disparity, stem cell source, MMF use and the intensity of the conditioning regimen as covariates. After adjustment for these factors, neutropenia, anemia and thrombocytopenia after day 28 were all significant independent risk factors for non-relapse mortality after HCT (neutropenia: HR=1.81, 95% CI 1.4–2.4, P<0.0001; anemia: HR=1.56, 95% CI 1.2–2.1, P=0.002; thrombocytopenia: HR=2.35, 95% CI 1.7–3.2, P<0.0001) (Table 6).

Table 2.Univariate risk factors for neutropenia (absolute neutrophil count <500/μL).

Table 3.Multivariate risk factors for neutropenia (absolute neutrophil count < 500/μL).

Discussion

This study provides novel and somewhat unexpected results on the risk of cytopenias after HCT. Older recipient age, low CD34 cell dose, an unrelated donor, and HLA mismatch were risk factors for cytopenias after transplantation. Non-myeloablative conditioning was associated with significantly reduced incidences of anemia and thrombocytopenia after day 28, but not of neutropenia.

Table 4.Univariate risk factors for significant anemia and thrombocytopenia after day 28.

Table 5.Multivariate risk factors for significant anemia and thrombocytopenia after day 28.

We hypothesized that non-myeloablative conditioning is associated with less neutropenia after day 28. Surprisingly, in this study we did not find a significant reduction of neutropenia either overall or in the context of ganciclovir use. Overall, neutropenia after day 28 occurred in 13% of patients. The exact contribution of MMF to the relatively high rates of neutropenia in NM-HCT recipients cannot be determined since MMF was given to all patients receiving non-myeloblative conditioning. MMF was significantly associated with neutropenia even after controlling for donor relatedness (which determined the duration of drug use). However, neutropenia is an important adverse effect of MMF and cumulative toxicity with ganciclovir is plausible and has been described.22 Our study also identified other factors that might explain the high rate of neutropenia in NM-HCT. We found older recipient age to be a risk factor for both neutropenia after day 28 and ganciclovir-related neutropenia. NM-HCT is more commonly done in older patients. The effect of older recipient age may be mediated by subclinical renal dysfunction (especially tubular function23), which may lead to inadvertent overdosing of myelotoxic drugs that are eliminated through the kidneys and whose doses are adjusted only by creatinine clearance (which does not measure tubular function). Such an effect would be consistent with the pharmacokinetic properties and the toxicity profile of ganciclovir, which includes predominantly neutropenia but not thrombocytopenia and anemia.

Table 6.Multivariate risk factors for non-relapse mortality. Landmark analysis at day 100, n=1489, 331 events.

There are limited data on cytopenia after day 28 relative to the intensity of the conditioning regimen. Severe GVHD, myelotoxicity associated with drugs such as ganciclovir, trimethoprim sulfamethoxazole or MMF, as well as viral and severe fungal or bacterial infections have all been associated with an increased risk of neutropenia after day 28.1,2,4,22 Furthermore, Bruno et al. previously reported unrelated donor, grade II–IV acute GVHD, impaired renal function, the combination of busulfan and cyclophosphamide, total body irradiation, stem cell dose and infections as risk factors for secondary failure of platelet recovery among M-HCT.5 The present study extended our previous findings that NM-HCT may also have protective effects against thrombocytopenia and anemia.24 We identified HLA mismatch, CMV serostatus and the CD34 cell count as additional risk factors for both outcomes in multivariable models. We speculate that the higher doses of CD34 cells and the reduced intensity of the conditioning regimen used in NM-HCT contributed to the lower rates of anemia and thrombocytopenia.

The CD34 cell dose rather than the stem cell source per se was an important risk factor for all cytopenias examined in this study. When the cell dose was included in the multivariable models the stem cell source was no longer significant, suggesting that the protective effect of peripheral blood stem cells for anemia and thrombocytopenia seen in the univariate analyses was mediated by the higher dose of CD34 cells (Tables 3 and 5).

CMV serostatus of the recipient was a risk factor for both anemia and thrombocytopenia requiring blood products (Table 5), a finding not previously appreciated in HCT recipients.25 When CMV serostatus and active CMV infection were included in a multivariable analysis, active CMV infection remained significant while CMV serostatus was no longer significant, suggesting that active CMV infection or pre-emptive therapy was responsible for the effect. The relative contribution of CMV infection compared to that of its treatment cannot be determined from this study. Ganciclovir has not been associated with thrombocytopenia or anemia or an increased use of blood products in several placebo-controlled randomized trials in HCT recipients.6,26,27 A previous risk factor analysis in myeloablative HCT recipients between 1990 and 1997 did not identify CMV serostatus as a risk factor for thrombocytopenia, but the use of platelet products was not analyzed in that study.5 Based on the lack of association with anemia and thrombocytopenia in randomized trials of ganciclovir, we speculate that CMV infection itself might be responsible for the effect.4,28

Our study has several limitations, including the retrospective nature of the analysis and that the analysis of concomitant medications was performed by protocol only. With regard to the non-myeloablative conditioning, the results can probably not be extrapolated to other types of reduced-intensity conditioning regimens. However, the strength of the analyses lies in the large sample size, the number of clinically important factors analyzed, a homogeneous transplant protocol, and highly standardized supportive care strategies.

In conclusion, the study provides a comprehensive analysis of factors associated with cytopenias after day 28 in HCT recipients. Unexpectedly, NM-HCT did not reduce the risk of neutropenia after day 28 overall or in the context of ganciclovir treatment. The high rates of neutropenia appear to be linked to the use of MMF and ganciclovir, emphasizing the need for less toxic immunosuppressive and anti-CMV drugs or strategies. In contrast, NM-HCT showed a protective effect against anemia and thrombocytopenia after day 28, probably through less toxic conditioning and higher doses of CD34 stem cells or almost exclusive use of peripheral blood stem cells. Finally, the study identified potentially modifiable factors that could be used before transplantation to minimize the risk of post-transplant cytopenias, including non-myeloablative conditioning, optimized HLA matching, and higher doses of CD34 cell infusions.

Acknowledgments

We thank Daniel Stachel, MD, for reviewing charts, and Gary Schoch and Craig Silva for database services.

Footnotes

• Funding: the authors are grateful for research funding from the National Institutes of Health, Bethesda, MD grants P01CA018029, P30CA015704, P01HL036444, P01CA078902, and K24HL093294. The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health nor its subsidiary Institutes and Centers. Hirohisa Nakamae was funded by the Graduate School of Medicine, Osaka City University, Osaka, Japan.
• Authorship and Disclosures The information provided by the authors about contributions from persons listed as authors and in acknowledgments is available with the full text of this paper at www.haematologica.org.
• Financial and other disclosures provided by the authors using the ICMJE (www.icmje.org) Uniform Format for Disclosure of Competing Interests are also available at www.haematologica.org.

• Received April 5, 2011.
• Revision received August 18, 2011.
• Accepted August 19, 2011.