|Home | About | Journals | Submit | Contact Us | Français|
Rapid lymphocyte recovery after bone marrow or peripheral blood transplantation is associated with improved survival. However, the impact of early lymphocyte recovery has not been examined after umbilical cord blood transplant (UCBT). We evaluated lymphocyte recovery in 360 consecutive patients with hematologic malignancy that underwent UCBT between 2001 and 2007. Uniform myeloablative (MA), reduced intensity conditioning (RIC) and graft vs. host disease prophylaxis regimens were used. In multivariate analysis, an ALC >200 ×106/L at day 30 (n=73) after MA conditioning was associated with superior 2-year overall survival (OS) (73% vs. 61%; p=0.02) [RR: 2.29; 95% CI: 1.15 – 4.56], progression-free survival (PFS) (68% vs. 54%; p=0.05) [RR: 1.96; 95% CI: 0.99 – 3.86] and less transplant related mortality (TRM) (8% vs. 28%, p<0.01) [RR: 4.38; 95% CI: 1.65 – 11.60] compared to ≤200×106/L (n=43). Similarly, an ALC >200 ×106/L at day 42 (n=105) after RIC was associated with superior 2-year OS (59% vs. 41%, p<0.01) [RR: 2.10; 95% CI: 1.3 – 3.41] and PFS (46% vs. 36%, p=0.05) [RR: 1.58; 95% CI: 1.01 – 2.49] compared to ≤200 ×106/L (n=55). There was no significant relationship between ALC and relapse. Rapid lymphocyte recovery early after UCBT predicts better survival in patients with hematologic malignancies.
Umbilical cord blood (UCB) has become an increasingly used stem cell source for the treatment of hematologic malignancies in both children and adults. Like other unrelated graft sources, post transplant complications including treatment related mortality (TRM) and relapse remain barriers to patient survival. The kinetics of neutophil recovery after UCB transplantation are delayed compared to other stem cell sources, such as bone marrow (BM) or peripheral blood (PB) and therefore may play a role in determining transplant outcomes(1–5).
Prior studies in bone marrow (BM) and peripheral blood (PB) transplant recipients show that the early differences in absolute lymphocyte count (ALC) are predictive of survival in patients with hematologic malignancies(6–11). To date, there is little data available on lymphocyte recovery and its influence on umbilical cord blood transplant (UCBT) outcomes. While low ALC appears to be associated with higher risk of infections(12), and lymphocyte function may be impaired after UCBT (13), a relationship between the kinetics of ALC recovery and survival has not yet been investigated. Based on the prior studies in BM and PB transplant recipients showing inferior survival(6–11), we hypothesized that delayed recovery of ALC after UCB transplantation might be associated with inferior transplant outcomes. We therefore evaluated whether differences in ALC early after UCBT were associated with transplant outcomes in patients with hematologic malignancies and assessed differences after myeloablative (MA) and reduced intensity conditioning (RIC) regimens.
This study included 360 consecutive patients with a hematologic malignancy who underwent either MA or RIC unrelated donor UCBT at the University of Minnesota between 2001 and 2007. Patient demographics are listed in table 1. Clear differences between these two populations were present, including age, time from diagnosis to transplant, disease, single vs. double UCB transplantation and HLA disparity (likely due to the addition of a second unit). Thirty-eight patients were excluded based on not having an available ALC reported (due to early death) or failed engraftment. All patients and/or guardians provided informed consent according to the principles of the Declaration of Helsinki prior to transplantation. Patient demographics and clinical outcomes data were prospectively collected and available from the University of Minnesota Bone Marrow Transplant Database.
Methods of HLA typing and UCB unit selection have been detailed elsewhere(14–15). Briefly, UCB units were typed with the recipient at intermediate resolution for HLA-A and -B and at allele level for HLA-DRB1. Other HLA loci were not considered in the selection algorithm. UCB units were selected if they were ≤2 locus HLA-mismatched with the recipient. The choice to receive one or two UCB units was based on cell dose available to the patient. Prior to 2002, the minimum acceptable cell dose for a single unit was 1.5 × 107 nucleated cells (NC) per kilogram (kg) and after 2002 the minimum cell dose was 2.5 × 107/kg. After 2003, the target cell dose was ≥3.0 × 107 NC/kg for those with HLA matched (i.e., 6/6) units and ≥4.0 × 107/kg for those with HLA mismatched (i.e., 4–5/6) units. If an UCB unit with the minimum cell dose was not available, the patient was transplanted with two partially HLA matched UCB units that were <2 locus HLA-mismatched with each other, as previously described (15–17).
The majority of patients from the MA group (n=111, 84%) received conditioning with cyclophosphamide (Cy) (120mg/kg), total body irradiation (TBI) (1320 cGy) and fludarabine (Flu) (75mg/m2), with the remaining patients receiving either the same Cy/TBI (without fludarabine, n=17, 13%) or a busulfan-containing preparative regimen (n=5, 4%). GVHD prophylaxis was cyclosporine (CsA)-based (CsA and mycophenolate mofetil [MMF] [n=111] and CSA/prednisone/anti-thymocyte globulin [ATG] [n=22]) for a minimum of 180 days after UCB transplantation as previously described (14, 17). For the patients who received RIC allo-HCT, the majority (n=176, 93%) received conditioning with Cy (50mg/kg), TBI (200cGy) and Flu (200mg/m2) (18). GVHD prophylaxis was CsA/MMF. The reason for using a RIC regimen was either due to age >45 years, concurrent infection or poor performance status.
All UCB units were thawed according to the methods of Rubinstein et al(19) and infused after hydration and premedication with acetaminophen and Diphenhydramine. UCB units were infused shortly after thawing/washing, and for patients receiving two units, they were infused sequentially within 30 minutes of each other. For double UCB unit transplant, order of unit infusion was random. Granulocyte colony stimulating factor (5μg/kg/day) was administered to all patients from day 1 until an absolute neutrophil count (ANC) ≥ 2.5 × 109/L was achieved for 2 consecutive days. All patients received Pneumocystis jiroveci pneumonia prophylaxis with trimethoprim-sulfamethaxole or pentamidine for the first 12 months of transplantation. Viral prophylaxis included acyclovir if seropositive for herpes simplex or cytomegalovirus prior to transplant. CMV surveillance was performed weekly on peripheral blood with ganciclovir treatment at the time of positive antigen or PCR testing.
GVHD was diagnosed clinically with histological confirmation when possible. Staging was based upon published criteria (20) and treatment of acute GVHD (aGVHD) clinical stage II or greater was with methlyprednisolone (≥48 mg/m2 intravenously or oral equivalent) daily for a minimum of 2 weeks prior to a taper over 8 weeks.
Kaplan-Meier estimates for 2-year overall survival (OS) and progression-free survival (PFS) are reported.(21) Cumulative incidence of relapse and transplant related mortality (TRM) at 2 years were calculated by treating deaths from other causes and relapse, respectively, as competing risks(22). Statistical comparisons of the time-to-event curves for OS and PFS were compared by the log-rank test and for relapse and TRM by Gray’s test. Event times were calculated from the date of transplantation to the event or the date of last contact. OS, PFS and time to relapse were censored at 4 years and TRM was censored at 2 years for all patients.
ALC cutoffs of 150, 200, 250 and 300 ×106/L at 21, 30, 42, 60 and 100 days after UCBT were evaluated in univariate models for each outcome; these cutoffs were chosen for comparison with previously published data on ALC outcomes in allo-HCT patients(6, 8, 10). Multivariate regression analyses with Cox regression and the Fine and Gray competing hazards methods were used as appropriate(23–24). Models were fit for each ALC cutoff while also adjusting for gender, age at transplant, total infused nucleated cell dose, CFU-GM, infused CD3 and CD34 cell doses, time from initial diagnosis to transplant, number of umbilical cord blood units transplanted, HLA-match (HLA for single transplant or minimum HLA for double transplant patients), recipient CMV status, and the development of acute GVHD (included as a time-dependent covariate). Acute GVHD and HLA-match were explored in multivariate models of relapse and TRM and were found to be not significant, therefore they were not adjusted for in the final relapse and TRM models to ensure convergence. All reported p-values are from adjusted models unless otherwise noted.
The total number of confirmed bacterial, fungal and viral infections among all patients was evaluated and the total number of early and late infections per patient determined; defined as before (early) and after (late) the ALC cutpoint day of interest. Patients were categorized into three groups: those with 0 infections, 1 infection or ≥2 infections during the particular time period. The number of patients in each group was compared by the ALC cutpoint using Fisher’s Exact test.
As the primary aim of this analysis was to identify the earliest time that an ALC may be predictive of outcomes, we reported the earliest day and most predictive ALC of outcome measures (p-values <0.10 for multiple outcomes) in our results for both MA and RIC cohorts, which were analyzed separately. Patient and transplant characteristics were compared by ALC cutoff using chi-squared and Fisher exact tests for categorical data and the Wilcoxon rank-sum test for continuous data. All analyses were performed using SAS version 9.1 and R version 2.7.
Patient ALCs were abstracted from medical records on or within 3 days of 21, 30, 42, 60 and 100 days after UCBT. ALC were recorded as missing at a particular time point if the WBC was <0.5 ×109/L as no manual differential counts were performed during severe leukopenia. If patients relapsed prior to 100 days, their ALC was included only up to the date of relapse. Of the 147 (41%) patients who received a MA regimen, 133 (90%) had neutrophil recovery and at least one ALC recorded. As shown in Table 1, 59 (44%) of the 133 patients had ALL and 74 (56%) had AML. Prior to UCBT, 58 (44%) patients were in first complete remission (CR1), 59 (44%) CR2, 9 (7%) CR3 and the remaining 7 patients were either primary induction failure (PIF) (n=2) or in relapse (n=5).
Of the total patients, 213 (59%) patients received a RIC regimen. One hundred and eighty nine (89%) had neutrophil recovery and at least one ALC recorded. As shown in Table 1, 108 had leukemia (AML n=62; ALL n=16; chronic myelogenous leukemia [CML] n=8; other leukemia/MDS n=22), 65 lymphoma (Hodgkins disease [HD] n=24; Non-Hodgkins lymphoma [NHL] n=41) and the remaining 16 patients had myeloproliferative diseases (n=9) or other hematologic malignancy (n=7).
At day 30 after MA UCB transplant patients with an ALC ≤200 ×106/L had significantly inferior OS (61% vs. 73%; p=0.02), PFS (54% vs. 68%; p=0.05) and greater TRM (28% vs. 8%; p=<0.01) at 2-years compared to patients with an ALC >200 ×106/L (Figure 1A,B; Table 2). Causes of TRM for patients ≤200 ×106/L were multi-organ failure (n=7), infection (n=4; fungal n=3, viral n=1), hemorrhage (n=1) and aGVHD (n=1) compared to infection (n=2; fungal n=1, bacterial n=1), aGVHD (n=2), multi-organ failure (n=1), and secondary late graft failure (n=1) for those >200 ×106/L. In contrast, ALC ≤ 200 ×106/L at day 30 was not significantly associated with risk of relapse (18% vs. 24%, p=0.72). In multivariate analysis, an ALC ≤200 ×106/L remained the only independent predictor of worse survival, PFS and higher TRM after adjusting for gender, age at transplant, total infused cell dose, CFU-GM, infused CD3 and CD34 cell dose, time from initial diagnosis to transplant, single vs. double umbilical cord blood transplant, HLA-match, recipient CMV status and development of acute GVHD (as a time-dependent covariate) (Table 2).
Pre-transplant patient and graft characteristics were compared across patients with a low or high ALC to determine whether these parameters were associated with day 30 ALC. While patients with an ALC ≤200 ×106/L at day 30 were more likely to receive 2 UCB units (p=0.05) and have slightly slower neutrophil recovery (22.8 vs. 20.9 days, p=0.054), there was no differences in age at transplant, disease status, preparative regimen, HLA-match, infused CD3 or CD34 cell dose or time to neutrophil engraftment (Table 3). In comparing the incidence of infections between MA patients, there were significantly more early infections (<30 days post-HCT) in patients with an ALC ≤200 ×106/L at day 30, but no significant difference in the number of late infections (days 31 – 100 post-HCT) based on day 300 ALC (Figure 2A and 2B).
The day 30 ALC ≤200 ×106/L after RIC UCB transplant did not show a significant difference in 2-year OS (48% vs. 54%; p=0.19), PFS (40% vs. 41%; p=0.35), relapse (38% vs. 34%; p=0.81) or TRM (22% vs. 25%; p=0.57) compared to patients >200 ×106/L. However, the day 42 ALC was predictive. Patients with a day 42 ALC ≤200 ×106/L had significantly inferior 2-year OS (41% vs. 59%; p<0.01) and PFS (36% vs. 46%; p=0.05) compared to patients >200 ×106/L (n=109) (Figure 3A; Table 2). There was no significant difference in either TRM (29% vs. 19%; p=0.24) (Figure 3B) or disease recurrence (44% vs. 31%, p=0.63) based on day 42 ALC ≤ or >200 ×106/L. Causes of TRM for patients ≤200 ×106/L at day 42 were multi-organ failure (n=8), infection (n=5; fungal n=2, bacterial n=2, viral n=1), GVHD (n=1), hemorrhage (n=1) and secondary malignancy (n=1) compared to multi-organ failure (n=5), infection (n=4; viral n=4), and GVHD (n=3) for those >200 ×106/L.
In comparing the incidence of infections between RIC patients, there was no significant difference in the number of early (≤ 42 days post-HCT) or late (days 43 – 100 post-HCT) infections based on the day 42 ALC (Figure 2). In multivariable analysis, after adjusting for gender, age at transplant, total infused cell dose, CFU-GM, infused CD3 and CD34 cell dose, time from initial diagnosis to transplant, single vs. double umbilical cord blood transplant, HLA-match, recipient CMV status and development of acute GVHD (as a time-dependent covariate), an ALC ≤ 200 ×106/L remained an independent predictor of worse survival and PFS (Table 2). Like the MA group, there were no significant differences in the pre-transplant and graft characteristics of RIC patients with day 42 ALC ≤ or >200 ×106/L (Table 4).
There have been a number of improvements made in the field of UCB transplantation over recent years, particularly in defining a critical infused cell dose needed for sustained neutrophil recovery and in developing approaches to overcome this limitation(15, 25–27). Despite the progress in recognizing a critical cell dose for engraftment, TRM and relapse remain significant causes of treatment failure for UCB recipients. TRM is especially distressing since it occurs in patients that might otherwise be cured from their disease. The ability to identify patients at increased risk for early death could lead to interventions directed toward preventing these negative outcomes. This single institution analysis of 360 patients who received UCB transplantation for hematologic malignancies, shows that early rapid lymphocyte recovery, after either myeloablative or RIC transplantation is an independent parameter associated with superior survival.
In this analysis, patients who received either a MA or RIC UCB transplant and had an ALC >200 ×106/L at day 30 or 42 respectively, had superior OS and PFS. Similar findings of improved outcomes for BM and PBSC recipients have been reported (6–11), but this issue has not yet been examined in UCB transplantation. The exact explanation for these findings is unknown, but may reside in the cells that make up the ALC at these early time points. NK cells recover to normal amounts rapidly after transplant (28), however, these cells are functionally impaired (29). Despite this, two prior studies correlating high ALC with improved transplant outcomes have implicated natural killer (NK) cells as being significantly associated with improved outcomes (30–31). Thus, perhaps the numbers of NK cells in the graft or levels of endogenous cytokine (i.e., IL-15) after transplant may differentially influence NK expansion and account for these findings. Alternatively, even though the number of T cells at day 30–60 after transplant are low, reconstitution is actively occurring and rapid recovery of the T cell compartment might also account for these observations. Both thymic independent (homeostatic proliferation) and dependent pathways of T cell reconstitution have been described (reviewed in (32)). The early increases in T cells after transplant are thought to be mainly due to homeostatic T cell proliferation, where a limited repertoire of T cells emerge in response to alloantigen stimulation and high serum cytokine concentrations. Whether these T cells are protective against relapse or infectious organisms remains an open question in humans, but both have been hypothesized (32–33). Perhaps in support of this, Hamza et al. showed that patients with rapid lymphoid recovery after UCBT and MUD BMT had a lower risk of infections (12). In contrast, IFN-γ producing CD8+ CMV reactive T cells could be detected in the blood of UCB recipients early after transplant (week 8), but their appearance did not correlate with CMV control (34). Alternatively, rapid lymphocyte recovery may reflect a more robust thymic dependent T cell production. In support of this Komanduri et al. showed that UCB recipients who had high numbers of CD4+ T cells that co-expressed CCR7+ (found on naïve and central memory cells) at day 30 UCB had superior survival, suggesting that new T cell formation and intact immune responses were critical to survival (13). In our analysis, MA patients with a day 30 ALC ≤ 200 ×106/L had significantly more early infections compared to patients >200 ×106/L, but this was not seen for late infections or for RIC patients (early or late infections) based on their day 42 ALC. Unfortunately, for a significant proportion of patients in our study, we do not have detailed lymphocyte subset data available to determine the composition of the patient’s ALC and this will be addressed in future studies.
Several prior studies that report the benefit of high ALC early after transplant show that the improvement in survival is due to lower relapse rates (6–11). For instance, in studies of patients transplanted with ALL(10), AML(9) or a mixture of hematological diseases(11), a high day 21 ALC was strongly correlated with a reduction in relapse and an improved DFS following matched sibling donor transplant. Similar results were shown by Savani et al. for T cell depleted sibling donor transplants for AML at day 30, but not for ALL (31). These findings by Savani et al., prompted us to perform a similar subgroup analysis (based on disease type) in our myeloablative and RIC patient cohorts. However, for both subgroups of diseases in patients undergoing myeloablative (ALL and AML) or RIC UCB transplantation (leukemia/MDS and lymphoma), we were unable to identify a significant impact of ALC on disease relapse in multivariate analysis (data not shown). In contrast to other reports on early lymphocyte recovery following BM or PBSC transplantation (6–11, 31), our results in UCB transplantation show that a low ALC after MA allo-HCT was associated with significantly higher TRM, with no impact on relapse. The early TRM identified in these patients may have been a competing risk to relapse, thus explaining these results. While it is tempting to speculate that high ALC early after transplant is somehow related to graft versus tumor reactions, the association of ALC on relapse is not limited to the allogeneic graft vs. tumor effect, since similar findings have been found after autologous transplant for non-Hodgkin’s lymphoma(36), Hodgkin’s disease(37), multiple myeloma(38) and Ewing sarcoma(39).
All prior studies investigating ALC on allo-HCT outcomes have been in the myeloablative setting (6–11). The time to lymphocyte recovery in our myeloablative UCB recipients are relatively similar to the studies in BM or PBSC recipients reported by others (6–7, 9–10). This is remarkable considering that UCB recipients receive 10–50× fewer MNCs/kg and that these cells have been cryopreserved and thawed, which negatively affects viability. Thus, these results might suggest that the proliferation kinetics and/or the homing properties of UCB lymphocytes differ compared to BM/PBSC.
Here, we also report the first large series of ALC data in patients undergoing RIC conditioning (n=189). In these patients, day 42 was the first time that ALC was significantly predictive for transplant associated outcomes. Based on our prior studies, the majority of patients show full donor lymphocyte chimerism at this time point (18). Although the kinetics of immune recovery are relatively similar between MA and RIC recipients (40), an explanation for the difference in time (day 30 vs. day 42) between MA and RIC in terms of an ALC of ≤200 x106/L predicting outcome may be the result of residual host cells competing with lymphoid growth factors in RIC patients. Alternatively, after RIC, the niches for stem cells and lymphoid engraftment may still be occupied with residual host cells which might also delay donor lymphocyte recovery.
While TRM remains a significant barrier to the success of UCB transplantation, there have been very few measures to identify patients at risk for this complication. For instance, Sorror and colleagues described the HCT-specific co-morbidity index (41), but such measures have only been partially validated in the UCBT setting, since they are predictive for TRM after myeloablative, but not RIC transplantation in adult recipients (42). Here we show that patients with hematologic malignancies receiving a MA or RIC UCBT and fail to achieve an ALC >200 ×106/L at day 30 or day 42 respectively, have a significantly inferior OS and PFS. These results may allow for interventions aimed at preventing mortality. Considering that the majority of deaths after UCB transplantation are within 180 days and are related to multi-organ failure and/or infectious complications, day 30 or 42 ALC may be useful to identify MA or RIC UCBT recipients at risk for early mortality. The findings presented here also suggest that limiting treatments that delay- or instituting approaches that accelerate lymphocyte recovery might have a significant impact on transplant outcomes. Potential strategies could include improved antimicrobial prophylaxis, or the use of agents that promote the recovery and survival of T cells (i.e., IL-7 (43)) and/or NK cells (i.e., IL-15 (44–45)). Lastly, agents aimed at protecting structures involved in de novo T cell production, including thymic epithelial protectants and/or stimulants have considerable promise (46). Further analyses investigating the impact of ALC following MA and RIC UCBT for hematologic malignancies are warranted to confirm these findings.
This work was supported in part by grants from the National Institutes of Health NCI P01-CA65493 (JEW, BRB, JSM and MRV), American Cancer Society RSG-08-181-LIB (MRV), Leukemia Research Fund (MRV) and Children’s Cancer Research Fund (MRV, MJB and JEW).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Author Contribution statement:Michael J. Burke: Conceived the study and writing of manuscript
Rachel Isaksson Vogel: Statistical analysis
Sanyukta K Janardan: data collection
Claudio G. Brunstein: Protocol development, reviewed data and writing of manuscript
Angela R. Smith: Protocol development and writing of manuscript
Jeffery S. Miller: Protocol development and writing of manuscript
Daniel J. Weisdorf: Protocol development and writing of manuscript
John E. Wagner: Protocol development and writing of manuscript
Michael R. Verneris: Conceived the study and writing of manuscript