|Home | About | Journals | Submit | Contact Us | Français|
Reduced-intensity conditioning (RIC) regimens in cord blood transplant (CBT) are increasingly utilized for older patients and those with comorbidities. However, the optimal conditioning regimen has not yet been established and remains a significant challenge of this therapeutic approach. Anti-thymocyte globulin (ATG) has been incorporated into conditioning regimens in order to decrease the risk of graft failure; however, use of ATG is often associated with infusion reactions and risk of post-transplant complications. We report the results of a non-ATG-containing RIC regimen, where patients received 2Gy TBI unless they were considered to be at higher risk of graft failure, in which case they received 3Gy of TBI. Thirty patients underwent CBT using this protocol for high-risk hematological malignancies. There was only one case of secondary and no cases of primary graft failure. At one year, estimates of NRM, OS and PFS were 29%, 53% and 45%, respectively. The cumulative incidences of grade III–IV acute and chronic GVHD were 14% and 18%, respectively. In summary, results of this study demonstrate that this non-ATG-containing conditioning regimen provides a low incidence of graft failure without increasing regimen-related toxicity.
For patients with hematologic malignancies, hematopoietic cell transplantation (HCT) is often the best therapeutic approach to achieve a cure . However, adults receiving conventional high-dose myeloablative conditioning regimens experience high rates of non-relapse mortality (NRM) with this approach. Furthermore, many patients in need of HCT do not have an available HLA-matched related or unrelated donor and require an alternative donor transplant [2,3].
Cord blood transplant (CBT) using reduced-intensity conditioning (RIC) regimens has emerged as an effective therapy for patients with advanced or high-risk hematological malignancies who cannot tolerate a conventional myeloablative conditioning regimen, and who lack a suitable related or unrelated donor. Because donor-recipient HLA-antigen disparity appears to be more tolerable in CBT than with other sources of stem cells, cord blood (CB) extends the donor pool for those who need a HCT .
However, despite the wider use of RIC regimens in CBT, the optimal conditioning regimen for RIC CBT has not yet been established and graft failure remains one of the greatest challenges of this therapeutic approach [5,6]. One approach to decrease the risk of primary graft failure has been the incorporation of anti-thymocyte globulin (ATG) into the conditioning regimens to facilitate engraftment [7,8]. However, there are several toxicities associated with the use of ATG, including an increased risk of EBV-driven post-transplant lymphoproliferative disorder (PTLD) in addition to its deleterious impact in the graft-versus-tumor effect and delay in immune-reconstitution [9–11].
Due to the potential toxicities associated with the use of ATG in RIC regimens, we investigated the use of a non-ATG-containing RIC regimen, with the goal of reducing ATG-based risks while not compromising sustained donor engraftment. Herein, we report our findings with this regimen, including the rate of engraftment, survival, relapse and incidence of graft-versus-host disease (GVHD) in adult patients with advanced or high-risk hematologic malignancies who were not eligible for conventional donor myeloablative HCT.
This protocol was reviewed and approved by the Institutional Review Board of Fred Hutchinson Cancer Research Center (FHCRC) and other participating institutions. Written informed consent was obtained from all patients before the enrollment and participation according to the Declaration of Helsinki. Patients with advanced or high-risk hematologic disease were eligible for CBT if they had no related donor matches at 5–6/6 human leukocyte antigen (HLA) loci (A, B, DRB1) or 10/10 matched unrelated donor, or if an unrelated donor was not available within a reasonable time frame, based on the patient’s disease.
Cord blood (CB) donors were selected on the basis of HLA-match and cell dose. The total nucleated cell (TNC) dose of the combined units had to contain at least 3.0×107 TNC per kilogram recipient weight with a minimum pre-freeze cell dose of 1.5×107 TNC per kilogram per unit. All CB units were required to be at least 4/6 HLA-matched to the recipient with no matching requirement between the two units. HLA matching was at the intermediate resolution for HLA-A and –B and at the allele level for HLA-DRB1. All patients underwent a double CBT (dCBT) and the CB units were infused sequentially on day 0.
All patients received fludarabine (40mg/m2/day; total dose 200mg/m2) from day −6 to day −2, and a single dose of cyclophosphamide (50mg/kg intraveneously) on day −6. Patients who had received a prior autologous transplant within 12 months or ≥ 2 cycles of multiagent chemotherapy with at least one cycle of therapy within the 3 months previous to CBT received a single fraction of 200cGy TBI on day −1. Otherwise, patients were considered at higher risk for primary graft failure and received a higher single-fraction dose of 300cGy of TBI on day −1. In the higher risk population, rules were also in place that allowed for an increase to the dose of TBI (to 400cGy or 450cGy) if the observed proportion of graft failure was consistent with a true rate in excess of 15%.
GVHD prophylaxis began on day −3 and consisted of cyclosporine A (CSA) and mycophenolate mofetil (MMF). CSA was continued for a minimum of 100 days and then tapered 10% per week starting on day +101 and discontinued no sooner than 6 months post-transplant in the absence of GVHD. MMF was administered at 1 gram every 8 hours until day 40 post-transplant and then in the absence of GVHD, tapered by 12 grams per week and discontinued after day + 96.
Supportive care was standard per each institution. At our Center (FHCRC), we adopted an intensive strategy to prevent CMV disease in seropositive CBT recipients .
Granulocyte-colony stimulating factor (G-CSF; 5µg/kg per day) from day 0 was administered to all patients until the ANC was greater than 2,500/µl for two consecutive days. The day of neutrophil engraftment was defined as the first of 3 consecutive days of an absolute neutrophil count of 500/µl or greater. Platelet engraftment was defined as the first of 7 consecutive days when the platelet count exceeded 50 × 106/µl (untransfused). Graft failure was defined as lack of neutrophil engraftment beyond day 42 after the infusion of CB stem cells.
Disease status was assessed on days 28, 80 and one year after transplantation (unless clinically indicated at other times). Host/donor chimerism studies were performed with restriction-length fragment polymorphism (RFLP) assays on both the bone marrow and peripheral blood. Whole (unsorted) bone marrow chimerism was performed on days 28 and 80 after transplantation. Patients were defined as complete donor chimerism if they had more than 95% donor cells. Single-donor dominance was defined as more than 90% single-unit contribution of CD3+ cells combined with at least 70% single-unit contribution of both CD56+ and CD33+ fractions by day 28 after transplantation.
The primary objective of this protocol was to estimate the probability of overall survival at one year, this estimate to be used as a benchmark for comparison for future trials. A further objective was to assess the probability of graft failure with the higher dose of TBI (300 cGy) among patients at higher risk of failure, with rules built in to allow for further escalation of the TBI dose in the event of excess graft failure. Probabilities of overall and PFS were estimated using the method of Kaplan and Meier . For analysis of survival, death from any cause was considered an event and for PFS, relapse, progression or death were considered events. Cumulative incidence estimates were used to summarize the probabilities of neutrophil and platelet recovery, acute and chronic GVHD, non-relapsed mortality (NRM), and relapse. For neutrophil and platelet recovery and GVHD, death without the event was considered a competing risk. For NRM, relapse was considered a competing risk, and for relapse, death without relapse was a competing risk. All outcomes were assessed among all patients (regardless of the dose of TBI).
Thirty consecutive patients underwent dCBT per this protocol between February 2006 and January 2011 (FHCRC, N= 17; Rocky Mountain Cancer Center, N=5; University of Colorado, N=7; Intermountain Blood and Marrow Transplant, N=1). Patient characteristics are summarized in table 1. Disease status included high-risk or advanced hematological malignancies. Nine patients (30%) had failed prior HCT due to relapsed disease, 5 after autologous stem cell transplant and 3 after allogeneic transplant. The ninth patient, with Hodgkins disease, had relapsed disease following an autologous stem cell transplant and then graft failure after undergoing a salvage haploidentical HCT. The median number of prior treatments and comorbidity score were 3 (range, 1–8) and 3 (range, 1–8), respectively . Minimal residual disease (MRD) was established based on the presence of anormal immunophenotype on flow cytometry. MRD was present at 0.2% to 0.4% in 9 (30%) of the patients. Graft characteristics are described in table 1. Notably, all patients received two CB units to achieve the required cryopreserved cell dose.
Twenty-seven of 30 patients had neutrophil recovery, the median time to this event being 17 days (range, 6–40 days; Figure 1A). Of the three patients who did not reach this event, two died from multiorgan failure at day 12 and 42 and the other one died on day 63 from necrotizing hemorrhagic meningitis after he relapsed on day 12. Two of these patients were considered to be at higher risk of graft failure and therefore received 300cGy TBI (those who died without recovery on days 12 and 63). There were no cases of primary graft failure, and only one case of secondary graft failure. This patient was diagnosed with MDS and had received only one cycle of chemotherapy prior to CBT. The patient received the higher dose of TBI (300 cGy) and died from multi-system organ failure and intracranial hemorrhage at day 54 post-transplant without recovery of platelet counts. Chimerism studies were consistent with secondary graft failure. Per the rules specified in the protocol for patients at higher risk of graft failure, the observed rate of graft failure was consistent with a true rate of less than 15%, and therefore the dose of TBI was not increased above 300cGy for such patients. Twenty (67%) of 30 patients had evidence of platelet recovery (Figure 1B) in the first 100 days post HCT, the median time to recovery being 55 days (range, 23 to 183 days). Each of the ten without evidence of platelet engraftment died, days of death ranging from 12 to 381 (median 63 days).
Patients were considered evaluable for donor chimerism if they were in remission at the time of evaluation. Assessment of whole bone marrow chimerism revealed that complete donor chimerism was achieved in 43% of the evaluable patients on day + 28 and 55% of the patients on day + 80. Analysis of donor/host peripheral blood chimerism was available in 22, 19 and 16 patients on days +28, +56 and +80, respectively. Single-donor dominance was achieved on 59% of the patients on day 28, 72% on day 56 and 75% on day 80 after transplant.
Seventeen of 28 (61%) patients who received a GVHD grade developed grades II–IV acute GVHD, with only 4 (14%) developing grade III (Figure 2A). There were no cases of grade IV acute GVHD. There were several patients with limited follow-up, specifically to a time before some cases of chronic were observed, and this resulted in an estimated probability of chronic GVHD at 2 years of 18% (Figure 2B). Of note, all cases of chronic GVHD were limited, except for one patient who developed extensive chronic GHVD and died of sepsis complications. Eight patients died before day 100. The conditioning regimen was well tolerated. There were 11 cases of NRM, with an estimated probability of 17% at 100 days, 20% at 180 days, and 29% at one year (Figure 3). Infectious complications are summarized in table 2. The majority of the virus infections were due to BK viruria. Of note, there were no cases of CMV disease, EBV reactivation or PTLD in this cohort of patients.
By last follow-up, there have been a total of 17 deaths. The median follow-up among the 13 survivors was 344 days (range, 114 to 1303 days). The estimated overall survival at one year was 53% (Figure 3). To date, 8 patients relapsed, with day to relapse ranging from 20 to 449 days (median day to relapse, day 70). All but two patients who relapsed eventually died from their disease. One-year PFS is 45% (Figure 3). Causes of death are summarized in Table 3.
Multiple centers have reported studies of RIC CBT [5–9,15–20]. The largest series reported to date by the Minnesota group included 110 patients with various hematological malignancies . In this study, ATG was added to the conditioning regimen to decrease the incidence of graft failure in high-risk patients. Although the incidence of graft failure was lower than in their previous series , ATG was associated with a higher risk of NRM at 180 days (38% vs 12% in the non-ATG group, p=0.02) in a univariate analysis.
There are multiple reasons why ATG may pose a particular hazard to recipients of CBT. First, ATG is well documented to delay T-cell recovery, which is an undesirable effect in these patients who are known to experience delayed immune-reconstitution even in the absence of ATG [10,21]. Second, ATG has been associated with an increased risk of EBV PTLD [8,21,22]. Third, ATG might abrogate the nascent graft-versus-leukemia effect . Finally, ATG is associated with high rates of infusion-associated reactions and serum sickness .
In our own experience (unpublished data), ATG-based RIC CBT showed a very high rate of CMV reactivation and an increased risk of other viral infections. We also had one death associated with the infusion of this drug. Based on this experience, we investigated a non-ATG-based conditioning regimen for patients undergoing RIC CBT to evaluate whether the use of a higher dose of TBI (300cGy as opposed to 200cGy) would decrease the toxicity rate associated with ATG in those patients with no prior therapy without compromising the rate and tempo of engraftment.
In this study, we observed a similar incidence of NRM as compared to ATG-based conditioning regimens, rapid neutrophil engraftment, no cases of primary graft failure, and only one case of secondary graft failure [8,9]. Furthermore, the survival rate is similar to that reported with conventional donor hematopoietic stem cells [25–27]. Most of the deaths were due to relapse, which is not surprising in the setting of RIC, as such patients typically present with very high-risk disease and approximately 30% of the population has failed prior transplants due to disease relapse.
Similar to other published reports of RIC CBT, we observed a transient mixed (i.e., presence of host cells) and double chimerism (i.e., presence of cells from each donor unit) at early time points after transplantation [6,7,17]. Although some series reported complete chimeras by day 100 , lack of single-unit dominance is not uncommon after RIC CBT [8,17], however the clinical significance of this remains unknown.
Our post-grafting immunosuppression is modified from the Minnesota’s regimen . Both regimens include CSA from day −3 through day +100, then tapered to be discontinued no sooner than 6 months post-transplant. However, in contrast to Minnesota, which has reported on using MMF only twice daily from day −3 to day +30, our regimen included MMF one gram three times daily from day −3 to day +30. Our dose and schedule of MMF was based on findings of studies from our institution showing that increased MMF dosing from 15mg/kg twice daily to 3 times daily resulted in reduced incidence of both graft failure and acute GVHD .
With respect to GVHD, our incidence of acute GVHD was similar to that reported by others [7–10]. As is usual in CBT, nearly all CB units were mismatched at 1 or 2 HLA alleles, yet the incidence of GVHD is similar to that seen from conventional donor sources that are more highly matched. Most cases were due to grade II GVHD. Moreover, although there were 4 cases of grade III GVHD, only one patient died due to complications from GVHD. Again, similar to others, only a small proportion of patients developed chronic GVHD and there were no cases of severe chronic GVHD, although we acknowledge that follow-up is limited in some patients and some of these could still develop chronic GVHD in the future. Thus, removing ATG did not appear to negatively impact the incidence of GVHD. Lower incidence of acute GVHD has been shown among patients who received ATG and although a higher incidence of acute GVHD could have been expected in our non-ATG based conditioning regimen, it is likely that our more prolonged use of MMF counterbalanced the effects of the removal of ATG from our conditioning regimen .
In summary, the results of this study suggest that this non-ATG-based regimen is safe and the removal of ATG did not negatively impact the incidence of graft failure or GVHD. The uniform use of 2 partially matched CB units in all transplants, the post-transplant immunosuppression practiced at our institution and the use of a higher dose of TBI (300cGy as opposed to 200cGy) in patients who are at high-risk for graft failure likely contributed to enhanced engraftment rates. Nonetheless, the small size of our study and its heterogeneous population prevents us from drawing any definitive conclusion and larger studies with longer follow-up are needed to confirm the significance of these findings. Future directions for optimization of RIC CBT should include improvement in the post-transplant immunosuppression to minimize graft failure and NRM, as well as to improve immune reconstitution. Strategies to prevent post-transplant relapse are also needed.
We are grateful to the patients and families who consented to the use of clinical research results and biologic specimens in these trials. We thank Denise Ziegler, Ivy Riffkin, MaryJoy Lopez, and Adrienne Papermaster for their assistance in the preparation of this manuscript.
This work was supported by the National Institutes of Health (grants K23 HL077446, C.D.; ALC CA18029 and CA 78902). C.D. is a Damon Runyon Clinical Investigator supported in part by the Damon Runyon Cancer Research Foundation (CI# 35-07). F.M. is a recipient of a Research Fellowship from the Fondazione Internazionale di Ricerca in Medicina Sperimentale, Torino, Italy sponsored by “Provincia di Benevento”.
Presentation of material in submitted manuscript: Presented in part at the 53nd Annual American Society of Hematology Conference in San Diego, Ca, December 10–13, 2011.
Authorship ContributionsF.O., F.M.,and C.D. participated in the study design, data analysis and interpretation of data for the manuscript. F.O. wrote the first draft and F.M., T.A.G., J. G., P. M., P. F., B. S., R. S. and C.D. provided revisions and critical review of the final manuscript. T.A.G. and F.M. performed the statistical analyses.
Conflicts of Interest
All authors declare no conflicts of interest.