|Home | About | Journals | Submit | Contact Us | Français|
Acute lymphoblastic leukemia (ALL) with high-risk features has a poor prognosis in adults despite aggressive chemotherapy. Reduced-intensity conditioning (RIC) is a lower toxicity alternative for high-risk patients requiring hematopoietic cell transplantation (HCT), however it has not been widely used for ALL. We conducted a retrospective study of 24 high-risk adult ALL patients who received an RIC regimen of fludarabine/melphalan prior to allogeneic peripheral blood stem cell transplant between 6/14/02 and 6/15/07 at City of Hope. Indications for the RIC regimen were: 1) age 50 or older (42%), 2) compromised organ function (54%), or 3) recipient of a previous HCT (37.5%). Patients had a median age of 47.5 years and the median follow-up was 28.5 months for living patients. Both overall survival and disease-free survival at two years was 61.5%. Relapse incidence was 21.1% and non-relapse mortality was 21.5% at two years. cGVHD developed in 86% of evaluable patients. In this series, no significant correlations were made between outcomes and patient age, presence of Philadelphia chromosome, relatedness of donor source or prior HCT. These high survival rates for high-risk ALL patients following RIC HCT may offer a promising option for patients not eligible for a standard myeloablative transplant.
Acute lymphoblastic leukemia has a poor prognosis in adult patients, with a five-year overall survival (OS) rate of 39-50% despite aggressive chemotherapy [1-3] and only 15% for patients over 50 years of age . In patients with high-risk disease, as determined by age, cytogenetics, remission status, and/or response to induction therapy, survival outcomes are even worse [1-3]. According to the MRC/ECOG ALL Trial of chemotherapy versus autologous and allogeneic transplant, allogeneic hematopoietic cell transplantation confers the greatest durable benefit for standard risk adult patients and is more effective than either chemotherapy or autologous transplant . Goldstone et al. also show, however, that for patients over 45 and others with high-risk ALL, a high non-relapse mortality (NRM) of 36% offsets any potential survival advantage of the reduced relapse rate conferred by myeloablative transplant . Patients requiring transplants who are over age 50, have impaired organ function, or have had previous ablative therapy are unable to withstand the toxicity of the myeloablative protocols that are standard of care.
In the last ten years, the goal of reducing transplant related mortality has led to the investigation of a variety of reduced-intensity, non-myeloablative conditioning protocols for allogeneic HCT in patients with multiple hematologic malignancies[5-8]. In a study comparing RIC to standard HCT, NRM (22% vs. 30%) and overall survival (OS) (59% vs. 52%) were comparable . Similar to observations in the myeloablative transplant setting, Mohty et al. describe a strong association of chronic graft-versus-host disease (GVHD) with a reduction in relapse incidence from 55% to 30% , suggesting a primary role for graft-versus-leukemia (GVL) in the achievement of remission via RIC HCT.
Four small (22–33 patients) prospective studies [9-12] and a larger (97 patients) retrospective study  have attempted to assess the feasibility and effectiveness of RIC specifically for treatment of ALL in high-risk populations. The two-year overall survival rates for these studies average 32% (median 31%, range 18-50%) , with variable, but high, relapse rates and transplant-related mortalities, depending upon remission status. Consensus conclusions from these studies appear to be that improved survival is associated with RIC transplant during first complete remission (CR1) and that relapse incidence (RI) is lower for patients exhibiting GVHD.
In most published studies of RIC in ALL [9-11, 13], multiple sources of hematopoietic stem cells and pre-transplant conditioning regimens were included in the same data set, making it difficult to obtain the best and most consistent results. In this study, we have removed the issue of transplant regimen variation; all patients were treated at City of Hope with peripheral blood stem cell transplant (PBSCT) following the same reduced intensity conditioning regimen. Based on this more homogeneous treatment and stem cell source, we report a high survival rate among high-risk ALL patients receiving RIC.
A retrospective analysis was conducted on 24 high-risk ALL patients treated between 6/14/02 and 6/15/07 with a uniform reduced-intensity stem cell transplant protocol at the City of Hope. The indications for the reduced intensity regimen were one or more of the following: 1) patient aged 50 or greater (42%), 2) compromised organ function (54%), or 3) recipient of a previous HCT (37.5%). Indications for HCT during CR1 included age over 35, high white blood cell count (> 50,000) at diagnosis, multiple rounds of induction chemotherapy required to achieve remission, and/or poor prognosis cytogenetics (e.g. Philadelphia chromosome or t(4;11)(q21;q23). Cytogenetic risk level, based on Pullarkat et al. , is indicated in Table 1 as Level 2, 3, 4 or Ph+ (in increasing order of severity). The City of Hope Institutional Review Board approved the retrospective study and analysis of this patient case series.
Salient patient characteristics for all 24 patients are displayed in Table 1. Fifteen women and nine men were part of the study with a median age of 47.5 years (range 23-68). Seven patients (29%) had a previous allogeneic HCT and two (8%) had a previous autologous HCT. In addition to the high-risk factors for which they were included in this study, the patient population exhibited additional risk factors that could affect outcome: 67% of the peripheral blood stem cell donors were unrelated to the recipients, 54% of patients were beyond CR1 at the time of transplant and 42% were Ph+. Three patients had ALL that was secondary to a prior malignancy: patient #1 had prior multiple myeloma, #4 breast cancer and #10 had both a history of chronic lymphocytic leukemia and germ cell cancer. One patient, #9, also had a coexisting myelodysplastic syndrome at the time of transplant.
Listed first in Table 1 are the ten patients in CR1 (42%), followed by one CR1 patient (#11) who was not in molecular remission (4%), four patients in second complete remission (CR2, 17%), one induction failure (4%), three in first relapse (R1, 12%), one in second relapse (R2, 4%) and four in third complete remission or beyond (≥CR3, 17%). For patients in CR1, the median time between remission and transplant was 2.3 months (range 0.5-5.7), 40% required two courses of chemotherapy to achieve remission and the median WBC at diagnosis was 21.5 (range 0.6-70). Patients not in remission at the time of transplant had a median WBC prior to conditioning of 3.3 (range 0.9-7.8), a median of 1% blasts in the bone marrow (range 0-15) and a median of 0% blasts in the blood (range 0-21).
Matched sibling donors were available for 8 of the 24 (33%) patients in the study. For the 16 patients lacking suitable related donors, HLA MUDs were successfully identified through the National Marrow Donor Program. HLA typing was performed using polymerase chain reaction (PCR) sequence-specific primers or PCR sequence-specific oligonucleotide probe techniques. Of the MUD transplants, mismatches were as follows: Ten patients were matched at 10 of 10 loci, three patients had 2 antigen mismatches, one patient was allele mismatched and one patient had 2 allele mismatches and an antigen mismatch.
The reduced-intensity conditioning regimen for all patients consisted of intravenous fludarabine at 25 mg/m2 daily for five days followed by intravenous melphalan at 140 mg/m2 for one day. Subsequently patients received allogeneic peripheral blood stem cells (PBSC) mobilized with granulocyte colony stimulating factor (GCSF). Eight of the ten Ph+ patients were treated with thymidine kinase inhibitors (TKI) pre and/or post transplant. Table 2 shows details of treatment for Ph+ patients, including type and duration of TKI therapy.
The GVHD prophylactic treatment regimen and the maximal extent of acute (aGVHD) and chronic (cGVHD) manifestations are listed in Table 1 for each patient. Prophylactic treatment of GVHD varied between patients depending on their HLA match and the available protocols at the time of transplant and included: cyclosporine (CSA) plus mycophenylate mofetil (MMF) in 2 patients, CSA/MMF plus methotrexate (MTX) in 8 patients, CSA/MMF plus anti-thymocyte globulin (ATG) in 1 patient, tacrolimus plus sirolimus in 7 patients, and tacrolimus/sirolimus plus MTX in 6 patients. No patient in this study group was treated using DLI.
Survival estimates were calculated based on the product-limit method, and 95% confidence intervals were calculated using the logit transformation and the Greenwood variance estimate . Differences between survival curves were assessed by the log rank test. The significance of demographic/treatment features collected from HCT recipients was assessed using survival analysis and univariate Cox regression analysis . The list of features included was determined from a literature review that identified factors found to be associated with survival and disease relapse in patients treated with allogeneic RIC HCT. Statistical significance was defined at the P value less than or equal to .05 level.
All twenty-four patients engrafted successfully. The median time, post-engraftment, to reach an absolute neutrophil count (ANC) of ≥ 500, was 15 days (range 10-26), and a platelet count ≥ 20,000 was 16.5 days (range 8-24). At day 30 post-engraftment, short tandem repeat (STR) analysis of bone marrow showed a median of 100% donor cells (range 90-100%).
Eighteen patients (75%) developed acute GVHD, which was graded according to consensus criteria . Three patients displayed grade I aGVHD, 15 had grade II-IV, with 5 of those suffering grade III or IV. Of the 21 patients evaluable 100 days post-engraftment, 18 developed chronic GVHD (86%); 13 were classified with extensive disease and 5 with limited cGVHD. Three patients did not develop any symptoms of cGVHD. The three patients not evaluable for cGVHD died prior to 100 days post-engraftment. We also noted that 9 of the patients had cGVHD with acute features: 4 cases developed progressively from aGVHD, 3 were acute/chronic overlap syndromes and 2 cases were delayed acute onset (after day100). Only 1 patient exhibited neither acute (grade II or above) nor chronic forms of graft-versus-host disease (4%).
Relapse and death events are reported as days post HCT and cause of death is listed in Table 1 in the outcome column. Fifteen of the 24 patients were living and disease-free at analysis date, one of whom had relapsed (patient #4), but was disease-free one year after a second RIC transplant. Karnofsky Performance Status (KPS) scores  for survivors are listed in the last column, with all patients at 80% or above and able to carry on relatively normal activity. The KPS score for 1 surviving patient was unobtainable. Of the 13 surviving patients with cGVHD, at the date of analysis, 6 patients had active disease and remained on immunosuppressant therapy, 5 patients had inactive disease and were tapering medications (3 tacrolimus only), and 3 patients were completely off cGVHD medications. Duration of TKI therapy for Ph+ patients is shown in Table 2. TKI therapy was discontinued due to side effects, relapse or with cessation of immunosuppressive therapy.
Median follow-up for living patients was 28.5 months (range 12.8–72.5). Overall survival and disease-free survival at two years were both 61.5% with a confidence interval (CI) of 48.1% to 72.5%. Survival curves for OS and DFS are displayed in Figures 1A and 1B, respectively. Figure 1B also shows the Relapse Incidence curve, with RI at 14.6% (CI of 5.8-33.9%) at one year and 21.1% (CI of 10.2-40.7%) at two years. NRM was 12.5% (CI of 4.8-30.2%) at day 100, 21.5% (CI of 11.4-38.3%) at one year and at two years and is charted in Figure 1A. For NRM causes of death in individual patients, see Table 1.
The Log Rank test was applied to these data to determine whether there were any significant correlations between OS, DFS, RI or NRM and known variables in the patient population. In contrast to other studies in the literature, at the date of analysis there were no significant relationships found. In particular, presence of Ph+ did not affect outcome (60% alive at time of analysis, see Table 2), nor did patient age ≥ 47 (75% living). Use of an unrelated donor and prior HCT also were not significant factors for outcome with 56% of MUD recipients and 44% of prior HCT recipients alive at the time of analysis. Comparison of patients in CR1 versus those beyond CR1 also yielded no significant differences with respect to outcome and the OS and RI curves are shown in Figure 2. Although some studies have shown a correlation between outcome and the occurrence of cGVHD, no meaningful tests could be performed comparing groups with and without symptoms of cGVHD, as the overwhelming majority of patients exhibited some level of GVHD. Of the 18 patients with cGVHD, 9 had some acute features, either progressive development, overlap syndrome or delayed acute onset. There was no significant difference in outcomes for the 9 patients with acute features compared to the rest of the population.
ALL is relatively rare in adults and its incidence increases dramatically with age. The cumulative incidence of ALL cases (per 100,000 of age-matched population) triples in the over-50 age group compared to the 25- 49 age group . In the over-50 group. Additionally, the older patient population has a higher incidence of Ph+, exemplified by the SWOG study in which the median age is 32 for the total study group and 47 for Ph+ patients . Older patients have a very poor prognosis without HCT, but are generally ineligible for myeloablative transplants due to high NRM as demonstrated in the MRC/ECOG ALL Trial . Goldstone et al. conclude based on their results in older patients, that research on RIC transplant regimens in ALL is imperative to reduce morbidity and offer a viable alternative to high-risk patients . RIC has the potential to extend the benefit of transplant to those older patients for whom standard chemotherapy is not effective long term.
Based on the limited ability of donor lymphocyte infusion (DLI) to produce a complete remission in relapsed ALL (only 18% as opposed to 60% for CML) , some believe that a GVL effect is not clinically important in ALL. Rowe et al. have suggested that the lack of response to DLI may be related to the fact that the DLI study involved patients in active relapse while transplant patients are generally in remission, allowing for a more effective allogeneic response to tumor cells . Several transplant studies do in fact suggest that there is a therapeutic GVL effect in ALL, based on a correlation between GVHD and a decrease in relapse incidence among ALL patients, following both myeloablative [22, 23] and RIC [7, 11] allogeneic HCT. The disparity between DLI-associated and transplant-associated GVL effects may also be due to differences in the ability of ALL blasts to present target antigens, the frequency of minor-antigen reactive T-cell precursors, cell cycle kinetics, or susceptibility to lysis .
The development of RIC for ALL would give high-risk patients transplant options and hope for disease control with fewer complications. Several studies have attempted to assess the efficacy of RIC for treatment of high-risk ALL, however, most were small and the patient populations and treatment regimens were heterogeneous. In this study at City of Hope, all patients received a fludarabine/melphalan conditioning regimen and the stem cell source was PBSC. Despite the poor prognostic indicators and limited sample size, the 2-year OS and DFS of 61.5% for this patient population are encouraging. In comparison, the recent MRC/ECOG ALL Trial shows that standard risk ALL patients receiving allogeneic transplants have a 5-yr overall survival of 62%, while the Ph− high-risk trial patients have an OS of 41% and Ph+ patient OS is only 22% .
The 21.5% non-relapse mortality at 2 years compares very favorably with the International ALL Trial 2-yr NRM of 35.8% for high-risk myeloablative transplant patients. The 2-yr relapse rate of 21.1% was also similar to the 2-yr rate of 25% from the International ALL Trial (interpolated from 10-year graph) . At 100 days NRM was 12.5%, however, an increase to 21.5% by 2 years post-transplant was attributable to cGVHD and associated infections.
Nearly all patients had GVHD to some extent (23 of 24 patients) and PBSCT is known to correlate with higher risk of GVHD . The incidence of cGVHD was a cause of concern, however in surviving patients, KPS scores were 80-100%, allowing a reasonably functional life after transplant. Jagasia finds that patients having cGVHD with any acute features have a poorer prognosis. In this group, half of the cGVHD cases had some acute features, only 2 of which were delayed acute presentation. There was no correlation between cGVHD with acute features and poor outcome. The low relapse rates in this study may be partially attributable to high GVHD, however direct statistical tests were inconclusive due to patients numbers.
This study is consistent_with the recent findings of Bachanova et al. using cord blood for ALL RIC, whose 22-patient study is of similar size, high-risk patient composition and outcome . Both studies have a relatively low incidence of relapse (Stein 2-yr: 21%, Bachanova 3-yr: 36%), especially for first remission patients after the second year. Although the stem cell sources and treatment details differ, both studies show significant improvement in OS compared to previously published data (Stein 2-yr: 61%, Bachanova 3-yr: 50%).
Analysis of this data from 24 high-risk ALL patients did not find a significant association between age and survival, however that is not unexpected given the skewed age distribution and sample size. While younger patients might be expected to demonstrate significantly better outcomes in a larger, more age-diverse study, we demonstrate highly acceptable survival rates for older patients receiving RIC. Patients too old to be candidates for myeloablative therapies may now be considered reasonable candidates for reduced-intensity PBSCT transplants. In this study, unrelated donor transplant patients had survival rates comparable to patients with related donors. A similar finding for unrelated donors is also recently reported for elderly patients in a large AML transplant study (368 patients)  and in the recent cord blood ALL study . For older patients, MUD transplant survivability is an important consideration, as the availability of eligible sibling donors diminishes with patient age.
Despite the high-risk nature of the patients in this study, all had low disease burden at the time of transplant, either in a complete remission or with low WBC and % blasts. Forty-two percent of patients were in CR1, however no significant difference in outcome was detected for these patients compared to those beyond CR1. We are aware that this result is in contrast to the published literature and attribute this to the small sample size.
For adult Philadelphia chromosome positive ALL patients, a safer alternative to standard allo-HSCT is especially important. Historically, this population had a dismal survival rate of 10-20% when treated with standard chemotherapy alone [1, 14, 27]. Use of imatinib mesylate and other tyrosine-kinase inhibitors is improving remission rates and durable responses, allowing for improved transplant  and non-transplant  survival. Despite this fact, allo-HCT is considered the only curative approach and is the standard of care in first complete remission for Ph+ ALL patients [27, 28], . Even though the presence of the Philadelphia chromosome was not part of the selection criteria for RIC in this study, 42% of patients were Ph+, reflecting the increased incidence of Philadelphia chromosome in older patients. Of the 10 Ph+ patients in this study, 5 have survived in remission, a DFS of 50% that is comparable to the DFS of Ph+ patients seen in a collaborative study by this group and Stanford (Ph+ DFS in CR1=48%, beyond CR1=26%) . Due to increasing prevalence of Ph+ with increasing age [28, 32], the availability of effective reduced-intensity transplants is a crucial addition to the treatment armamentarium for these patients.
In conclusion, this retrospective study demonstrates optimistic survival rates for patients with high-risk ALL undergoing RIC, comparable to those seen in the literature for standard-risk patients undergoing myeloablative allo-HCT. These data also contribute to the body of evidence supporting a role for GVL in the treatment of ALL. Studies describing RIC transplant for ALL are scarce in the literature and there are as yet, no single-protocol prospective trials. While our findings are exciting, the caveats of the retrospective nature of the study and its small sample size can be addressed only through future prospective trials in adults with ALL. In order to validate these findings and those of other studies, we plan to use this preliminary data as the basis of a proposal for a large phase II prospective clinical trial of reduced intensity transplant in patients with high-risk ALL.
This study was supported in part by grants from the National Institute of Health, CA30206 and CA33572. We would like to thank Karen Chang for hematopathology and Sean Wang for stem cell collection. We would like to acknowledge the dedicated nurses of the City of Hope Bone Marrow Transplant Unit for excellent care of our patients and support of medical research.
Conflict-of-Interest Disclosure: The authors declare no competing financial interests.
Contribution: A.S.S. designed the original study, performed the research, analyzed the data and wrote the paper; J.M.P. and M.L.S. analyzed the data and critically reviewed the paper, N.M.K., R.T.S., N.C.T., K.C. S.W. and D.S. generated data, M.R.D. and D.S.S. critically reviewed the paper, S.H.T. analyzed the data and wrote the paper, and S.J.F. designed the original study and critically reviewed the paper.