Disease burden may identify patients more likely to benefit from second allogeneic hematopoietic stem cell transplantation to treat relapsed acute myelogenous leukemia


The major cause of failure after allogeneic hematopoietic stem cell transplantation (HSCT) for acute myelogenous leukemia (AML) is disease relapse or progression. We analyzed the outcome of second HSCT for treatment of patients with relapsed, refractory AML/myelodysplastic syndrome (MDS) at our institution. A total of 72 patients were eligible for this analysis. In all, 25 (35%) patients received salvage chemotherapy prior to the second transplant procedure and only two (3%) patients were in complete remission at the time of the second transplant. A total of 20 patients (28%) had low leukemia burden as measured by the absence of peripheral blood blasts and 5% blasts in the bone marrow at the time of the second transplant. Although, the overall median survival after the second transplant was 6 months, a subset of patients who had low leukemia burden at the time of the second transplant had a 5-year survival of 25 vs 12% in those with a high leukemia burden. Thus, a second transplant may offer the possibility of long-term disease control in a subset of patients who have a ‘low bulk’ disease at the time of transplantation.

In adults with acute leukemia, disease relapse remains the major cause of treatment failure after allogeneic hematopoietic stem cell transplantation (HSCT).1 These patients generally have a poor prognosis, with a median survival of 3–4 months without active treatment.2, 3 Many of these patients receive donor leukocyte infusions (DLI) to induce a graft-versus-leukemia effect (GVL).2, 3, 4 While DLI produces responses in chronic myelogenous leukemia (CML), patients with acute leukemia rarely benefit from this strategy. Some patients with a good performance status may be considered for a second allogeneic HSCT. Although the treatment-related mortality (TRM) after the second HSCT can be substantial (2-year nonrelapse mortality of 51% in the study by Radich et al),5 some studies indicate that the leukemia-free survival (LFS) after the second allogeneic HSCT may be as high as 52% at 3 years in a subset of patients who relapse late after the first HSCT and are in a complete remission (CR) at the time of the second transplant.6 However, most studies are limited by the small number of patients and include a variety of diagnoses from CML-chronic phase to acute leukemias (both myelogenous and lymphoid). Most studies also included both adults and children. There is no consensus on which conditioning regimen to use, the source of stem cells, and whether patients should undergo further chemotherapy with a goal of attaining a CR before undergoing the second transplant.

In an attempt to identify factors that may help to determine which patients would benefit from a second HSCT, we reviewed outcome in adults with refractory acute myelogenous leukemia (AML), myelodysplastic syndrome (MDS), or CML myeloid blast crisis who underwent a second allogeneic HSCT after failure of the first transplant at our institution.

Patients and methods

Patient population

In total, 847 patients with AML, MDS, or myeloid blast crisis of CML underwent allogeneic HSCT between May 1989 and November 2003. Of the 847, 346 patients relapsed and 72 underwent a second allogeneic HSCT from the same or different donor. Patients who had a different donor readily available were transplanted using a different donor, whereas patients who did not have another donor were transplanted using the same donor that was used for the first transplant. The data were collected from Blood and Marrow Transplant Database of The University of Texas M.D. Anderson Cancer Center or by review of patient charts. All patients were treated on protocols approved by the Institutional Review Board or under the compassionate indication mechanism. All patients signed informed consent. The Institutional Review Board granted permission for retrospective chart review.

Statistical analysis

Primary end points for the study were overall survival (OS) (time from second HSCT until death or last follow-up) and LFS (time from second HSCT until disease relapse, disease progression, death during remission, or last follow-up). Actuarial OS and LFS rates were estimated by the Kaplan–Meier method. Only patients who achieved a CR after the second transplant were evaluated for LFS. All patients were evaluated for response around day 30 after transplant and therefore that was considered to be the date of CR.

TRM was defined as a death not attributed to disease. It was estimated by the cumulative incidence method considering death due to disease as competing risk. Acute graft-versus-host disease (GVHD) was defined as per consensus criteria.7 Chronic GVHD was defined by clinical criteria in patients surviving more than 100 days from transplant with sustained donor engraftment.

Comprehensive evaluation of prognostic factors for the primary outcomes was not possible due to the lack of statistical power to detect associations with Hazards ratio (HR) below two-fold. We performed univariate analysis on an exploratory basis for the most relevant clinical factors associated with survival. Comparisons of survival rates at 18 months after the second transplant were performed using the Cox's proportional hazards model and log-rank tests. Differences in patient characteristics were evaluated by the Wilcoxon rank sum test for continuous variables and by the chi-square test for categorical variables.8 Similar analysis was performed after excluding patients who underwent a second transplant for graft failure after the first transplant. Statistical significance was defined at 0.05. Analysis was performed using the statistics software package STATA 7.0.9


Patient characteristics

A total of 72 patients underwent a second HSCT at our institution for myeloid malignancies. Median age at the time of second transplant was 42 years (range, 14–75 years). In all, 97% of the patients had intermediate-risk or poor-risk cytogenetics at diagnosis (Table 1).

Table 1 Patient pretransplant and transplant characteristics

Indication for second transplant was disease relapse or disease progression in the majority of cases (87%). Nine patients (13%) underwent the second transplant because of failure to engraft after the first transplant. Median time to relapse/disease progression after first transplant was 6.3 months (range, 0–88 months). Median time between the first and second transplant was 7 months (range, 1–90 months).


At the time of second HSCT, the majority of patients (50%, 36/72) were in untreated relapse and only 2/72 (3%) were in CR. A total of 46 (63.9%) patients had more than 5% blasts in their BM and 39 patients (54.1%) had circulating PB blasts. The source of stem cells was mobilized PB in 60 patients (83%). In 79% of the patients, the donor for the second transplant was the same as the first transplant and in 21% it was a different donor. GVHD prophylaxis was tacrolimus based in 40/72 (56%) patients and cyclosporine based in 22/72 (31%) patients. A bone marrow aspiration and biopsy was carried out approximately 30 days after transplantation to assess disease response.

Conditioning regimens for second transplant procedure

A variety of different myeloablative and reduced-intensity conditioning regimens were used. Preparative regimens were considered myeloablative if they were likely to produce profound pancytopenia for >28 days without transplantation and if, after transplantation, hematopoietic recovery was likely to be donor derived (complete chimerism in >80% of patients). All other regimens were categorized as reduced-intensity regimens.10 Myeloablative preparative regimens included cyclophosphamide in combination with intravenous busulfan (seven patients) or total body radiation±thiotepa (five patients);11, 12 decitabine (eight patients);13 single agent melphalan (13 patients);14 intravenous busulfan in combination with fludarabine (two patients), or cyclophosphamide and thiotepa (three patients).15, 16

Reduced-intensity regimens included fludarabine, cytarabine, and idarubicin or cisplatin (two patients), fludarabine and melphalan with/without antithymocyte globulin (19 patients);17, 18, 19 fludarabine plus antithymocyte globulin (six patients); fludarabine plus intravenous busulfan (one patient); gemtuzumab ozogamicin, fludarabine, and cytarabine (four patients) or melphalan (one patient).


In total, 62 patients (87.3%) patients engrafted with donor cells following the second transplant. Median time to neutrophil engraftment (absolute neutrophil count 500 μl) after the second HSCT procedure was 12 days (range, 6–50). A total of 24 patients never attained a platelet count >20 000/μl. In patients who did engraft with platelet counts of 20 000/μl, the median time to engraftment was 11.5 days (range, 3–94). Out of the nine patients who underwent the second HSCT for graft failure, seven received PBSC grafts from the same (four patients) or different donor (three patients) and two were rescued by unrelated umbilical cord blood grafts. Seven of nine patients ultimately engrafted with donor cells and two patients had autologous reconstitution.


After the first transplant, 15 out of 72 patients (21%) developed acute GVHD and of the 60 evaluable patients 16 (27%) developed chronic GVHD. The incidence of acute and chronic GVHD after the second transplant was 35% (25 of 71 evaluable patients) and 31%, respectively (16 of 51 evaluable patients). Seven patients had extensive and eight patients had limited chronic GVHD after second transplant.


A total of 12 patients were alive at the time of this analysis, 10 in CR. The most common cause of death was disease relapse/progression. Causes of death are listed in Table 2. The TRM after second transplant was 36% (s.e.±6). Seven patients died before day 30, and 24 patients died within the first 100 days of the second HSCT.

Table 2 Causes of death

LFS and OS

A total of 52 patients (74%) attained a CR after the second transplant. After a median follow up time of 25 months (range, 3–125 months) in surviving patients, the median survival was 6 months. At 68 months, the OS for the entire group was 14% (95% CI, 7–24), Figure 1. For patients achieving a CR after the second transplant (52 patients), the LFS was 15% (95% CI, 7–26) and the OS was 18% (95% CI, 8–30).

Figure 1

Kaplan–Meier estimate of OS from second transplant.

Univariate analysis (entire cohort)

We studied various pre- and post transplant characteristics to determine the predictors of response and survival after second transplant. The factors evaluated include age at second transplant, gender, reason for transplant (relapse vs graft failure), cytogenetics at diagnosis, donor type, same vs different donor, myeloablative vs reduced-intensity conditioning, GVHD prophylaxis, presence or absence of acute and chronic GVHD after the first and second transplant, time to relapse/disease progression after second transplant, disease status at the time of second transplant, and leukemia burden at second transplant (percentage of BM and PB blasts). Results of univariate analysis for OS are summarized in Table 3.

Table 3 Univariate analysis for overall survival

Patients who had a high leukemia burden at the time of the second HSCT (as measured by BM blasts >5% and presence of circulating PB blasts) had a worse outcome (HR 1.8, 95% CI 0.9–3.7, P=0.08), Figure 2.

Figure 2

Kaplan–Meier estimate of survival after second transplant for patients with no circulating blasts plus 5% bone marrow blasts (—)-vs those with circulating blasts plus >5% bone marrow blasts (······).

When survival was compared at 18 months after the second HSCT, patients who had relapsed or progressed more than 1 year after the first transplant had better outcome (HR 2.4, 95% CI, 0.99–5.7) when compared to patients who relapsed or progressed within 1 year (P-value=0.04).

Similar outcomes (P-value=NS) were noted for patients who had intermediate-risk cytogenetics at diagnosis, using different vs same donor for the second HSCT, use of reduced-intensity regimen for both transplants, and transplanting in untested relapse vs treated relapse.

Univariate analysis (excluding graft failures)

The overall results were similar when the analysis was performed excluding patients who underwent their second transplant for failure to engraft after the first transplant. Patients who had a high leukemia burden at the time of the second HSCT (as measured by BM blasts >5% and presence of circulating PB blasts) had a worse outcome (HR 2.9, 95% CI 1.1–7.7, P=0.03). Patients who had relapsed or progressed more than 1 year after the first transplant had again had a better outcome (HR 2.4, 95% CI, 0.99–5.6) when compared to patients who relapsed or progressed within 1 year (P-value=0.05).


The prognosis for patients who relapse after an allogeneic transplant is generally poor. DLI may salvage a subset of patients who experience a smoldering relapse. Similarly, chemotherapy may result in long-term cures in a small number of patients who relapse late.1 Second HSCT may be another alternative in some patients; however, the factors that predict long-term remissions in this patient population are not well defined.

Like other studies our study is also limited by the number of patients and by the heterogeneity of the transplant regimens used. An analysis based on specific combinations of preparative regimens used for first and second transplant procedures and their impact on outcomes could not be carried out due to this heterogeneity. Also, there is an inherent selection bias in this retrospective analysis. Patients who were able to undergo a second transplant had to survive long enough after their relapse to be able to undergo a second transplant. However, the major strengths of this study are that only adult patients with myeloid malignancies were included. Also, the majority of patients had either refractory or untreated relapse at the time of the second transplant.

In our study, disease relapse was the major cause of treatment failure after the second HSCT. This may be because the median time to relapse after the first HSCT was short (median 176 days) and time interval between first HSCT and relapse has been described as an important prognostic factor by several authors.6, 20, 21, 22 Accordingly, in the present study, the subset of patients who relapsed within a year of first transplant had a worse outcome. In a study by the International Bone Marrow Transplant Registry, the probability of LFS and relapse were 7 and 77%, respectively, in patients who relapsed within 6 months of first HSCT vs 28 and 59%, respectively, who relapsed more than 6 months after the first HSCT.21

Eapen et al22 reported increased rate of relapse in patients who received reduced-intensity regimen for the second HSCT, indicating that dose intensity may be important in these patients. However, in general, myeloablative regimens are associated with a higher TRM. In this study, patients who received reduced-intensity preparative regimen for the second transplant had a similar OS to those subjects whose second regimen was myeloablative.

In the present study, the TRM was 36% after the second HSCT, which is comparable to that reported by others. Radich et al5 reported a nonrelapse mortality at 1 year after second HSCT of 45%. In that study the severe regimen-related toxicity rate was also high at 39% and the 1-year relapse rate was 70% resulting in a 1-year LFS of only 14%. Similarly in the analysis by International Bone Marrow Transplant Registry, the 2-year TRM after second HSCT was 41% and the incidence was 3.9 times greater in patients undergoing the second transplant within 6 months of the first HSCT.21 The French group23 studied 150 patients with acute and chronic leukemias who underwent a second allogeneic transplant procedure. After a median follow-up time of 30 months, the overall LFS, TRM, and relapse rates were 30, 68, and 24%, respectively.

Other factors that have been reported to impact the outcome after the second transplant are the absence of aGVHD at the time of second HSCT,5, 23 remission status at second transplant,6, 20, 21 using a female donor,13 development of chronic GVHD after the second HSCT,23 and source of stem cells.24

Al-Qurashi et al25 and Eapen et al22 did not find any advantage to using a different donor for the second transplant procedure. We saw a trend towards improved survival using a different donor though it did not reach statistical significance given the small number of patients.

Leukemia burden at time of transplantation has been considered to be a bad prognostic factor.26 Our group has shown that in patients with refractory or relapsed AML/MDS undergoing allogeneic HSCT, PB and BM blast count impacted on the OS and EFS. In the present study also, patients who had >5% BM blasts at the time of second HSCT had a worse outcome (HR 1.7, 95% CI 0.9–3.1, P=0.08). When the total leukemia burden was taken into account, patients with low leukemia burden had an OS of 25% at 5 years compared to 12% in those with a high burden (Figure 2). This was true even after patients with graft failures were excluded from the analysis.

There were a total of 10 patients who underwent two unrelated donor transplants. At the time of this report all have died. There were three deaths due to disease relapse and seven died due to transplant-related causes. Only five of 10 survived beyond day 100 after the second transplant. Based on this it is difficult to recommend second unrelated donor transplant to patients with relapsed, refractory leukemia.

Graft failure after myeloablative hematopoietic stem cell transplantation presents a major management problem. Most of these patients are at high risk of infections because of the prolonged period of neutropenia. Treatment strategies for these patients have included administration of growth factors, infusion of additional stem cells, immunosuppression, and stem cell infusion from the same or different donor.27 The results are generally poor. In our series, 7/9 patients who underwent the second HSCT for graft failure engrafted; however, only four survived beyond 100 days and only one patient is a long-term survivor.

In conclusion, although the OS was poor, this retrospective analysis identified a subgroup of patients who had lower leukemia burden at the time of second HSCT, 40% of whom survived at least 2 years. Within the limitations of sample size and the heterogeneity of the preparative regimens, it is reasonable to recommend that patients with circulating peripheral blood blasts and excess bone marrow blasts are unlikely to benefit from a second allogeneic transplant.


  1. 1

    Frassoni F, Barrett A, Granena A et al. Relapse after allogeneic bone marrow transplantation for acute leukemia: a survey by the EBMT of 117 cases. Br J Haematol 1988; 70: 317–320.

  2. 2

    Collins R, Shpilberg O, Drobyski W et al. Donor leukocyte infusions in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation. J Clin Oncol 1997; 15: 433–444.

  3. 3

    Bacigalupo A, Soracco M, Vassallo F et al. Donor lymphocyte infusion (DLI) in patients with chronic myeloid leukemia following allogeneic bone marrow transplantation. Bone Marrow Transplant 1997; 19: 927–932.

  4. 4

    Kolb H, Schattenberg A, Goldman J et al. Graft-versus-leukemia effect of donor lymphocyte transfusions in marrow grafted patients: European Group for Blood and Marrow Transplantation Working Party Chronic Leukemia. Blood 1995; 86: 2041–2050.

  5. 5

    Radich JP, Gooley T, Sanders JE et al. Second allogeneic transplantation after failure of first autologous transplantation. Biol Blood Marrow Transplant 2000; 6: 272–279.

  6. 6

    Bosi A, Laszlo D, Labopin M, et al, for the Acute Leukemia Working Party of the European Blood and Marrow Transplant Group. Second allogeneic bone marrow transplantation in acute leukemia: results of a survey by the European Cooperative Group for Blood and Marrow Transplantation. J Clin Oncol 2001; 19: 3675–3684.

  7. 7

    Przepiorka D, Weisdorf D, Martin P et al. 1994 Consensus conference on acute GVHD grading. Bone Marrow Transplant 1995; 15: 825–828.

  8. 8

    Wilcoxon F . Individual comparisons by ranking methods. Biometrics 1945; 1: 80–83.

  9. 9

    Stata Corporation. StataCorp: Stata Statistical Software: Release 7.0. Stata Corporation: College Station, TX, 2001.

  10. 10

    Champlin R, Khouri I, Shimoni A et al. Harnessing graft versus-malignancy: non-myeloablative preparative regimens for allogeneic haematopoietic transplantation, an evolving strategy for adoptive immunotherapy. Br J Haematol 2000; 111: 18–29.

  11. 11

    Andersson BS, Kashyap A, Gian V et al. Conditioning therapy with intravenous busulfan and cyclophosphamide (IV BuCy2) for hematologic malignancies prior to allogeneic stem cell transplantation: a phase II study. Biol Blood Marrow Transplant 2002; 8: 145–154.

  12. 12

    Clift RA, Buckner CD, Thomas ED et al. Marrow transplantation for chronic myeloid leukemia: a randomized study comparing cyclophosphamide and total body irradiation with busulfan and cyclophosphamide. Blood 1994; 84: 2036–2043.

  13. 13

    Ravandi F, Kantarjian H, Cohen A et al. Decitabine with allogeneic peripheral blood stem cell transplantation in the therapy of leukemia relapse following a prior transplant: results of a phase I study. Bone Marrow Transplant 2001; 27: 1221–1225.

  14. 14

    de Lima M, van Besien K, Gajewski J et al. High-dose melphalan and allogeneic peripheral blood stem cell transplantation for treatment of early relapse after allogeneic transplant. Bone Marrow Transplant 2000; 26: 333–338.

  15. 15

    de Lima M, Couriel D, Thall PF et al. Once-daily intravenous busulfan and fludarabine: clinical and pharmacokinetic results of a myeloablative, reduced-toxicity conditioning regimen for allogeneic stem cell transplantation in AML and MDS. Blood 2004; 104: 857–864.

  16. 16

    Bibawi S, Abi-Said D, Fayad L et al. Thiotepa, busulfan, and cyclophosphamide as a preparative regimen for allogeneic transplantation for advanced myelodysplastic syndrome and acute myelogenous leukemia. Am J Hematol 2001; 67: 227–233.

  17. 17

    Giralt S, Estey E, Albitar M et al. Engraftment of allogeneic hematopoietic progenitor cells with purine analog-containing chemotherapy: harnessing graft-versus-leukemia without myeloablative therapy. Blood 1997; 89: 4531–4536.

  18. 18

    Giralt S, Thall PF, Khouri I et al. Melphalan and purine analog-containing preparative regimens: reduced-intensity conditioning for patients with hematologic malignancies undergoing allogeneic progenitor cell transplantation. Blood 2001; 97: 631–637.

  19. 19

    Khouri IF, Keating M, Korbling M et al. Transplant-lite: induction of graft-versus-malignancy using fludarabine-based nonablative chemotherapy and allogeneic blood progenitor-cell transplantation as treatment for lymphoid malignancies. J Clin Oncol 1998; 16: 2817–2824.

  20. 20

    Kishi K, Takahashi S, Gondo H et al. Second allogeneic bone marrow transplantation for post-transplant leukemia relapse: results of a survey of 66 cases in 24 Japanese institutes. Bone Marrow Transplant 1997; 19: 461–466.

  21. 21

    Mrsic M, Horowitz MM, Atkinson K et al. Second HLA-identical sibling transplants for leukemia recurrence. Bone Marrow Transplant 1992; 9: 269–275.

  22. 22

    Eapen M, Giralt SA, Horowitz MM et al. Second transplant for acute and chronic leukemia relapsing after first HLA-identical sibling transplant. Bone Marrow Transplant 2004; 34: 721–727.

  23. 23

    Michallet M, Tanguy ML, Socie G et al. Second allogeneic haematopoietic stem cell transplantation in relapsed acute and chronic leukaemias for patients who underwent a first allogeneic bone marrow transplantation: a survey of the Societe Francaise de Greffe de Moelle (SFGM). Br J Haematol 2000; 108: 400–407.

  24. 24

    Russell J, Bowen T, Brown C et al. Second allogeneic transplants for leukemia using blood instead of bone marrow as a source of hematopoietic cells. Bone Marrow Transplant 1996; 18: 501–505.

  25. 25

    Al-Qurashi F, Ayas M, Al Sharif F et al. Second allogeneic bone marrow transplantation after myeloablative conditioning analysis of 43 cases from single institution. Hematology 2004; 9: 123–129.

  26. 26

    Wong R, Shahjahan M, Wang X et al. Prognostic factors for outcomes of patients with refractory or relapsed acute myelogenous leukemia or myelodysplastic syndromes undergoing allogeneic progenitor cell transplantation. Biol Blood Marrow Transpl 2005; 11: 108–114.

  27. 27

    Grandage VL, Cornish JM, Pamphilon DH et al. Second allogeneic bone marrow transplants from unrelated donors for graft failure following initial unrelated donor bone marrow transplantation. Bone Marrow Transplant 1998; 21: 687–690.

Download references

Author information

Correspondence to C Hosing.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hosing, C., Saliba, R., Shahjahan, M. et al. Disease burden may identify patients more likely to benefit from second allogeneic hematopoietic stem cell transplantation to treat relapsed acute myelogenous leukemia. Bone Marrow Transplant 36, 157–162 (2005).

Download citation


  • allogeneic transplantation
  • acute myelogenous leukemia

Further reading