- © 2006 by American Society of Clinical Oncology
Long-Term Follow-Up After Reduced-Intensity Conditioning Allogeneic Transplantation for Acute Myeloid Leukemia/Myelodysplastic Syndrome: Late CNS Relapses Despite Graft-Versus-Host Disease
To the Editor:
In 2003, we reported the interim analysis of a prospective clinical trial of reduced-intensity conditioning (RIC) allogeneic hematopoietic stem-cell transplantation (AHSCT) for patients with predominantly poor-risk acute myeloid leukemia/myelodysplastic syndromes (AML/MDS).1
Twenty-five patients have now undergone transplantation from fully human leukocyte antigen–matched family donors on this study between April 1999 and March 2005 (Table 1). No in vivo or in vitro T-cell depletion was used. Treatment-related mortality (TRM) remained very low at 4%, with only one patient dying free of relapse at day +155 from pulmonary infection (with bone marrow aplasia after withdrawal of immunosuppression). There was little treatment-associated morbidity, with a median inpatient stay of 27 days (range, 17 to 44 days); 20% of patients did not require intravenous antibiotics for neutropenic fever. We observed a much lower frequency of cytomeoplovirus (CMV) reactivation (12%) than reported with conditioning protocols employing in vivo T-cell depletion despite donor and/or recipient CMV-seropositivity in 84% of donor/recipient pairs.2,3 All patients achieved primary engraftment, and stable neutrophil engraftment took a median of 15 days (range, 2 to 25 days). The median whole-blood chimerism assessed at day +28 was 89% donor (range, 40% to 100%). Mild (grade 1 to 2) acute graft-versus-host disease (GvHD) was seen in only four patients (16%) and no cases of severe (grade 3 to 4) acute GvHD occurred.
Patients With Predominantly High-Risk AML/MDS Undergoing Non-T-Cell Depleted Reduced-Intensity Conditioning Sibling Allogeneic Hematopoietic Stem-Cell Transplantation With Fludarabine and Cyclophosphamide Conditioning (n = 25)
After cessation of post-transplantation immunosuppression, 14 patients (56%) developed GvHD. No patients died as a result of GvHD post–cessation of immunosuppression. No significant relationship was seen between the development of GvHD after cessation of immunosuppression and the dose of either CD34+ or CD3+ T cells received.
With a median follow-up of surviving patients of 3.5 years, 14 patients (56%) are alive, of whom 13 (52%) are free of disease. Treatment failure (TRM or relapse) occurred in 14 patients (56%), although two patients who relapsed have undergone repeat RIC AHSCT and are alive and disease free. Estimated actuarial overall survival and event-free survival are 55% and 40%, respectively, at 4 years (Fig 1A).
(A) Estimated actuarial event-free survival (EFS) and overall survival (OS) in acute myeloid leukemia/myelodysplastic syndromes (AML/MDS) patients receiving reduced intensity conditioning sibling allogeneic hematopoietic stem-cell transplantation (AHSCT). (B) Cumulative incidence of relapse after reduced-intensity conditioning AHSCT for AML/MDS and the effect of the development of graft-versus-host disease (GvHD) after cessation of post-transplantation immunosuppression. Log-rank P = .0009. Only patients with no evidence of disease at day +100 are shown.
Significantly fewer relapses were observed in patients who developed GvHD after cessation of post-transplantation immunosuppression (Fig 1B), consistent with a clinically significant graft-versus-leukemia (GvL) effect, presumably targeting alloantigens present on both recipient GvHD target tissues and leukemic cells. Recent reports have confirmed our observation of the protective effect of GvHD on relapse risk in RIC AHSCT for AML/MDS.4
Despite the development of GvHD after cessation of immunosuppression, isolated CNS relapse occurred in two patients late after transplantation; one patient with AML M1 (patient 3) at D+780 and another with MDS/AML (patient 5) at D+1889. Neither patient had prior CNS leukemia, nor had either received any CNS-directed therapy.
Ten percent to 30% of patients with AML who relapse after AHSCT do so at isolated extramedullary anatomic sites, most commonly patients with myelomonocytic or monocytic differentiation, although isolated CNS relapse is rare.5 The GvL effect after RIC AHSCT may not target the CNS as efficiently as the bone marrow. Certainly responses to donor lymphocyte infusions after RIC AHSCT are less effective in AML patients with extramedullary sites of relapse.6 The only randomized trial examining CNS prophylaxis as part of primary treatment for AML7 reported no reduction in CNS relapse in those who received prophylactic intrathecal (IT) chemotherapy. There is considerable heterogeneity in current practice of IT prophylaxis and AHSCT, but on the basis of the rarity of isolated CNS relapse after full-intensity conditioning AHSCT, the European Group for Blood and Marrow Transplantation does not currently recommend IT prophylaxis for AML patients without prior CNS involvement.8
However, administration of prophylactic IT treatment after RIC AHSCT for AML/MDS may prevent the occurrence of disease relapse within the CNS, (which may not be prevented by ongoing GvHD), and further improve the durable remission rate of 40% that we have observed.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
