Complete molecular responses are achieved after reduced intensity stem cell transplantation and donor lymphocyte infusion in chronic myeloid leukemia

The recommended first-line therapy for patients with chronic phase (CP) chronic myeloid leukemia (CML) is imatinib mesylate (IM; Novartis, Basel, Switzerland).¹  IM is effective in inducing durable responses in the majority of patients.²  The use of allogeneic stem cell transplantation (SCT) for CML is declining yet remains a curative option for those with available donors.³  Transplant regimens with reduced intensity conditioning (RISCT) are less toxic, appropriate for a broader patient group, and have particular efficacy in CML.^4,5  The requirement for IM in the context of RISCT is not established, in particular whether IM is necessary after transplantation where the safety and efficacy of escalating doses of donor lymphocyte infusions (DLIs) have been shown.^6-8  We established a protocol for newly diagnosed patients using IM to establish disease control (complete cytogenetic response [CCyR]) before RISCT. IM was discontinued before RISCT and escalating doses of DLI were given after transplantation to those with residual disease (BCR-ABL ≥ 0.02%). Because residual disease after IM therapy may relate to stem cell persistence, RISCT followed by prophylactic DLI delivered when disease burden is low should induce graft-versus-leukemia effect within the stem cell population, leading to disease eradication.

Approval was obtained from the Multicentre Research Ethics Committee of London for these studies. Informed consent was obtained in accordance with the Declaration of Helsinki.

Patients were 18 to 65 years of age with CML in first CP and a matched sibling donor who was screened and fit to donate. Standard definitions of disease stage and treatment response were used.⁹  Patients commenced 400 mg per day IM (600 mg/day from 3 months if no CCyR) with RISCT performed at 6 to 12 months once CCyR was established. The conditioning regimen was fludarabine 30 mg/m² (day −7 to −3), melphalan 140 mg/m² (day −2), and alemtuzumab (MabCampath; Bayer Schering Pharma, Newbury, United Kingdom) 20 mg (once, day −8) and 10 mg twice a day (from day −7 to day −4). All SCTs were obtained from peripheral blood with cells (median dose, 5.53 × 10⁶ CD34⁺ cells/kg) infused on day 0. Granulocyte colony-stimulating factor was given at 5 μg/kg from day 6 until the absolute neutrophil count was maintained at more than 10⁹/L. Graft-versus-host disease (GVHD) prophylaxis with cyclosporin started at day −1 (therapeutic range, 200-300 μg/L), with a reducing dose from day 30 of 10% per week, aiming to stop by day 100. Viral prophylaxis was aciclovir (400 mg twice a day for low risk and 800 mg 4 times a day for high risk of cytomegalovirus [CMV] reactivation). All patients were monitored weekly for CMV, Epstein-Barr virus, and adenovirus reactivation. Standard definitions of acute and chronic GVHD were used.¹⁰  DLIs were considered for those with less than 100% donor chimerism or 0.02% or more BCR-ABL⁺ (on 2 occasions separated by > 2 weeks) after 6 months following RISCT. DLIs were given in increasing doses from 5 × 10⁵ CD3⁺ cells/kg using half-log increments at 3-month intervals to a maximum of 2 doses of 10⁸CD3⁺ cells/kg. The aim was to achieve 100% chimerism and undetectable BCR-ABL, although in practice further DLI was not given once BCR-ABL was consistently less than 0.02%. Patients who had more than grade 2 GVHD at any time were ineligible for DLI. IM therapy was considered for those with BCR-ABL levels fulfilling the criteria for DLI earlier than 6 months after transplantation, those with 0.02% or more BCR-ABL after maximal DLI, those for whom DLI was contraindicated because of GVHD, or those with rapidly rising BCR-ABL levels.

Between 2001 and 2006, 18 patients from 4 centers were recruited. Of these, 3 received myeloablative SCT: 2 for progression to lymphoid blast crisis on IM before SCT and the third for failing to reach a major cytogenetic response on IM. Fifteen of 18 patients with median age 39 years (range, 21-56 years) achieved a CCyR and are considered here; all received a sibling RISCT following the trial protocol. The largest center recruited 11 of 15 cases, and these were consecutive CP CML patients referred for consideration of SCT at diagnosis. Follow-up data extend to a median of 31 months (range, 12-61 months) after transplantation. The patient demographics and transplant data are seen in Table 1.¹¹ 

RISCT with DLI after this protocol was well tolerated, safe, and effective. In line with previous studies, patients underwent early engraftment (median, 10 days, platelets ≥ 20 × 10⁹/L; 11 days, absolute neutrophil count ≥ 0.5 × 10⁹/L) with corresponding short inpatient stays.^5,12  In the first year, 7 of 15 (47%) patients required hospital readmission (range, 1-91 days) because of infection, with GVHD contributing significantly in one case (Table 1). One patient died 12 months after RISCT with no deaths thereafter (overall survival 93% at 1 year by Kaplan-Meier calculation), and all now live independently. Viral reactivation was common, with CMV reactivation in 6 of 7 (86%) patients at risk. This high incidence has been reported previously with alemtuzumab and is related to effects on immune reconstitution.¹³  In addition, 2 patients developed Epstein-Barr virus⁺ posttransplantation lymphoproliferative disorder but responded well to rituximab. GVHD was seen in 8 of 15 (53%) of all RISCTs and 6 of 13 (46%) receiving DLI. The majority had limited chronic GVHD/acute GVHD, and extensive chronic GVHD was seen infrequently. GVHD was probably minimized by the use of alemtuzumab combined with escalating doses of DLI. Patients routinely stopped IM at the time of RISCT. Further IM was required after transplantation to establish disease control in 4 of 15 patients. This was effective in all cases, and IM was safely discontinued (after 4-13 months). All 15 patients are now off IM. The use of IM and DLI combined with BCR-ABL data after transplantation for individual patients is illustrated in Figure 1A.

Evidence for the efficacy of the approach was seen as a progressive reduction in BCR-ABL levels (Figure 1B). Currently, 8 of 15 (53%) patients have sustained undetectable BCR-ABL (defined as 2 undetectable results taken > 2 weeks apart). Of these, 7 of 8 required DLI (dose 0.1-1.6 × 10⁷ CD3⁺ cells/kg in 1-3 infusions), 5 of 8 had GVHD, 2 of 8 had received IM after transplantation, and all were European Group for Blood and Marrow Transplantation (EBMT) score 1 or 2. There has been no cytogenetic or hematologic relapse after transplantation; and for patients in whom longer follow-up is available, responses are maintained.

Since this protocol was established, the role for SCT has diminished, with a shift toward first-line IM for all CP patients.¹⁴  This is reflected in a recent proposal outlining treatment strategies for CP CML where SCT is considered only for patients with initial high-risk assessments and those intolerant or responding poorly to IM.¹  Published comparisons between drug treatment and transplantation promote this stance, although survival curves may cross as follow-up progresses.^2,15-17  Interpretation is further complicated by the relatively recent introduction of IM, the availability of new more potent tyrosine kinase inhibitors (TKIs), improvement in supportive care for SCT, and the lack of evaluation of RISCT in this setting. It should also be borne in mind that maintaining transplantation as a “last resort” for those intolerant or resistant to other therapies will result in patients who receive transplants having higher EBMT risk scores as a result of the longer time to transplantation, increased age, and the risk of disease progression.

It is probable that the majority of patients in this study would have maintained a CCyR if maintained on IM in lieu of RISCT. We think that the proportion achieving undetectable BCR-ABL after RISCT and DLI is comparable with the response that would be expected with IM alone (∼41% at 4 years² ) and is consistent with published data on undetectable BCR-ABL after SCT (71% at median 6.6 years of follow-up¹⁵ ). The RISCT procedure was well tolerated, and all patients are currently off any CML therapy. Regular monitoring is necessary because follow-up is relatively short and the risk of late relapse must be significant. However, if this occurs, it should be DLI responsive and is to be compared with the published annual rates of IM resistance and of disease progression on IM. We note that, in our series, one of 15 patients died and transplantation-related mortality of 23% at 2 years after RISCT for CML has been reported (although this value includes all phases of CML).¹⁸  We agree with the use of IM first-line in CML in CP and have discontinued RISCT at diagnosis in our institution. However, our approach highlights the safety and efficacy of RISCT as a potential therapeutic option that might be used when there are financial constraints or local access issues with TKIs, when the patient has a strong preference for transplantation and against continuous drug therapy, and for cases with severe TKI intolerance or early evidence of TKI resistance in CP.

The authors thank all medical and nursing staff and data managers who have been involved in the care of these patients and in the collection of trial data.

This work was supported by Novartis and Chugai Pharma. N.B.H. is a Leukemia Research Fund–sponsored clinical research fellow. M.C. is a Medical Research Council clinical research fellow. G.M.S. has received research support and honoraria from Novartis.

Contribution: N.B.H., M.C., and K.S. collected and analyzed the data; M.C., J.G., A.N.P., I.G.M., G.M.S., C.C., P.S., and T.L.H. designed the research; N.B.H., M.C., and T.L.H. wrote the paper.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Tessa Holyoake, Section of Experimental Haematology, Royal Infirmary, 10 Alexandra Parade, Glasgow, G31 2ER, United Kingdom; e-mail: tlh1g@clinmed.gla.ac.uk.