Dose-dependent effects of in vivo antithymocyte globulin ... - Nature

5 downloads 0 Views 156KB Size Report
genous leukemia (CML) in first chronic phase. Median age was 36 years (16–51). In all, 40 patients were transplanted within 2 years after diagnosis and 57 later.
Bone Marrow Transplantation (2003) 32, 243–250 & 2003 Nature Publishing Group All rights reserved 0268-3369/03 $25.00

www.nature.com/bmt

Dose-dependent effects of in vivo antithymocyte globulin during conditioning for allogeneic bone marrow transplantation from unrelated donors in patients with chronic phase CML M Schleuning1, W Gu¨nther2, J Tischer1, G Ledderose1 and H-J Kolb1,2 1 Medical Clinic III, University Hospital Großhadern, Ludwig-Maximilians-University, Munich, Germany; and 2Clinical Cooperative Group Hematopoietic Cell Transplantation, GSF-National Research Center for Environment and Health, Munich, Germany

Summary: We conducted a dose–escalation study with antithymocyte globulin (ATG) in patients undergoing unrelated donor bone marrow transplantation (URD-BMT). This study analyzes the results for 97 patients with chronic myelogenous leukemia (CML) in first chronic phase. Median age was 36 years (16–51). In all, 40 patients were transplanted within 2 years after diagnosis and 57 later during disease. ATG-S (Fresenius) 20–120 mg/kg body weight (b.w.) was given prior to transplantation. A total of 31 patients received less than 60 mg/kg b.w. and 66 patients received 60 mg/kg b.w. or more. All patients except one were grafted with bone marrow, and graftversus-host disease (GVHD) prophylaxis consisted of cyclosporin A and methotrexate. Graft failure did occur in one patient. Grade II–IV acute GVHD developed in 56.7% and extensive chronic GVHD in 11.3% of the patients. The relapse rate was 13.4%. With a median follow-up of 5.8 years (1.5–12.1), 5-year disease-free and overall survival for all patients were 56 and 66%, and for patients transplanted within 2 years of diagnosis it was 72 and 82%. A lower dose of ATG was a significant risk factor for poor outcome. In summary, URD-BMT remains an excellent treatment option for patients with early phase CML, if a sufficient amount of ATG is included in the preparative regimen. Bone Marrow Transplantation (2003) 32, 243–250. doi:10.1038/sj.bmt.1704135 Keywords: CML; anti-thymocyte globulin

The therapeutic options for patients with chronic myelogenous leukemia (CML) have changed dramatically during the past years. In particular, the introduction of the bcrabl-specific tyrosine kinase inhibitor imatinib mesylate (Gleevecs) now offers a highly effective and well-tolerated oral therapy for CML patients. In contrast, allogeneic bone

Correspondence: Prof Dr M Schleuning, Center for Bone Marrow Transplantation, Deutsche Klinik fu¨r Diagnostik, Aukammallee 33, 65191 Wiesbaden, Germany Received 6 February 2003; accepted 2 March 2003

marrow transplantation (BMT), which is still the only curative treatment option, is associated with considerable mortality and morbidity. It, therefore, has been questioned if BMT should be offered to CML patients early during disease, especially if they lack an HLA-identical sibling donor.1 Given the fact that, by the combined effort of the international registry networks, there are now 8 million HLA-typed volunteer donors available, the chance of finding a suitable donor in reasonable time for patients lacking an HLA-matched family member now exceeds 75%. Therefore, unrelated donar BMT (URD-BMT) is a realistic treatment option for a large group of patients in need. It was, however, reported to be associated with a higher risk of treatment failure.2 In an analysis of more than 1400 patients with CML who underwent URD-BMT, graft failure and acute and chronic graft versus host disease (GVHD) were identified as the major obstacles of URD-BMT in these patients, whereas relapses occurred in only 6% of the patients. By contrast, transplant in chronic phase, within 1 year of diagnosis, younger recipient age, a cytomegalovirus (CMV) seronegative recipient and the lack of severe acute GVHD were independent factors for improved disease-free survival.3 Although the routine use of high-resolution molecular typing of HLA class II and recently class I antigens already improved the incidence of GVHD to some extent,4 further reducing the risks of graft failure and acute GVHD remain the major tasks in the setting of URD-BMT in CML patients. For this purpose, ex vivo T cell depletions of the marrow grafts have been performed in patients with CML, which led to a decreased incidence of acute GVHD.5,6 However, ex vivo T cell depletions carry the risk of higher relapse rates since the graft-versus-leukemia effect widely depends on the presence of donor T cells.7,8 An alternative strategy employs an in vivo T cell depletion with either monoclonal T cell-specific antibodies, like Campath1-H,9 or with polyclonal antibodies, like antithymocyte globulin (ATG).10,11 During the years 1990– 1997, we conducted a dose-escalation study with ATG for all the patients receiving marrow grafts from unrelated donors. This analysis evaluates the outcome of URD-BMT for patients with CML in first chronic phase with special respect to the dose of ATG applied during conditioning, and to the timing of transplantation.

ATG in unrelated transplants for CML patients M Schleuning et al

244

Patients and methods Patients As analyzed by November 2002, 97 patients with CML in first chronic or accelerated phase received bone marrow grafts from unrelated volunteer donors between September 1990 and May 2001 after giving their informed consent to the procedure at the BMT unit of the medical clinic 3 of the University Hospital Grosshadern in Munich, Germany. Accelerated phase was defined either by the occurrence of B symptoms, or by increasing splenomegaly, or more than 20% basophils or eosinophils, or increasing numbers of leukocytes despite adequate therapy, or platelet counts of less than 100 000/ml or anemia with hemoglobin levels of less than 9 g/dl, or more than 10% blasts and promyelocytes in the peripheral blood or in bone marrow. Patients with advanced disease (blastic transformation or second chronic phase), with a follow-up of less than 18 months and patients receiving reduced intensity conditioning regimens were excluded from the study. Patient characteristics are summarized in Table 1. Median patient age was 36 years (range 16 – 51 years). The median time interval from initial diagnosis to transplantation was 30 months (range 6 –176 months).

Donor selection Unrelated volunteer donors were preselected centrally at the central bone-marrow donor registry for Germany Table 1

Patients characteristics

Number of patients Median patient age Gender (male/female) Median follow-up Number of patients Transplanted within 2 years after diagnosis Transplanted later than 2 years after diagnosis Disease status First chronic phase Accelerated phase Recipient/donor sex matching Male/male Male/female Female/male Female/female CMV status (patient/donor) Neg/neg Neg/pos Pos/neg Pos/pos CMV reactivation Antigenemia or PCR positivity Conditioning 12 Gy TBI—Cy 200 12 Gy TBI—Cy 120 HLA-typing Identical 1-antigen mismatch 2-antigen mismatch Cell source Bone marrow (median MNC) PBSC (CD34+ cells)

Bone Marrow Transplantation

97 36 (16 – 51) 57/40 5.8 (1.5 – 12.1) years 40 57

72 (74.2%) 25 (25.8%) 36 20 21 20

(37.1%) (20.6%) (21.6%) (20.6%)

37.9% 24.2% 20.0% 17.9% 27% 58 39 87 8 2 96 (3.05  108/kg) 1 (10.4  106/kg)

(ZKRD) in Ulm, Germany. They were matched by serotyping of HLA class I antigens and by DNA typing of the HLA class II antigens DRB1 and DQB1. In total 87 donor–recipient pairs were HLA identical with respect to HLA-A, HLA-B, HLA-DRB1 and HLA-DQB1. Owing to the improbability of finding better matched unrelated donors one antigen difference was accepted in eight pairs (HLA-A, n ¼ 2; HLA-B, n ¼ 1; HLA-DRB1, n ¼ 4; HLADQB1, n ¼ 1) and two antigen differences were accepted in two pairs (HLA-A+HLA-DRB1, n ¼ 1; HLADRB1+HLA-DQB1, n ¼ 1). In the case of two or more HLA-matched volunteer donors, final donor selection was done by additional criteria, like gender, history of pregnancies or blood transfusions, age and CMV immunization, at the transplantation unit.

Conditioning regimen All patients received a standard conditioning regimen consisting of 12 Gy total body irradiation (TBI) and cyclophosphamide. TBI was administered in three fractions of 4 Gy from day 8 to day 6 in patients receiving a 4-day course of cyclophosphamide and from day 6 to day 4 in patients receiving a 2-day course of cyclophosphamide. Cyclophosphamide was given at a daily dose of 50 mg/kg body weight (b.w.) on days 5 to day 2. In order to reduce toxicity starting October 1997, the total dose of cyclophosphamide was reduced from 200 mg/kg b.w. (n ¼ 58) to 120 mg/kg b.w. (n ¼ 39) and it was administered at a daily dose of 60 mg/kg b.w. on day 3 and day 2. All patients received forced diuresis and dose-equivalent Mesna during the days of cyclophosphamide treatment.

In vivo T cell depletion with ATG In 1990, we initiated a dose escalation study with antithymocyte globulin (ATG-S, Fresenius, Bad Homburg, Germany) for all patients receiving grafts from unrelated volunteer donors.12 All eligible patients were entered in a sequential manner and after inclusion of 30 patients to each dose level, the dose was escalated to the next level. From September 1990 to September 1996, the dose was escalated from 20 mg/kg b.w. to 120 mg/kg b.w. (calculated on actual body weight). Owing to the serious adverse events at the highest dose level (two cases of intracerebral hemorrhage due to severe thrombocytopenia in non-CML patients), the dose was de-escalated to 60 mg/kg b.w. during the following year. This dose level was maintained in the following years. In all, 31 patients with CML in first chronic or accelerated phase received a total dose of ATG of less than 60 mg/kg b.w. and 66 received 60 mg/kg b.w. or more. For details see Table 2. ATG was administered in four equivalent fractions in the morning 8 h before cyclophosphamide from day 5 to day 2 in patients receiving a 4-day course of cyclophosphamide and in three equivalent doses from day 3 to day 1 in patients receiving a 2-day course of cyclophosphamide. Initially, ATG was given without premedication over 3–4 h. Later, the infusion time was prolonged to 6–8 h and patients received 250 mg of prednisone i.v. prior to ATG infusion and 4 h later.

ATG in unrelated transplants for CML patients M Schleuning et al

Table 2

ATG doses applied

Total ATG dose (mg/kg) 20 30 40 50 60 75 80 120

Number of patients (%) 18 1 11 1 38 1 22 5

(18.6) (1) (11.3) (1) (39.2) (1) (22.7) (5.2)

Transplantation All but one patient, where the donor refused to donate bone marrow, received unselected bone marrow as graft. In cases of ABO blood group incompatibility, either red cells were removed by sedimentation facilitated by gelatine (Gelafundins, Braun-Melsungen, Melsungen, Germany) or plasma was removed by centrifugation or both. The graft was transfused via a tunneled central nutrition line over 1–2 h. The mean mononuclear cell (MNC) count of the graft was 3.05  108 cells/kg b.w. of the recipient.

GVHD prophylaxis and treatment All patients received cyclosporine A (CSA) from day 1 to day +180 and a short-course of methotrexate (MTX) for GVHD prophylaxis. CSA was started as continuous infusion at day 1 at a dose of 5 mg/kg and reduced to 3 mg/kg at day +3. After day +3, the dose of CSA was adjusted to serum trough levels. MTX was given intravenously at a dose of 15 mg/m2 on day +1, and at a dose of 10 mg/m2 on days +3 and +6. In case of severe liver toxicity or mucositis, patients received leukovorin 15 mg i.v. on day +4 and +7, 24 h after the MTX injection. Acute GVHD was graded by standard criteria and treatment with high-dose prednisone (3 mg/kg b.w.) was initiated if at least grade II acute GVHD developed. In responders, prednisone was then tapered by 10% every 3 days. Nonresponders received various second-line treatments, like muromonab-CD3 (Orthoclone OKT3s, Janssen-Cilag, Neuss, Germany) or horse-derived antilymphocyte globulin (Lymphoglobulin Merieuxs, Institut Merieux Transplant, Leimen, Germany).

Supportive care All patients were nursed in single rooms with hepa-filtered positive air pressure. Standard antibiotic prophylaxis consisted of trimethoprim–sulfametoxazole (320 mg/ 1600 mg/day) pretransplant until day 1, and after engraftment 480 mg/2400 mg twice weekly. Additionally, all patients received oral colistinsulfate and amphotericin B. Broad spectrum intravenous antibiotic treatment in combination with fluconazole (200 mg/day) was started in the event of fever according to good clinical practice procedures. For antiviral prophylaxis, patients received intravenous acyclovir (3  10 mg/kg b.w./day) and polyvalent intravenous immunoglobulin (0.3 g/kg b.w.) every other week. For VOD prophylaxis, patients over the age of

40 or with a history of liver disease either received prostaglandin E1 (Prostavasines, Schwarz-Pharma, Mannheim, Germany) at a dose of 320 mg in four equivalent doses daily or defibrotide (Prociclides, Crinos SPA, Villaguardia, Italy) at a dose of 3200 mg in four equivalent doses daily. Only CMV-negative, leukocyte depleted and irradiated blood products were used. Weekly screening for the detection of CMV reactivation was performed in blood and urine samples by PCR and antigen assays. Gancyclovir or foscarnet treatment was initiated either prophylactically in case of prednisone treatment of acute GVHD in patients at risk (recipient or donor or both CMV positive) or as pre-emptive therapy in patients with repeated positive test results or increasing genome equivalents.

245

Engraftment Engraftment is defined as the first of three consecutive days when the patient’s neutrophil counts exceeded 0.5  109/l.

In vivo detection of ATG We measured the concentration of rabbit IgG, the development of anti-ATG-antibodies by ELISA, and the amount of rATG by flow cytometry analysis in the serum of four patients who received 60 mg/kg b.w. (three patients) or 90 mg/kg b.w. Blood samples were taken prior to ATG injection and up to 50 days after stem cell transplantation. Rabbit IgG and anti-rabbit IgG-antibody (ELISA). Rabbit IgG were detected by an immunoenzymatic assay. The 96-well microtiter plates were coated with 1 mg goat antirabbit IgG (Jackson Immuno Research Lab., West Grove, PA, USA). After incubation of patient sera taken at various time points, bound rabbit-IgG was detected by alkaline phosphatase-conjugated goat anti-rabbit F(ab)2 (Immunotech, Hamburg, Germany). Absorbance was measured at 405 nm using an ELISA plate reader. To detect anti-rabbit antibodies, a similar procedure was applied except that microtiter plates were coated with diluted ATG and the extinction was revealed by peroxidase-conjugated goat anti-human IgG. FACS analysis. The amount of free ATG in sera of patients was assessed by flow cytometry using Ficoll separated PBL. Sera were diluted 1:2 in PBS. A 50 ml volume of these dilutions were mixed with PBL and incubated for 30 min, washed twice with PBS, incubated with a FITC-conjugated goat anti-rabbit IgG antibody (Jackson Immuno Research Lab., West Grove, PA, USA), and washed with PBS. The amount of ATG in sera was estimated by plotting the medium fluorescence intensity (MedFI) from these probes against a standard curve of ATG (200 mg/ml to 0.1 mg/ml), as described.13

Statistical analysis All statistical analyses were performed utilizing SPSS for Windows 10.0. (SPSS Software GmbH, Mu¨nchen, Germany). Kaplan–Meier estimates were used to calculate the probability of overall and event-free survival. Event was Bone Marrow Transplantation

ATG in unrelated transplants for CML patients M Schleuning et al

246

defined as relapse or death from any cause. For univariate analyses, Kaplan–Meier estimates were compared with the log-rank test. A P-value o0.05 (a ¼ 5%) was considered significant. For multivariate analyses all, variables that have been significant at a level of a ¼ 20% have been selected for Cox regression analyses. The final model was calculated by backward selection. For simple comparisons, Student’s t-test was used. A two-sided P-value of o0.05 was considered significant.

Results ATG levels in patients sera and ATG binding to lymphocytes Serum levels of rabbit IgG started to decrease between days 21 and 37 after transplantation, while antibodies against ATG were already found to raise between days 11 and 24 after transplantation. rATG in serum was measured by FACS analysis after blocking the Fc receptor. Levels of rATG in serum were distinguishable down to 0.8 mg/ml. Free rATG increased till day 1 up to 40 mg/ml serum and was detectable with a median half-life of 11 (range 4–15) days in concentrations above 1.5 mg/ml between day 11 and day 30 after stem cell transplantation (Table 3). Blood taken from a patient who engrafted 16 days after transplantation revealed an average binding of 70% of leukocytes even at this time after transplantation. Approximately 90% of CD3+, CD4+ and CD8+ lymphocytes had detectable rATG on their surface (Table 4), and 82% of NK cells had a positive signal for rATG.

Engraftment and graft failure All but one patient surviving more than 30 days after transplantation engrafted. The one patient who primarily did not engraft received only a total dose of 40 mg/kg b.w.

Table 3 Serum levels of free rabbit IgG, anti-rabbit-antibody and free rabitt ATG in patients

Patient Patient Patient Patient

Table 4

Free rabbit-IgG Start of decrease (day)

Anti rabbit antibody Start of increase (day)

Free rATG level > 1.5 mg/ml, detectable till day

32 25 37 21

17 11 24 18

16 24 30 26

1 2 3 4

Antibodies binding to lymphocytes 14 days after SCT

CD3+ CD3 CD4+ CD4 CD8+ CD8

Bone Marrow Transplantation

ATG pos.

ATG neg.

46.5 33.5 17.0 52.0 19.9 50.5

2.2 17.8 1.0 29.0 1.7 27.9

ATG during conditioning and died from infectious complications on day +42. Three more patients died from treatment-related toxicity before engraftment. The median time to engraftment was 23 days for patients receiving o60 mg/kg b.w. ATG and 20 days for patients receiving at least 60 mg/kg/ b.w. This difference was highly significant (Student’s t-test: P ¼ 0.008). The median number of transfused mononuclear cells was 3.05  108 (range 0.6 – 8.5  108) cells/kg b.w. of the recipient. There was a significantly better overall survival and a trend toward better event-free survival for patients receiving at least 3.0  108 cells/kg b.w. (log-rank test: P ¼ 0.0338 and 0.0895, respectively). The one patient grafted with GCSF-mobilized peripheral blood stem cells received 10.4  106 CD34+ cells/kg b.w. and engrafted on day +16.

GVHD Acute GVHD grade II–IV, requiring treatment, occurred in 55 patients (56.7%). Severe acute GVHD grade III–IV developed in 17 patients (17.5%). The incidence of acute GVHD grade III–IV was significantly higher in those patients receiving ATG doses less than 60 mg/kg b.w. (10.6 vs 32.5%; Student’s t-test: P ¼ 0.025). Chronic GVHD occurred in 32 patients (32.9%) and was extensive in 11 patients (11.3%). Seven of the 32 patients developed chronic GVHD only after being treated for relapse with donor lymphocyte infusions (DLI). There was no significant difference for the incidence of chronic GVHD with regard to the dose of ATG used. Five patients (5.2%) died from acute GVHD, three of them had received less than 60 mg/kg b.w. of ATG. The incidence of fatal acute GVHD was 9.7% in the group of patients who received less than 60 mg/kg b.w. of ATG, and 3.0% in the group of patients who received 60 mg/kg b.w. or more of ATG. Two patients (2.1%) died from chronic GVHD.

Adverse reactions of ATG Independent of the dose applied, adverse reactions to ATG were frequent, especially during the first day of application and have been detailed elsewhere.14 By using a longer infusion schedule and steroid premedication, side effects could be markedly reduced although not totally avoided. As has been published, the most frequent adverse reactions included fever, chills, hemolysis, thrombocytopenia and diarrhea.10,11 At the highest dose level of 120 mg/kg b.w. two patients, not included in this analysis because of different underlying diseases, died from intracerebral hemorrhage possibly because of the observed profound thrombocytopenia. In CML patients, thrombocytopenia also occurred but was less severe than in patients with acute leukemia and all patients reported in this study received and tolerated the scheduled dose of ATG.

Treatment failure Treatment-related mortality at day +100 was 12.6%. and overall mortality not related to relapse was 30.9%. Causes and time of death are detailed in Table 5. A total of 13 patients (13.4%) relapsed and 12 of them responded to

ATG in unrelated transplants for CML patients M Schleuning et al

247 Table 5

Causes of nonrelapse-related death

n=30

CAUSE of death

Day of death

5 2 1 19 2 1 6

+37, +39, +42, +47, +77 +372; +372 +123

6

Acute GVHD Chronic GVHD Hemorrhage Infections Aspergillosis CMV-pneumonia Pneumonia (bacterial, viral) Sepsis

2 2 1 1 1

Multiorgan failure Toxoplasmosis Veno-occlusive disease Secondary malignancy Unknown

+23, +148 +269 +23, +108, +530, +545, +663; +1479 +17, +23, +112, +119, +192, +1325 +77, +195 +42, +65 +12 +780 +762

DLI. Three of the responders later died for various reasons (acute GVHD after DLI for second relapse, gastrointestinal bleeding, bronchiolitis obliterans). The risk of relapse was not significantly different in patients receiving less (6/31 or 19.4%) or at least 60 mg/kg b.w. of ATG (7/66 or 10.6%) and was not correlated to the disease status at transplantation since only three of the relapsed patients were transplanted in accelerated phase.

Overall and event-free survival With a median follow-up of 5.8. years (range 1.5–12.1 years), 5-year disease-free and overall survival for all patients was 56 and 66%, respectively. There was significant better overall survival (log-rank test: P ¼ 0.0429) and significant better event-free survival (log-rank test: P ¼ 0.0293) for patients receiving 60 mg/kg b.w. or more of ATG (Figure 1). In patients transplanted within 2 years of diagnosis,event-free and overall survival were 72 and 82%, respectively. In this subgroup the differences with respect to the ATG dose used were even more pronounced. Patients receiving 60 mg or more of ATG had a 5-year disease-free and overall survival of 83 and 90%, respectively and for patients receiving less than 60 mg the survival rates were 32 and 55%, respectively (Figure 2). These differences could not be observed in the subgroup of patients transplanted later than 2 years after diagnosis (data not shown).

Univariate analysis For overall and event-free survival, the following risk factors were analyzed by log-rank test: patient age (o45 vs X 45 years), disease status at transplantation (first chronic phase vs accelerated disease), time to transplantation (o2 vs X2 years), recipient–donor sex combination (male recipient/female donor vs all other combinations), HLA mismatch, graft MNC number (o3  108 vs X3 108 kg b.w), the occurrence of acute GVHD, grade II–IV, the occurrence of chronic GVHD, the occurrence of relapse (only for overall survival), CMV reactivation (defined by PCR positivity or CMV antigenemia), CMV status of recipient and donor, dose of ATG (o60 vs X60 mg/kg b.w.), dose of cyclophosphamide and transplantation period (1990–1996 vs 1997–2001). Significant (Po0.05)

Figure 1 Kaplan-Meier estimates of (a) overall and (b) event-free survival of all CML patients included in the study. The solid line indicates patients who received 60 mg/kg of ATG or more, as compared to those who received less than 60 mg/kg of ATG (broken line). The statistical significance, as indicated, has been calculated by log-rank test.

factors and additional factors (Po0.2) selected for multivariate analysis for overall and event-free survival are shown in Table 6. The same analysis just omitting the parameter time to transplantation was performed for the subgroup of patients transplanted within 2 years after diagnosis. The results are depicted in Table 7. Whereas univariate analysis for all patients revealed the occurrence of acute GVHD, the higher dose of cyclophosphamide, disease status at transplantation and patient’s age of at least 45 years as the most important adverse factors for overall survival, the subgroup analysis for patients transplanted within 2 years of diagnosis found disease status at transplantation and o60 mg/kg b.w. of ATG as the only significant adverse factors for overall survival. For eventfree survival time to transplantation, disease status at transplantation, a lower dose of ATG and a dose of 200 mg/kg b.w. of cyclophosphamide were significant risk factors for an unfavorable outcome in the whole group, whereas in the subgroup of patients transplanted within 2 years of diagnosis low-dose ATG and disease status at transplantation again proved to be the only significant risk factors. In the subgroup of patients transplanted later than Bone Marrow Transplantation

ATG in unrelated transplants for CML patients M Schleuning et al

248 Table 7 Univariate analysis for overall and event-free survival for patients transplanted within 2 years of diagnosis (asterisks indicate significant factors) Factor Disease status Transplantation period Dose of ATG Dose of cyclophosphamide

Overall survival *

P=0.0087 P=0.1925 P=0.0113* P>0.2

Event-free survival P=0.0219* P=0.1187 P=0.0014* P=0.1860

Cox regression analysis The most important independent risk factors for unfavorable outcome with respect to overall survival of the whole group were the occurrence of acute GVHD, patient’s age of at least 45 years, graft cell numbers of less than 3.0  108/ kg. b.w. of the recipient and a time interval of more than 2 years from diagnosis to transplantation (Table 8). Time to transplantation was also a significant factor influencing event-free survival, as was dose of cyclophosphamide, a female donor for a male recipient and less than 60 mg/kg of ATG during conditioning. Owing to the relatively small number of patients and a strong correlation between dose of ATG, dose of cyclophosphamide and the transplantation period in the subgroup of patients transplanted within 2 years after diagnosis, Cox regression analysis was not performed for this subgroup.

Discussion

Figure 2

Kaplan–Meier estimates of (a) overall and (b) event-free survival of patients transplanted within 2 years after diagnosis. The solid line indicates patients who received 60 mg/kg of ATG or more, as compared to those who received less than 60 mg/kg of ATG (broken line). The statistical significance, as indicated, has been calculated by log-rank test.

Table 6 Univariate analysis for overall and event-free survival for all patients (asterixes indicate significant factors) Factor Patient age Disease status Time to transplantation Recipient–donor sex combination Graft cell number Acute GVHD, grade II–IV Transplantation period CMV status of recipient Dose of ATG Dose of cyclophosphamide

Overall survival

Event-free survival

P=0.0112* P=0.0092* P=0.0266* P=0.0717 P=0.0338* P=0.0012* P>0.2 P=0.1083 P=0.0429* P=0.0098*

P=0.0898 P=0.0165* P=0.0146* P=0.0668 P=0.0895 P=0.0468* P=0.1990 P>0.2 P=0.0293* P=0.0140*

All factors with a P value of less than 0.2 were selected for Cox regression analysis.

2 years after diagnosis, the most important risk factors were the occurrence of acute GVHD, lower graft cell numbers and a female donor for a male recipient (data not shown). Bone Marrow Transplantation

Many groups have reported that the interval between diagnosis and transplantation is a very important factor for the outcome of allogeneic stem cell transplantation in patients with CML.3,15–17 Our data further confirm these observations. The Cox regression analysis revealed that the risk of an unfavorable outcome was more than doubled in patients transplanted later than 2 years after diagnosis. Although the overall incidence of acute and chronic GVHD was not different in patients transplanted within 2 years of diagnosis or later, the severity of GVHD was enhanced in patients transplanted late during the disease, thus resulting in a doubled rate of nonrelapse-related mortality. In addition, there was also a higher percentage of relapses in these patients. Our study also confirms and extends the data of other groups10,11 that by the inclusion of ATG into the preparative regimen the incidence and, especially, the severity of acute GVHD can be reduced. Severe acute GVHD grade III–IV developed in 17.5% of our patients. This incidence is much lower than those reported from other studies (430%), which did not include ATG in the preparative regimen, but used standard GVHD prophylaxis mainly with CSA and MTX.3,18 However, this benefit seems to be restricted to, and then being even more pronounced in, those patients who received at least 60 mg/kg b.w. of ATG. Furthermore, the low incidence of graft failure is a very important finding of our study which favorably compares to the results of up to 16% of graft failure reported in studies not incorporating ATG into the preparative regimen.6

ATG in unrelated transplants for CML patients M Schleuning et al

249 Table 8

Cox regression analysis for overall and event-free survival for all patients

Factor

Patient age Time to transplantation Recipient–donor sex combination Graft cell number Acute GVHD, grade II–IV Dose of cyclophosphamide Dose of ATG

Overall survival

Event-free survival

P

RR

95% CI

0.017 0.019

0.41 0.36

0.19–0.85 0.15–0.85

0.038 0.005

2.25 0.27

1.05–4.86 0.11–0.67

Therefore, the dose of ATG seems to be a critical factor, which so far has not been analyzed systematically for patients with CML. After a median follow-up of 5.8 years, the Kaplan–Meier estimate revealed significant differences for the probability of event-free (68 vs 39%) and overall survival (72 vs 52%) in favor of ATG doses of at least 60 mg/kg b.w. In multivariate analysis of overall survival, the effect of the ATG dose used was hidden by other important prognostic factors like time interval between diagnosis and transplantation or graft mononuclear cell number. However, Cox regression analysis of event-free survival proved that inadequate dosing of ATG represents an independent risk factor. In patients receiving less than 60 mg/kg b.w. of ATG, the risk of relapse or death was more than doubled. Additionally, we did a subgroup analysis for those patients transplanted within 2 years of diagnosis. In univariate analysis with Kaplan–Meier estimates, the effects of the ATG dose were highly significant for event-free and overall survival giving excellent survival rates (ES 88%, OS 90%) with the higher ATG dose used. To the best of our knowledge, similar survival rates have not been reported for CML patients transplanted from unrelated volunteer donors so far. Interestingly, these differences could not be observed in the subgroup of patients transplanted later during their disease, indicating that the risk and severity of acute GVHD may be increasing and more difficult to control by duration of the disease. The incidence of chronic GVHD was with 32.9 and 11.3% extensive similar to the incidences reported from studies incorporating ATG,10 but lower than those reported by other investigators using standard GVHD prophylaxis (53–67%).2,3,18 Other strategies to reduce the incidence of acute GVHD in URD-BMT including T cell depletion of the graft with the T10B9 monoclonal antibody5 or ex vivo or in vivo T cell depletion with Campath-1M or -1G also resulted in less acute GVHD but resulted in graft failure rates up to 16% at 1 year19 and higher relapse rates.20 By contrast, in our study, we only observed one case of graft failure. Indeed, T cell depletion of the graft appeared to be a negative predictive factor for event-free survival in a survey from the European Group for Blood and Marrow Tranplantation.21 ATG has been used in marrow transplantation since nearly 30 years. In the early years, no benefit was reported in canine studies or in matched sibling transplantations for patients with leukemia.22,23 However, there are different preparations of anti-T lymphocyte globulins from various animal sources available. The specificity of these different

P

RR

95% CI

0.018 0.007

0.25 0.37

0.08–0.79 0.18–0.76

0.001 0.107

0.18 2.26

0.06–0.50 0.84–6.11

preparations is not comparable. The ATG used in our study is derived from rabbits immunized with the T lymphoblastic leukemic Jurkat cell line. Since the immunophenotype of this cell line differs from thymocytes, the activity of the purified immunoglobulin is probably not comparable to the activity of preparations that are derived from animals immunized with fetal thymic tissue. These preparations may have a much broader spectrum of activity, thus possibly resulting in higher rates of graft failures, relapses and infections. In particular, alemtuzumab has been reported to be associated with an exceptionally high risk of CMV reactivation (85%), whereas in our study the rate of CMV reactivations (27%; Table 1) was comparable to preparative regimens without in vivo T cell depletion.24 In addition, as mentioned above, the dose of ATG seems to be a critical factor. In an earlier study, we observed a higher relapse rate in high-risk patients receiving 80 mg/kg b.w. or more of ATG.25 In the current study with low-risk CML patients, we could not observe similar results regarding the relapse rates. Nevertheless a dose of 60 mg/kg b.w. seems to be sufficient to ensure engraftment and reduce acute GVHD to an acceptable level in unrelated transplants for CML. ATG is expected to enhance T cell depletion in combination with TBI and cyclophosphamide in recipients of allogeneic grafts prior to stem cell transplantation and also to deplete in vivo the donor T cell inoculum. As free rATG is still detectable up to 30 days after infusion, it might further contribute to the reduction of acute GVHD. However, our data provide evidence that recovery of T cells occurs even in the presence of measurable rATG serum levels. Thus it seems questionable if ATG still remains functional at these time points. The mechanism(s) by which ATG applied during the preparative regimen reduces the incidence of acute GVHD, therefore, remains to be determined. In conclusion, allogeneic BMT from unrelated volunteer donors offers an excellent treatment option to patients with chronic phase CML, provided it is performed early during disease and a sufficient amount of ATG is included in the preparative regimen. The notion that it may safely be delayed for first conducting a trial with interferon-a or imatinib mesylate is not supported by our data.

References 1 Hehlmann R. Trial of IFN or STI571 before proceeding to allografting for CML? Leukemia 2000; 14: 1560–1562. Bone Marrow Transplantation

ATG in unrelated transplants for CML patients M Schleuning et al

250 2 Szydlo R, Goldman JM, Klein JP et al. Results of allogeneic bone marrow transplants for leukemia using donors other than HLA-identical siblings. J Clin Oncol 1997; 15: 1767–1777. 3 McGlave PB, Shu XO, Wen W et al. Unrelated donor marrow transplantation for chronic myelogenous leukemia: 9 years’ experience of the national marrow donor program. Blood 2000; 95: 2219–2225. 4 Nademanee A, Schmidt GM, Parker P et al. The outcome of matched unrelated donor bone marrow transplantation in patients with hematologic malignancies using molecular typing for donor selection and graft-versus-host disease prophylaxis regimen of cyclosporine, methotrexate, and prednisone. Blood 1995; 86: 1228–1234. 5 Ash RC, Casper, JT, Chitambar CR et al. Successful allogeneic transplantation of T-cell-depleted bone marrow from closely HLA-matched unrelated donors. N Engl J Med 1990; 322: 485–494. 6 McGlave P, Bartsch G, Anasetti C et al. Unrelated donor marrow transplantation therapy for chronic myelogenous leukemia: initial experience of the National Marrow Donor Program. Blood 1993; 81: 543–550. 7 Apperley JF, Mauro FR, Goldman JM et al. Bone marrow transplantation for chronic myeloid leukaemia in first chronic phase: importance of a graft-versus-leukaemia effect. Br J Haematol 1988; 69: 239–245. 8 Horowitz MM, Gale RP, Sondel PM et al. Graft-versusleukemia reactions after bone marrow transplantation. Blood 1990; 75: 555–562. 9 Kottaridis PD, Milligan DW, Chopra R et al. In vivo CAMPATH-1H prevents graft-versus-host disease following nonmyeloablative stem cell transplantation. Blood 2000; 96: 2419–2425. 10 Kroger N, Zabelina T, Kruger W et al. Anti-thymocyteglobulin as part of the preparative regimen prevents graft failure and severe graft versus host disease (GvHD) in allogeneic stem cell transplantation from unrelated donors. Ann Hematol 2001; 80: 209–215. 11 Finke J, Bertz H, Schmoor C et al. Allogeneic bone marrow transplantation from unrelated donors using in vivo anti-T-cell globulin. Br J Haematol 2000; 111: 303–313. 12 Kolb HJ, Holler E, Ledderose G et al. Conditioning treatment with antithymocyte globulin (ATG) modifies graft-vs-host disease in recipients of marrow from unrelated donors. Blood 1995; 86(Suppl. 1): 952a (abstract). 13 Preville X, Nicolas L, Flacher M, Revillard J. A quantitative flow cytometry assay for the preclinical testing and pharmacological monitoring of rabbit antilymphocyte globulins (rATG). J Immunol Methods 2000; 245: 45–54. 14 Pihusch R, Holler E, Mu¨hlbauer D et al. The impact of antithymocyte globulin on short-terme toxicity after allogeneic

Bone Marrow Transplantation

15

16

17

18

19

20

21

22

23

24

25

stem cell transplantation. Bone Marrow Transplant 2002; 30: 347–354. Copelan EA, Penza SL, Theil KS et al. The influence of early transplantation, age, GVHD prevention regimen, and other factors on outcome of allogeneic transplantation for CML following BuCy. Bone Marrow Transplant 2000; 26: 1037–1043. van Rhee F, Szydlo RM, Hermans J et al. Long-term results after allogeneic bone marrow transplantation for chronic myelogenous leukemia in chronic phase: a report from the Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant 1997; 20: 553–560. Carreras E, Tomas JF, Sanz G et al. Unrelated donor bone marrow transplantation as treatment for chronic myeloid leukemia: the Spanish experience. The Chronic Myeloid Leukemia Subcommittee of the GETH. Grupo Espanol de Trasplante Hemopoyetico. Haematologica 2000; 85: 530–538. Hansen JA, Gooley TA, Martin PJ et al. Bone marrow transplants from unrelated donors for patients with chronic myeloid leukemia. N Engl J Med 1998; 338: 962–968. Spencer A, Szydlo RM, Brookes PA et al. Bone marrow transplantation for chronic myeloid leukemia with volunteer unrelated donors using ex vivo or in vivo T-cell depletion: major prognostic impact of HLA class I identity between donor and recipient. Blood 1995; 86: 3590–3597. Goldman JM, Apperley JF, Jones L et al. Bone marrow transplantation for patients with chronic myeloid leukemia. N Engl J Med 1986; 314: 202–207. Devergie A, Apperley JF, Labopin M et al. European results of matched unrelated donor bone marrow transplantation for chronic myeloid leukemia. Impact of HLA class II matching. Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant 1997; 20: 11–19. Kolb HJ, Storb R, Graham TC et al. Antithymocyte serum and methotrexate for control of graft-versus-host disease in dogs. Transplantation 1973; 16: 17–23. Weiden PL, Doney K, Storb R, Thomas ED. Antihuman thymocyte globulin for prophylaxis of graft-versus-host disease. A randomized trial in patients with leukemia treated with HLA-identical sibling marrow grafts. Transplantation 1979; 27: 227–230. Perez-Simon JA, Kottaridis PD, Martino R et al. Nonmyeloablative transplantation with or without alemtuzumab: comparison between 2 prospective studies in patients with lymphoproliferative dieorders. Blood 2002; 100: 3121–3127. Holler E, Ledderose G, Knabe H, Kolb HJ. ATG serotherapy during pretransplant conditioning in unrelated donor bmt: dose dependent modulation of GVHD. Bone Marrow Transplant 1998; 22 (Suppl. 1): 105 (abstract).