Outcomes after HLA-matched sibling transplantation or chemotherapy ...

Leukemia (2008) 22, 281–286 & 2008 Nature Publishing Group All rights reserved 0887-6924/08 $30.00 www.nature.com/leu

ORIGINAL ARTICLE Outcomes after HLA-matched sibling transplantation or chemotherapy in children with acute lymphoblastic leukemia in a second remission after an isolated central nervous system relapse: a collaborative study of the Children’s Oncology Group and the Center for International Blood and Marrow Transplant Research M Eapen1,2, M-J Zhang1, M Devidas3,4, E Raetz5, JC Barredo6, AK Ritchey7, K Godder8, S Grupp9, VA Lewis10, K Malloy1, WL Carroll5, SM Davies11 and BM Camitta2 1

Department of Medicine, Center for International Blood and Marrow Transplant Research, Medical College of Wisconsin, Milwaukee, WI, USA; 2Department of Pediatrics, Hematology, Oncology and Transplantation, Medical College of Wisconsin, Milwaukee, WI, USA; 3Statistical Center, The Children’s Oncology Group, National Childhood Cancer Foundation, Arcadia, CA, USA; 4Department of Statistics, University of Florida, Gainsville, FL, USA; 5Department of Pediatrics, Hematology and Oncology, New York University Medical Center, New York, NY, USA; 6Department of Pediatrics, Hematology, Oncology and Stem Cell Transplantation, University of Miami Miller School of Medicine, Miami, FL, USA; 7Department of Hematology and Oncology, Children’s Hospital of Pittsburgh, Pittsburgh, PA, USA; 8Department of Pediatrics, Hematology and Oncology, Medical College of Virginia Hospital, Richmond, VA, USA; 9Department of Hematology and Oncology, Children’s Hospital of Philadelphia, Philadelphia, PA, USA; 10Department of Pediatrics, Southern Alberta Children’s Cancer Center, , Calgary, AB, Canada and 11 Department of Hematology and Oncology, Children’s Hospital Medical Center, Cincinnati, OH, USA

In children with acute lymphoblastic leukemia (ALL) with isolated central nervous system (CNS) relapse and a human leucocyte antigen (HLA)-matched sibling, the optimal treatment after attaining second remission is unknown. We compared outcomes in 149 patients enrolled on chemotherapy trials and 60 HLA-matched sibling transplants, treated in 1990–2000. All patients achieved a second complete remission. Groups were similar, except the chemotherapy recipients were younger at diagnosis, less likely to have T-cell ALL and had longer duration (X18 months) first remission. To adjust for time-totransplant bias, left-truncated Cox’s regression models were constructed. Relapse rates were similar after chemotherapy and transplantation. In both treatment groups, relapse rates were higher in older children (11–17 years; RR 2.81, P ¼ 0.002) and shorter first remission (o18 months; RR 3.89, Po0.001). Treatment-related mortality rates were higher after transplantation (RR 4.28, P ¼ 0.001). The 8-year probabilities of leukemiafree survival adjusted for age and duration of first remission were similar after chemotherapy with irradiation and transplantation (66 and 58%, respectively). In the absence of an advantage for one treatment option over another, the data support use of either intensive chemotherapy with irradiation or HLA-matched sibling transplantation with total body irradiation containing conditioning regimen for children with ALL in second remission after an isolated CNS relapse. Leukemia (2008) 22, 281–286; doi:10.1038/sj.leu.2405037; published online 22 November 2007 Keywords: central nervous system relapse; acute lymphoblastic leukemia; chemotherapy; bone marrow transplantation; leukemiafree survival

(ALL) has led to significant decreases in meningeal relapse. Nevertheless, approximately 2–10% of children with ALL undergoing risk-directed therapy experience isolated CNS relapse.1–4 Until recently, the outcome for children with isolated CNS relapse was poor. Two recent chemotherapeutic trials conducted by the Pediatric Oncology Group (POG) that focused on allowing initial intensification of chemotherapy with drugs known to have effective penetration of the blood–brain barrier and delayed cranial or craniospinal irradiation report 4-year overall event-free survival rates of approximately 70%.2,5 Others have utilized hematopoietic stem cell transplantation as a strategy to improve event-free survival after isolated CNS relapse with mixed results.1,6–9 To date there are no randomized studies comparing these very different treatment approaches and only one8 that retrospectively compared outcomes after chemotherapy and transplantation utilizing matched-pair analysis. However, this report included only 10 patients with isolated extramedullary relapse. Although the best approach for treatment remains uncertain, there is general agreement that when relapse occurs early (o18 months from diagnosis) the outcome of either treatment approach is disappointing. The aim of this report was to identify the role of allogeneic transplantation (if any) in children with isolated CNS ALL treated on risk-directed front-line and relapse therapies in the current era. Thus we compared outcomes in 149 patients enrolled on two recent POG chemotherapy trials2,5 and 60 human leucocyte antigen (HLA)-matched sibling transplant recipients reported to the Center for International Blood and Marrow Transplant Research (CIBMTR) during the same period.

Introduction The incorporation of pre-emptive central nervous system (CNS) therapy as part of treatment for acute lymphoblastic leukemia Correspondence: Dr M Eapen, Center for International Blood and Marrow Transplant Research, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226, USA. E-mail: [email protected] Received 30 August 2007; revised 8 October 2007; accepted 17 October 2007; published online 22 November 2007

Patients, materials and methods

Data collection Patients who received chemotherapy were enrolled on POG clinical trials and detailed demographic and clinical data were obtained from POG statistical center. Data on patients undergoing transplantation were obtained from the statistical center of CIBMTR. CIBMTR is a working group of more than 500 transplant centers worldwide. Participating centers register basic

ALL in second CR after isolated CNS relapse M Eapen et al

282 information on consecutive transplantations to a statistical center at the Medical College of Wisconsin. Detailed demographic and clinical data are collected on a representative sample of patients in the registry using a weighted randomization scheme and all patients followed longitudinally. The Institutional Review Board of the Medical College of Wisconsin approved the study.

Inclusion criteria The study included patients with ALL, younger than 18 years at diagnosis and isolated CNS relapse (as determined by morphology) and attained second clinical remission. All patients received treatment between 1990 and 2000. Chemotherapy recipients were enrolled on POG trials 9061 (n ¼ 79) and 9412 (n ¼ 70). All transplant recipients (n ¼ 60) had an HLA-matched sibling donor and were transplanted in 35 centers. Patients who failed to attain a second remission or had a second relapse prior to transplantation (for transplant recipients) or the median time to transplantation (for chemotherapy recipients) were excluded.

End points Any death occurring during continuous remission was defined as treatment-related mortality. Relapse was defined as morphologic leukemia recurrence at any site and leukemia-free survival defined as survival in a state of continuous complete remission.

Treatment regimens Chemotherapy recipients were treated on POG trials 9061 or 9412.2,5 The first 6 months of therapy on POG 9061 and 9412 were similar and included the following: four-drug induction regimen with weekly triple intrathecal therapy for 4 weeks and consolidation/intensification from weeks 5 to 22 (intensive systemic therapy and emphasized drugs with better CNS penetration plus triple intrathecal therapy). Thereafter, children enrolled on POG 9061 received craniospinal irradiation (24 Gy to the cranium and 15 Gy to the spine) with weekly vincristine, daily dexamethasone and thrice weekly asparaginase for 3 weeks. POG 9412 differed in that intensive systemic therapy plus triple intrathecal therapy continued for a further 6 months (weeks 23–50) followed by cranial or craniospinal irradiation beginning at week 51. The radiation site and radiation dose was tailored such that children with short duration first remission (o18 months) received a dose similar to POG 9061 and for those with a longer duration first remission irradiation was limited to the cranium at a dose of 18 Gy. All patients received vincristine, asparaginase and dexamethasone during this period as per POG 9061. Continuation therapy for both protocols included daily 6-mercaptopurine and weekly intramuscular methotrexate for 6 weeks alternating with 4-weekly vincristine and intravenous cyclophosphosamide for 72 weeks. The decision to offer transplantation and all aspects of transplant regimens were at the discretion of transplant centers. All patients received T-cell replete bone marrow grafts from an HLA-matched sibling. Fifty patients (83%) received total body irradiation (TBI) containing conditioning regimens and 22, additional craniospinal radiation (as therapy for relapse) prior to transplantation. Among the 10 patients who did not receive TBI, 9 received craniospinal irradiation as part of relapse therapy prior to transplantation. Fifty-four patients (90%) received cyclosporine alone or cyclosporine and short course methotrexate. The remaining 10 patients (10%) received methotrexate with steroids or anti-thymocyte globulin. Twenty-five patients (42%) were cytomegalovirus seropositive at transplantation. Leukemia

Statistical analysis Variables related to patient and disease characteristics were compared among the groups using the w2 statistic for categorical variables and the Kruskal–Wallis test for continuous variables. The probabilities of overall and leukemia-free survival were calculated using left-truncated Kaplan–Meier estimator.10 For analysis of survival rates, death from any cause was considered an event, and surviving patients were censored at the time of last follow-up. For leukemia-free survival, relapse or death, (that is, treatment failure) were considered as events, and surviving patients were censored at the time of last follow-up. The probabilities of treatment-related death and relapse were calculated using a left truncated cumulative incidence function method.10 For treatment-related death, relapse was the competing event and for relapse, treatment-related death was the competing event. Surviving patients were censored at last follow-up. Confidence intervals (CI) were calculated using log transformation.10 The primary objective of the analysis was to compare outcomes after chemotherapy with irradiation and transplantation. When comparing outcomes of two different treatment groups, adjustments are required for differences in time-totreatment. This results from the fact that children must survive in complete remission commonly for 2–3 months to be eligible for a transplant, possibly selecting a better prognosis group. Therefore, left-truncated Cox’s proportional hazards regression models were constructed for all outcomes of interest.11 Multivariate models were built using a stepwise forward selection, with a P-value of 0.05 or less considered to indicate statistical significance. Results are expressed as relative risk (RR), that is, the relative occurrence of the event with chemotherapy as compared with transplantation. The variable for treatment type (chemotherapy versus transplantation) was retained in all steps of model building. Other variables considered were age at diagnosis (p10 versus 410 years), sex, white blood cell at diagnosis (WBC) count (p100 versus 4100 109 per liter), National Cancer Institute (NCI) risk score (good risk: age o10 years and WBC o50 109 per liter versus poor risk: all other categories of age and WBC) and time to first relapse (o18 versus X18 months). Transplant center effect on outcome by geographic region was examined and there were none.12 P-values are two-sided. SAS software, version 9.1 (SAS Institute, Cary, NC, USA) was used for analyses.

Results

Patients Table 1 shows the characteristics of patients and their disease. Chemotherapy recipients were younger, more likely to have B-precursor ALL and longer durations of first remission (418 months). The median durations of first remission were 26 and 15 months in the chemotherapy with irradiation and transplantation groups, respectively. The median time to transplantation after achieving a second remission was 2.5 months (range o1–8). All transplant recipients were in second remission and chemotherapy recipients maintained their duration of second remission to at least the median time to transplantation. The median follow-ups of the study population were 8 years after chemotherapy with irradiation and 9 years after transplantation.

Relapse After adjusting for age at diagnosis and duration of first remission, rates of second leukemia relapse were similar after chemotherapy and transplantation (Table 2). In both treatment groups, a second relapse was more likely in older (11–17 years)


283 children (RR 2.81, 95% CI 1.49–5.33, P ¼ 0.002) and those with early (o18 months) first relapse (RR 3.85, 95% CI 2.05–7.23, Po0.001). The 8-year probabilities of second relapse after Table 1

Characteristics of patients and their disease Chemotherapy Transplanta

Variables

N (%)

N (%)

149

60

Age at diagnosis (years)b o5 5–10 11–17

84 (56) 45 (30) 20 (13)

19 (32) 23 (38) 18 (30)

0.002

Male

96 (64)

38 (63)

0.881

139 (93) 10 (7)

46 (77) 14 (23)

Number of patients

ALL subtype B precursor T cell

o0.001

WBC at diagnosis, 109 lL o100 4100 Unknown

88 (59) 20 (13) 41 (28)

43 (72) 13 (22) 4 (7)

NCI risk groupc Good risk Poor risk Unknown

61 (41) 47 (32) 41 (28)

28 (47) 28 (47) 4 (6)

45 (30) 104 (70)

31 (52) 29 (48)

Time to 1st relapse o18 months X18 months

P-value

chemotherapy with irradiation and transplantation for children with early first relapse were 40 and 50%, respectively (Figure 1). Corresponding probabilities for children with late first relapse were 13 and 10% (Figure 1). Second leukemia recurrence after chemotherapy with irradiation and transplantation were equally likely to involve bone marrow and CNS. Cure after a second leukemia relapse was poor; 2 of 17 patients after transplantation and 3 of 31 patients after chemotherapy are alive. Among transplant recipients, we examined the impact of conditioning regimen (TBI-containing versus non-TBI regimens) on second leukemia relapse. Seven of 10 recipients of non-TBI conditioning regimens relapsed compared to 10 of 50 recipients of TBI-containing regimens. In subset analysis restricted to recipients of TBI-containing regimens, second leukemia relapse rates were similar after chemotherapy with irradiation and TBI-containing conditioning regimens (RR 1.19, 95% CI 0.68– 2.11, P ¼ 0.533).

Treatment-related mortality 0.477

0.429

0.004

Abbreviations: ALL, acute lymphoblastic leukemia; NCI, National Cancer Institute; WBC, white blood cell count. a Transplantations were performed in 35 centers (North American: 17 centers, 25 patients; South America: 4 centers, 5 patients; Western Europe: 9 centers, 17 patients; Australia/New Zealand: 3 centers, 9 patients and the Middle East: 2 centers, 4 patients. b Two patients in the chemotherapy group were o1 year at diagnosis. All transplant recipients were older than 1 year at diagnosis. c NCI risk group: Age and WBC at diagnosis criteria used to define good risk group (children 1–o10 years with WBC o50 000/mm3) and poor risk group (all others) for both B-precursor and T-cell ALL.

Compared to chemotherapy recipients, transplant recipients were more likely to die from a treatment-related complication (RR 3.19, 95% CI 1.45–7.02, P ¼ 0.004). The 8-year probabilities of treatment-related mortality were 9 and 22% after chemotherapy with irradiation and transplantation, respectively. Among transplant recipients, the incidence of treatment-related deaths was similar after TBI-containing and non-TBI conditioning regimens. Treatment-related mortality rates after transplantation did not vary by geographic location of center (P ¼ 0.9110).

Leukemia-free survival Treatment failure rates (that is, second leukemia relapse or death) were similar after chemotherapy with irradiation and transplantation (Table 2). In both treatment groups, failure rates were higher in older (11–17 years) children (RR 2.48, 95% CI 1.47–4.19, Po0.001) and those with early (o18 months) first relapse (RR 2.34, 95% CI 1.44–3.82, Po0.001). The 8-year probabilities of leukemia-free survival for the entire cohort, adjusted for age and duration of first remission were 66 and 58% after chemotherapy with irradiation and transplantation, respectively (Figure 2). Leukemia-free survival rates after transplantation did not vary by geographic location of center (P ¼ 0.2006).

Table 2 Relative risks of second relapse, treatment-related mortality, treatment failure and overall mortality after chemotherapy and HLAmatched sibling transplants Outcome

N1/N2

Relative risk 95% confidence interval

P-value

Relapse Chemotherapy HLA-matched sibling transplant

31/149 17/60

1.00 1.09 (0.58–2.05)

0.799

Treatment failure (inverse of leukemia-free survival)b Chemotherapy HLA-matched sibling transplant

45/149 29/60

1.00 1.53 (0.93–2.53)

0.095

Overall mortalityc Chemotherapy HLA-matched sibling transplant

42/149 27/60

1.00 1.48 (0.88–2.48)

0.140

a

Abbreviations: HLA, human leucocyte antigen; N1 ¼ number of events; N2 ¼ total number evaluable. Relapse: 31 of 149 patients in the chemotherapy group and 17 of 60 patients in the transplantation group had a second leukemia recurrence. b Treatment failure (inverse of leukemia-free survival, relapse or death): 45 of 149 patients (N ¼ 31 with second leukemia recurrence and N ¼ 14 died from non-relapse causes) in the chemotherapy group and 29 of 60 patients (N ¼ 17 with leukemia recurrence and N ¼ 12 died from non-relapse causes) failed treatment. c Overall mortality: 42 of 149 patients in the chemotherapy group and 27 of 60 patients in the transplantation group died from second leukemia recurrence or other causes. a

Leukemia


284

Figure 1 The probabilities of second leukemia recurrence after chemotherapy and transplantation against duration of first remission. The 2-, 5- and 8-year probabilities of recurrence after chemotherapy are 31, 40 and 40%, respectively, after an early first relapse. Corresponding probabilities after transplantation are 47, 50 and 50%. The 2-, 5- and 8-year probabilities of recurrence after chemotherapy are 6, 13 and 13%, respectively, after late first relapse. Corresponding probabilities after transplantation are 10, 10 and 10%.

Figure 2 The probabilities of leukemia-free survival after chemotherapy and transplantation, adjusted for age and duration of first remission. The 2-, 5- and 8-year probabilities of leukemia-free survival after chemotherapy are 77, 66 and 66%, respectively. Corresponding probabilities after transplantation are 62, 60 and 58%.

We performed two subset analyses comparing leukemia-free survival rates after chemotherapy with irradiation and transplantation: (1) limited to children with B-precursor ALL and (2) limited to transplant recipients who received TBI conditioning regimens. Consistent with the main analysis, in patients with B-precursor ALL, the 8-year probabilities of leukemia-free survival adjusted for age and duration of first remission were similar after chemotherapy with irradiation and transplantation, 71 and 68%, respectively (Figure 3). Similarly, the 8-year adjusted probabilities of leukemia-free survival after chemotherapy with irradiation and transplantation with TBI-containing conditioning regimens were 67 and 65%, respectively.

Po0.001). The 8-year probabilities of overall survival adjusted for age and duration of first remission were 67 and 62% after chemotherapy with irradiation and transplantation, respectively. Recurrent leukemia was the commonest cause of death in both treatment groups and occurred in 28 of 42 deceased patients in the chemotherapy with irradiation group, and 15 of 27 deceased patients in the transplant group. Other causes of death after transplantation include graft-versus-host disease (GvHD; n ¼ 4), infection (n ¼ 5), interstitial pneumonitis (n ¼ 1) and adult respiratory distress syndrome (n ¼ 1). Other causes of death after chemotherapy with irradiation include infection (n ¼ 8), second neoplasm (n ¼ 5) and adult respiratory distress syndrome (n ¼ 1).

Overall survival Overall mortality rates were similar after transplantation and chemotherapy with irradiation (Table 2). In both treatment groups, mortality rates were higher in older (11–17 years) children (RR 2.29, 95% CI 1.33–3.94, P ¼ 0.003) and those with early (o18 months) first relapse (RR 2.32, 95% CI 1.40–3.82, Leukemia

Discussion Following reports of event-free survival rates of over 70% for children treated on two recent North American POG trials for isolated CNS relapse,2,5 the role of transplantation is less


285

Figure 3 The probabilities of leukemia-free survival after chemotherapy and transplantation (with TBI-containing regimens), adjusted for age and duration of first remission for patients with B-precursor ALL. The 2-, 5- and 8-year probabilities of leukemia-free survival after chemotherapy are 82, 71 and 71%, respectively. Corresponding probabilities after transplantation are 68, 68 and 68%.

clear.6,9,13 In this report, outcomes after HLA-matched sibling transplantation reported to a large observational database were compared to that after intensive chemotherapy with irradiation (the above-mentioned POG trials). There are several limitations to the approach adopted. The decision to offer transplantation was at the discretion of the transplant center and the availability of an HLA-matched sibling. Patient selection also differed between the two groups; transplant recipients were more likely to be older and had a shorter duration of first remission. Data on therapy prior to relapse are lacking and cytogenetic data available for few patients in both treatment groups, as treatment occurred in an era when cytogenetic testing, was not mandatory. Nevertheless, our ability to adjust for key risk factors made a controlled though not randomized comparison possible. Significant prognostic factors in the current analysis were older age and short duration of first remission. After adjusting for these factors, we observed similar rates of second leukemia relapse, overall and leukemia-free survival after intensive chemotherapy with irradiation regimens and transplantation. The absence of a leukemia-free survival advantage after transplantation observed in this report is consistent with the findings of Borgmann and colleagues but differ from others.1,6,7,13 There are important differences between the current report and others1,13 that support a role for transplantation. Messina and colleagues1 reported higher disease-free survival rates after autologous transplantation (n ¼ 19; 56%) compared to chemotherapy (n ¼ 41; 13%). In their report, disease-free survival rates after chemotherapy are inferior to that reported for POG trials 9061 and 9412, the comparison cohort in the current analysis. Improvements in chemotherapy outcomes in the more recent POG studies may explain the absence of significant differences in outcome between the two treatment groups in this report. Bordigoni and colleagues13 observed disease-free survival rates of approximately 70% after transplantation with a TBI regimen for relapsed ALL (n ¼ 31). This is similar to that reported for POG trials 9061 and 9412, and after transplantation in the current report for recipients of TBI-containing conditioning regimens, supporting our observation that the results of transplantation are similar to an intensive chemotherapy with irradiation regimen in the current era. We observed similar rates of second leukemia relapse after chemotherapy with irradiation and transplantation. This differs

from results observed after bone marrow relapse, where improved disease control can be achieved with transplantation when relapse occurs early.8,14 In the absence of a graft-versusleukemia effect in the CNS and higher rates of treatment-related deaths after transplantation, transplantation does not offer an additional advantage over an intensified chemotherapeutic approach with irradiation. As reported previously, among transplant recipients, TBI-based regimens play a critical role in preventing a subsequent leukemia relapse in ALL.14,15 Should transplantation be offered, it is preferred that the conditioning regimen chosen includes TBI. The results of transplantation using chemotherapy-only preparative regimens are surprisingly poor. It is possible that these patients were selected for this treatment based on prior irradiation to the CNS or received less CNS-directed therapy than children treated with chemotherapy and irradiation alone. Treatment-related mortality rates were high in both treatment groups. Chemotherapy recipients were subject to intensive systemic therapy and irradiation resulting in deaths from infections and second malignant neoplasm. Most nonrelapse deaths in the transplantation group were also due to infections or GvHD-related and likely occurred at a higher rate than expected as transplantations spanned over a decade with variable supportive care. Importantly, transplant outcomes did not vary by transplant center or geographic location of the center, a concern when using data reported to an observational database. Consistent with other reports,6,8 the duration of first remission was an important predictor of outcome. Children with early (p18 months) relapse fared poorly after either treatment. Older age at diagnosis was also associated with higher rates of second leukemia relapse, and consequently lower overall and leukemia-free survival in both treatment groups. Other variables such as WBC count at diagnosis, sex, NCI risk group and aspects of transplantation regimen such as GvHD prophylaxis and donorrecipient cytomegalovirus status did not affect treatment outcomes. The relatively small numbers of patients may explain our inability to observe an association between treatment outcome and NCI risk group rather than a lack of biological or clinical significance. Small numbers of patients with T-cell ALL in both treatment groups limited our ability to examine directly the effect of ALL immunophenotype on the outcome. However, subset analysis Leukemia


286 restricted to patients with B-precursor ALL showed similar leukemia-free survival rates after chemotherapy with irradiation and transplantation, adjusted for duration of first remission and age confirming that the similar outcomes observed in the current report is not due to the higher proportion of T-cell ALL patients in the transplantation group. These data support use of either an intensive chemotherapy regimen with irradiation similar to POG 9061 and 9412 or HLAmatched sibling transplantation with a TBI-containing conditioning regimen in children with ALL, in second remission after an isolated CNS relapse. CNS irradiation and TBI containing conditioning regimens are known risk factors for neurocognitive deficits, cardiopulmonary and renal toxicities, growth impairment and second cancers in long-term survivors.5,16–19 This is of particular concern with increasing numbers of long-term survivors and future prospective studies should focus on reducing craniospinal irradiation where appropriate, as demonstrated in POG 9412.5 Future prospective studies should explore using sensitive techniques for detection of minimal residual disease in patients with isolated extramedullary relapse20 to aid in identifying patients at higher risk of leukemia recurrence as well as those who might benefit from a less intense treatment regimen in an effort to lower treatment-related morbidity/ mortality and yet preserve or improve upon current rates of leukemia-free survival.

Acknowledgements The study was supported by Public Health Service Grant U24-CA76518-08 from the National Cancer Institute, the National Institute of Allergy and Infectious Diseases and the National Heart, Lung and Blood Institute and U10-CA098543 from the National Cancer Institute and the Department of Health and Human Services.

References 1 Messina C, Cesaro S, Rondelli R, Rossetti F, Locatelli F, Pession A et al. Autologous bone marrow transplantation for childhood acute lymphoblastic leukaemia in Italy. AIEOP/FONOP-TMO group. Italian association of paediatric haemato-oncology. Bone Marrow Transplant 1998; 21: 1015–1021. 2 Ritchey AK, Pollock BH, Lauer SJ, Andejeski Y, Barredo J, Buchanan GR. Improved survival of children with isolated CNS relapse of acute lymphoblastic leukemia: a pediatric oncology group study. J Clin Oncol 1999; 17: 3745–3752. 3 Silverman LB, Declerck L, Gelber RD, Dalton VK, Asselin BL, Barr RD et al. Results of Dana-Farber Cancer Institute Consortium protocols for children with newly diagnosed acute lymphoblastic leukemia (1981–1995). Leukemia 2000; 14: 2247–2256. 4 Gaynon PS, Trigg ME, Heerema NA, Sensel MG, Sather HN, Hammond GD et al. Children’s Cancer Group trials in childhood acute lymphoblastic leukemia: 1983–1995. Leukemia 2000; 14: 2223–2233. 5 Barredo JC, Devidas M, Lauer SJ, Billett A, Marymont M, Pullen J et al. Isolated CNS relapse of acute lymphoblastic leukemia treated with intensive systemic chemotherapy and delayed CNS radiation: a pediatric oncology group study. J Clin Oncol 2006; 24: 3142–3149.

Leukemia

6 Borgmann A, Hartmann R, Schmid H, Klingebiel T, Ebell W, Gobel U et al. Isolated extramedullary relapse in children with acute lymphoblastic leukemia: a comparison between treatment results of chemotherapy and bone marrow transplantation. BFM Relapse Study Group. Bone Marrow Transplant 1995; 15: 515–521. 7 Borgmann A, Schmid H, Hartmann R, Baumgarten E, Hermann K, Klingebiel T et al. Autologous bone-marrow transplants compared with chemotherapy for children with acute lymphoblastic leukaemia in a second remission: a matched-pair analysis. The BerlinFrankfurt-Munster Study Group. Lancet 1995; 346: 873–876. 8 Borgmann A, von Stackelberg A, Hartmann R, Ebell W, Klingebiel T, Peters C et al. Unrelated donor stem cell transplantation compared with chemotherapy for children with acute lymphoblastic leukemia in a second remission: a matched-pair analysis. Blood 2003; 101: 3835–3839. 9 Yoshihara T, Morimoto A, Kuroda H, Imamura T, Ishida H, Tsunamoto K et al. Allogeneic stem cell transplantation in children with acute lymphoblastic leukemia after isolated central nervous system relapse: our experiences and review of the literature. Bone Marrow Transplant 2006; 37: 25–31. 10 Klein JP, Moeschberger ML. Survival Analysis: Techniques of Censored and Truncated Data. 2nd edn. Springer-Verlag: New York, NY, 2003, 92–133. 11 Cox DR. Regression models and life tables. J R Stat Soc B 1972; 34: 187–200. 12 Andersen PK, Klein JP, Zhang MJ. Testing for centre effects in multi-centre survival studies: a Monte Carlo comparison of fixed and random effects tests. Stat Med 1999; 18: 1489–1500. 13 Bordigoni P, Esperou H, Souillet G, Pico J, Michel G, Lacour B et al. Total body irradiation-high-dose cytosine arabinoside and melphalan followed by allogeneic bone marrow transplantation from HLA-identical siblings in the treatment of children with acute lymphoblastic leukaemia after relapse while receiving chemotherapy: a Societe Francaise de Greffe de Moelle study. Br J Haematol 1998; 102: 656–665. 14 Eapen M, Raetz E, Zhang MJ, Muehlenbein C, Devidas M, Abshire T et al. Outcomes after HLA-matched sibling transplantation or chemotherapy in children with B-precursor acute lymphoblastic leukemia in a second remission: a collaborative study of the Children’s Oncology Group and the Center for International Blood and Marrow Transplant Research. Blood 2006; 107: 4961–4967. 15 Davies SM, Ramsay NK, Klein JP, Weisdorf DJ, Bolwell B, Cahn JY et al. Comparison of preparative regimens in transplants for children with acute lymphoblastic leukemia. J Clin Oncol 2000; 18: 340–347. 16 Socie G, Curtis RE, Deeg HJ, Sobocinski KA, Filipovich AH, Travis LB et al. New malignant diseases after allogeneic marrow transplantation for childhood acute leukemia. J Clin Oncol 2000; 18: 348–357. 17 Socie G, Stone JV, Wingard JR, Weisdorf D, Henslee-Downey PJ, Bredeson C et al. Long-term survival and late deaths after allogeneic bone marrow transplantation. Late effects working committee of the International Bone Marrow Transplant Registry. N Engl J Med 1999; 341: 14–21. 18 Brown RT, Maden-Swain A, Walco GA, Cherrick I, Ievers CE, Conte PM et al. Cognitive and academic late effects among children previously treated for acute lymphoblastic leukemia receiving chemotherapy as CNS prophylaxis. J Pediatr Psychol 1998; 23: 333–340. 19 Pui CH, Cheng C, Leung W, Rai SN, Rivera GK, Sandlund JT et al. Extended follow-up of long-term survivors of childhood acute lymphoblastic leukemia. N Engl J Med 2003; 349: 640–649. 20 Goulden N, Langlands K, Steward C, Katz F, Potter M, Chessels J et al. PCR assessment of bone marrow status in isolated extramedullary relapse of childhood B-precursor acute lymphoblastic leukaemia. Brit J Haematol 1994; 87: 282–285.