- © 2001 by American Society of Clinical Oncology
Is There an Association Between Total-Body Irradiation and Secondary Acute Myelogenous Leukemia/Myelodysplastic Syndrome in Patients With Relapsed/Refractory Hodgkin’s Disease Treated With Autologous Stem-Cell Transplantation?
To the Editor:
In the March 1, 2001 issue of the Journal of Clinical Oncology, Sureda et al from the Grupo Español de Linfomas/Transplante Autólogo de Médula Ósea (GEL/TAMO) Spanish Cooperative Group1 summarized their registry results of autologous stem-cell transplantation (ASCT) for 494 patients with Hodgkin’s disease. Fifty-one patients (10%) received a total-body irradiation (TBI)–based high-dose regimen and 443 patients received high-dose chemotherapy only. Sixteen patients developed a secondary malignancy at a median of 24.5 months after ASCT (four in the TBI arm; 12 in the chemotherapy arm). Multivariate analysis showed that the administration of TBI in the conditioning regimen predicted the development of secondary malignancy after ASCT (relative risk = 4.64; 95% confidence interval, 1.21 to 16.66; P = .01). On the basis of this analysis, despite only four cases of secondary malignancy occurring in the TBI arm, the authors concluded that TBI should be avoided in the conditioning regimen because of a significantly higher rate of late complications, including secondary malignancies.
In this study, the criteria for selecting a TBI-based regimen are not mentioned. In addition, pretransplantation exposure to alkylating agent dose or topoisomerase II inhibitor was not examined in this analysis. Furthermore, the length of follow-up for the two arms (TBI v non-TBI) was not reported. The pathophysiology of therapy-related myelodysplastic syndrome/acute myelogenous leukemia (MDS/AML) is currently under active investigation. It’s still unclear whether t-MDS/t-AML is related to pretransplantation chemotherapy and radiotherapy, transplantation conditioning, stem-cell priming regimens, or the source of stem cells or whether it is a cumulative effect of all these exposures. Hence, it’s important to include all these factors in the analysis to provide a more meaningful assessment. Because of the limitation of a registry-based retrospective study, it would be more appropriate to perform a nested case-controlled study to examine all the potential variables. In addition, in a nested case-controlled study, both study arms will also have the same amount of follow-up time.
Between 1986 and 1998, 218 patients with Hodgkin’s disease underwent ASCT at City of Hope.2 Eleven patients developed t-MDS/t-AML, with the estimated cumulative probability approaching 8.1% ± 2.5% at 6 years. The median time to development of t-MDS/t-AML was 3.8 years from diagnosis and 0.9 years from transplantation. Multivariate analysis revealed priming with etoposide to be the only independent factor associated with increased risk of t-MDS/t-AML (relative risk = 8.3; 95% confidence interval, 1.6 to 43.9; P = .01). A short interval between transplantation and MDS suggests that pretransplantation therapy contributes significantly to the subsequent development of t-MDS/t-AML. In our study, TBI was not found to be associated with increased risk of developing secondary leukemia/MDS.
Two other studies have demonstrated an important role of prior therapy in the development of secondary AML/MDS. In a retrospective analysis of morphologically normal pretransplantation marrow or stem-cell specimens from 12 patients who subsequently developed t-MDS, the cytogenetic abnormality observed at the time of t-MDS was the same as those obtained from the pretransplantation specimens using fluorescence in situ hybridization techniques.3 At our institution, all patients undergo cytogenetic study before autologous stem-cell collection, and those with cytogenetic abnormalities are excluded from the transplant. Furthermore, Harrison et al4 performed a case-controlled study of 4,576 patients with Hodgkin’s disease from the British National Lymphoma Investigation and University College Medical School, London database, which includes 595 patients who had undergone ASCT. This analysis identified quantity of prior therapy and exposure to the mechlorethamine, vincristine, procarbazine, and prednisone combination or lomustine as risk factors in developing subsequent MDS/AML, whereas ASCT is not significant.
In patients with non-Hodgkin lymphoma, TBI has been associated with t-MDS/t-AML in multiple studies.5,6 However, the increased risk of secondary MDS/AML for patients with follicular lymphoma could be related to treatment before ASCT, in particular, repeated courses of alkylating agent.7 As for patients with Hodgkin’s disease, the study from GEL/TAMO is the only study that has reported a possible association of TBI with therapy-induced MDS/AML. All other Hodgkin’s disease transplantation series, including our cohort of 218 patients,2 do not find such an association.
At City of Hope, all patients with Hodgkin’s disease undergoing ASCT received the combination of fractionated TBI, etoposide, and cyclophosphamide as conditioning regimen,8,9 unless precluded by previous irradiation. Between 1987 and 1997, 70 patients with refractory or recurrent Hodgkin’s disease underwent ASCT by using a TBI-based conditioning regimen.10 The 3-year actuarial probability of overall survival, disease-free survival, and relapse was 76%, 81%, and 22%, respectively. Patients who received a TBI-based regimen tend to have an improved outcome, but this is not statistically significant. Interestingly, all relapses occurred within the first 18 months after transplantation. Similar results were reported from a Southwest Oncology Group multicenter study.11
In summary, the current literature does not provide any firm evidence regarding the association between TBI and the subsequent development of secondary MDS/AML after ASCT. Given the small number of events and relatively short follow-up period, the conclusion derived from the GEL/TAMO’s data should be interpreted with caution.1 On the other hand, given the encouraging results from our institution and other phase II studies,12 TBI-based regimens should continue to be explored, and in particular, for those patients who are at high risk of relapse after transplantation.
References
Response
- Anna Sureda,
- Eulogio Conde and
- For the Grupo Español de Linfomas/Transplante Autólogo de Médula Ósea Spanish Cooperative Group
In Reply:
Fung et al have raised the outstanding concern of predisposing factors for secondary malignancies (SM) in patients with Hodgkin’s disease (HD) undergoing autologous stem-cell transplantation (ASCT). In our multicenter retrospective analysis, we found the use of total-body irradiation (TBI) within the conditioning regimen to be a significant predictive factor for developing an SM, namely a myelodysplastic syndrome/acute myelogenous leukemia (MDS/AML) after transplantation, along with an increased age (≥ 40 years) and the use of adjuvant radiotherapy (RT) pretransplantation.1 Although the proportion of patients receiving TBI is only 10% of the total series and the number of events after transplantation is rather low (16 SMs, 12 in the chemotherapy arm and four in the TBI arm), the administration of TBI increased more than four times the risk of developing an SM after transplantation (relative risk = 4.64; 95% confidence interval, 1.31 to 16.66; P = .01). It was not possible to obtain from participating centers the reason for using TBI as part of the conditioning regimen. Nevertheless, other questions pointed out by Fung et al have been addressed in this study; median follow-up between TBI and non-TBI treated patients is similar (32 [range, 0 to 139] months v 30 [range, 0 to 137] months; P = .62). The exact amount of alkylating agents or topoisomerase II inhibitors could not be evaluated because the information was not available in a significant proportion of patients, but in our hands, prior administration of nitrogen mustard was not a significant prognostic factor when included in the multivariate analysis. We also analyzed the impact of the number of treatment blocks received by the patients, considering the transplantation conditioning regimen as a treatment block in the development of SM after ASCT; in the multivariate analysis, there was a nonsignificant trend to develop an SM in patients receiving more than two treatment blocks with respect to those receiving one or two.
MDS/secondary leukemia, a complication of HD treatment known to be associated with cumulative doses of alkylating agents and possibly further promoted by wide-field radiotherapy, is emerging as a major cause of morbidity and mortality in patients successfully treated with high-dose therapy and autografting. The crucial issue is whether the occurrence of MDS/secondary leukemia posttransplantation is due to the high-dose therapy per se or is merely indicative of the fact that patients who undergo transplantation have an increased risk because of their previous, often extensive, therapy. Several authors have addressed this issue, although the exact role of TBI is somewhat difficult to assess, probably because the vast majority of HD patients are autografted using chemotherapy protocols but not the combination of TBI and chemotherapy. In the series of Harrison et al,2 ASCT therapy did not have a significant impact on the development of SM after transplantation, but TBI as part of the conditioning regimen was not evaluated because almost all patients were conditioned with the BEAM protocol (carmustine 300 mg/m2 intravenously [IV] on day 1, etoposide 200 mg/m2 IV on days 2 through 5, cytarabine 200 mg/m2 twice daily on days 2 through 5, and melphalan 140 mg/m2 IV on day 6). In their nested case-controlled study including a group of 218 patients with HD,3 the City of Hope group found that priming with etoposide (relative risk = 8.3; 95% confidence interval, 1.6 to 43.9; P < .05) was the only significant prognostic factor for developing an MDS/secondary leukemia after transplantation. However, the number of patients who received TBI in the conditioning regimen was not specified. Moreover, the impact of priming with etoposide on the incidence of MDS/secondary leukemia in this series has probably obscured a possibly deleterious effect of previous chemotherapy treatments or even the effect of the conditioning regimen.
TBI has been specifically considered as a contributing factor to the development of MDS/secondary leukemia after ASCT. Darrington et al4 found an increased risk of secondary MDS in non-Hodgkin’s lymphoma patients ≥ 40 years of age who received TBI in the conditioning regimen; Milligan et al5 in a retrospective analysis from the European Bone Marrow Transplantation Group also found that TBI was a significant prognostic factor for the development of a secondary myelodysplasia. Interestingly, TBI was also a weak risk factor in the multivariate analysis for HD patients until the time interval between diagnosis and ASCT was included in the model. This may suggest that more heavily pretreated HD patients were selected for TBI conditioning. In our study, time interval between diagnosis and transplantation was only a weak factor that lost its significance after including conditioning regimen in the protocol; moreover, as stated above, the number of treatment blocks received before transplantation had only a marginal impact on the incidence of myelodysplasia after ASCT in our study.
In conclusion, there are no definitive data in the literature to ascertain the real impact of pretransplantation chemotherapy or conditioning regimen in the development of SM after ASCT. In a significant proportion of patients with HD, the use of TBI is precluded because of prior radiation therapy. Although the use of TBI has demonstrated its efficacy in terms of disease-free survival and overall survival in some subgroups of patients,6,7 on the basis of the results of our study, we would not recommend the routine use of TBI outside of specific clinical trials.
