Moreau et al reported a randomized trial comparing pretransplantation conditioning with 200 mg/m2 melphalan (HDM200) to 8 Gy total body irradiation (TBI) plus 140 mg/m2 melphalan (HDM140 + TBI) in patients with multiple myeloma.1 They found that the HDM140 + TBI regimen was associated with more severe mucositis, longer hospitalization, and worse overall survival. The durations of both neutropenia and thrombocytopenia were also significantly longer in the patients who received HDM140 + TBI. Moreau et al postulated that this “slower engraftment” 1,(p735) might be attributed to a greater degree of stromal damage caused by the TBI-containing regimen. Whereas it is tempting to speculate that such a differential effect on the marrow microenvironment may be operative, we doubt this to be the case. Marrow stromal/mesenchymal cells are relatively radiation resistant in comparison to hematopoietic cells, and it is unlikely that damage induced by TBI within 6 days of autograft reinfusion would have a profound influence on engraftment kinetics.2 Further, studies have demonstrated that marrow stromal cells are susceptible to damage by chemotherapy alone.3,4 Finally, 4 randomized studies comparing busulfan and cyclophosphamide to cyclophosphamide and TBI prior to allografting did not demonstrate any significant differences in the tempo of either neutrophil or platelet engraftment between the conditioning regimens.5 Taking these together, we suggest that, while HDM140 + TBI is more toxic than HDM200, the reason for this has little to do with the effects of TBI on the capacity of the marrow microenvironment to support hematopoiesis.
Moreau et al reported engraftment in terms of the overall duration of neutropenia and thrombocytopenia rather than the more conventionally reported median time to recovery of neutrophils and platelets following autograft reinfusion on day 0. These 2 disparate measures of engraftment should not be interpreted as interchangeable. The overall duration of cytopenia is a function both of the damage to residual recipient hematopoiesis caused by the conditioning regimen (time to nadir), as well as the quantity and quality of the hematopoietic stem and progenitor cells contained within the autograft (time to recovery). In contrast, the time to neutrophil and platelet recovery, measured from the day of autograft reinfusion, is a truer reflection of the quantity and quality of cells within the autograft, as well as the ability of the marrow stroma to support hematopoiesis. We speculate that the difference in the overall duration of neutropenia and thrombocytopenia observed by Moreau et al is a reflection of the variability in the effects of the 2 regimens on residual host hematopoiesis, particularly if one considers that the HDM140 + TBI regimen began on day − 6 while the HDM200 regimen began on day − 2. We have previously reported that up to 30% of patients given a conditioning regimen (including HDM200) that started within 4 days of autograft reinfusion experienced a transient leukocytosis (defined as a transient rise in leukocyte count of at least 1 × 109/L in the first 72 hours) following day 0.6 This was typically seen in patients who were also given a hematopoietic cytokine on or around day 0 and was never seen after administration of TBI. We speculate that this effect was due not to cells contained within the autograft but rather to maturing or demarginating recipient hematopoietic cells that survive when conditioning is administered so close to the autograft infusion, as in the study by Moreau et al. This phenomenon has previously been described by Arseniev et al in allograft recipients conditioned with a non–TBI containing regimen.7 We would be interested to know whether there was any difference in the tempo of neutrophil and platelet recovery measured from day 0 between the 2 treatment arms, as this would be a better reflection of any differential effect on stromal support of hematopoiesis. We suggest that in the Moreau et al trial, the median neutrophil and platelet counts on day 0 differed between the HDM140 + TBI and HDM200 recipients and that as a result, the HDM200 recipients experienced a shorter duration of pancytopenia.
Hematologic recovery after high-dose therapy for multiple myeloma
We thank Drs Devine, Jahagirdar, and van Besien for their interesting comments on the duration of neutropenia and thrombocytopenia following high dose melphalan conditioning and autologous transplantation in multiple myeloma. We confirm that, in our study, the median neutrophil and platelet counts on day 0 were lower in the total body irradiation (TBI) arm (TBI given over 4 days from day − 6) as compared with the melphalan alone arm (melphalan given on day − 2). But engraftment data were not calculated from the date of autologous transplantation; rather, they were calculated from the first day of neutropenia below 500/mL and of thrombocytopenia below 20 000/mL. The difference between the median duration of neutropenia in the 2 arms was 2 days (8 versus 10), and the median duration of thrombocytopenia was the same (7 days) in the 2 arms, but more patients in the TBI group experienced prolonged thrombocytopenia, giving a statistical advantage to the 200 mg/m2 melphalan (HDM200) arm. These engraftment kinetics are comparable to those reported previously. Fermand et al reported a median duration of neutropenia of 12 days (below 1000/mL) after peripheral blood stem cell (PBSC) transplantation performed as part of first-line therapy following a TBI-based conditioning regimen.1-1 In the total therapy program, Barlogie et al reported a median duration of neutropenia of one week in newly diagnosed patients after the first high-dose therapy, which consisted of 200 mg/m2melphalan followed by PBSC transplantation.1-2 The Spanish group recently updated the results of a retrospective study comparing HDM200 and HDM140 + TBI in patients with multiple myeloma.1-3 In this study, they showed that time to neutrophil recovery was the same in patients treated with HDM200 (11 days ± 1 day) as compared with HDM140 + TBI (11 days ± 7 days), but this variable was calculated from the day of infusion (day 0) when no patient was neutropenic in the HDM200 group, while a number of patients were already neutropenic in the TBI + HDM group. In this latter trial, the duration of hospitalization was significantly shorter in the HDM200 group.
The mechanisms for hematologic toxicity following high-dose therapy (HDT) are numerous and may involve the microenvironment, the residual host hematopoiesis, and the stem cell compartment as mentioned accurately by Devine et al in their letter.
