To the Editor:
We read with interest the report by Kawano et al1concerning the use of granulocyte colony-stimulating factor (G-CSF) after peripheral blood stem cell transplantation (PBSC) in pediatric patients. The main conclusions of this study are that filgrastim does not provide a marked clinical benefit in the setting of patients with more than a threshold level of PBSC infused, despite the existence of a significantly accelerated neutrophil recovery in the group of patients that receive G-CSF. Unfortunately, this effect does not imply fewer infections or lower the transfusion measures (only registered for red blood cells) for these patients. All of those facts, with an unfavorable trend for platelet engraftment in the group treated with G-CSF, move the investigators to recommend a more cautious use of this expensive treatment strategy in children.
Despite its indubitable interest (as one of the few randomized trials published to assess the use of G-CSF postransplantation using PBSC in which a nontreatment group has been included), in our opinion there are some points in the report that diminish its value. First, this multicentric study was developed in 14 different institutions, with a relatively small number of patients recruited for the different pathologies investigated. Moreover, the supportive measures were not uniform for all the centers involved, making difficult the interpretation of possible differences in incidence and type of infections and number of transfusions between the two groups. Second, the unfavorable trend for platelet engraftment observed for the patients treated with G-CSF could be irrelevant, if this fact does not imply a higher number and severity of hemorrhagic episodes with an increased number of platelet transfusions (not displayed in the report) for this group. Finally, other important items to analyze in this field could not be performed by this study due to the low number of patients recruited. Among other objectives, we suggest studies of the economics, the duration of hospitalization, the possible benefit of G-CSF in patients with lower threshold of PBSC infused, and the possible impact on incidence of mucositis or on survival.
In our hospital, to assess the use of G-CSF (Filgrastim) after autotransplantation, two consecutive prospective and randomized trials were performed over the last 3 years. Only adult patients with breast carcinoma, acute leukemia, or lymphoma were recruited. Chemotherapy conditioning (BuCy2, BEAC, or STAMP-V) and G-CSF mobilization for PBSC were used in all cases. No other cytokines were used. There were no statistical differences for age, sex, diagnosis, remission status, number of mononuclear and CD34+ cells infused by body weight, platelet count before transplantation, or conditioning used among the different groups. The observation period analyzed ran from PBSC infusion to day +100 posttransplantation. Multiple variables in the evolution of patients posttransplantation were recorded and analyzed to display any possible difference between the established groups.
In the first trial (from May 1994 to June 1997), the patients randomly received G-CSF (Filgrastim) at 5 μg/kg subcutaneously from day +2 (36 cases) or +5 (44 cases) after infusion of PBSC. No differences could be established between these two groups for aplasia time (granulocytes, <500 × 109/L), time to recover more than 20 × 109/L platelets, days with parenteral nutrition support, transfusional support, infectious complications, days of hospitalization, or number of days with fever and any other possible complications. A substantial economic benefit was obtained with a delayed use of G-CSF. These results are in accordance with those published previously by Faucher et al.2
In the second trial (from July 1997), the patients randomly received G-CSF: one group received G-CSF (Filgrastim) at 5 μg/kg subcutaneously from day +5 and the other received no cytokines. From the analysis of the data available at the time of this report, with a total of 62 patients recruited (30 with G-CSF and 32 without it), only minor differences regarding the use of G-CSF could be inferred. A faster granulocyte engraftment was evident in the group treated (mean of 10 v 12 days to achieve >0.5 × 109/L granulocytes; P < .001), without any differences between groups in incidence and severity of infections, days of fever, and duration of antibiotic treatment. As in the report of Kawano et al1 and that of Berstein et al,3 a subtle slower engraftment for platelets appeared in our series (mean of 15 days in the group with G-CSF v 12 days in the other group to achieve >20 × 109/L platelets; P = not significant), but without differences in incidence and severity of hemorrhages or platelet transfusion support.4
From our data it could be concluded that a delayed rather than an early administration of G-CSF post-PBSC transplantation is preferable (if used) and that there is no clear beneficial impact of its administration whatsoever.
We believe, as Kawano et al1 suggest for children, that the use of G-CSF post-PBSC transplantation could be marginal also for adults. Nevertheless, more ample randomized studies are necessary before condemning or supporting its use in this particular clinical setting.
Response
To the Editor:The word “transplantation” does not necessarily imply that targeted patients will always develop prolonged cytopenia, for which treatment with G-CSF is recommended. It must be stressed that de novo hematopoietic recovery kinetics after PBSCT versus those after bone marrow transplantation or after transplants with a minimum number of hematopoietic stem/progenitor cells versus after those with more than threshold numbers should be quite different. It is clear that there are age-dependent differences in stem cell kinetics and the quantity/quality of PBSC in grafts, which certainly affect the profile of the clinical response to G-CSF. In this regard, we thank Ojeda et al for revisiting the now widely accepted clinical procedure of using G-CSF after PBSC transplantation (PBSCT) by referring to our data published in a recent issue ofBLOOD.1-1 In that study, we reported the results of a prospective randomized trial with 74 children who were scheduled to undergo high-dose chemotherapy followed by autologous PBSCT. In the data reported by us with children and by Ojeda et al with adult patients, a 1- or 2-day earlier recovery of granulocytes did not have any meaningful practical clinical impact. Moreover, we and Ojeda et al found that the marginal clinical benefit of an earlier recovery of granulocytes could be offset by the delayed recovery of platelets, which has been well reported by other groups,1-2,1-3 although the exact mechanism of this phenomenon is unclear.
The quality of our multicenter study might be questioned. However, in a series of studies we have shown that morbidity and mortality related to PBSCT are essentially negligible, particularly in pediatric patients. This makes the biases that originate from differences in clinical management procedures among the participating institutes less significant. Indeed, in our intent-to-treat analysis, none of the patients suffered from serious bleeding episodes or clinically significant infectious complications. Hence, the value of extending the analysis to include minor items, such as mucositis, is questionable. All of the patients with ALL (n = 27) and neuroblastoma (n = 29) were treated uniformly according to the protocols of the Japanese Cooperative Study Group of PBSCT. Thus, it is unlikely that the quality of our multicenter trial and trivial differences in supportive care systems, if any, greatly affected the outcome of our study. We could not perform a critical cost analysis, because in Japan all medical costs related to cancer therapy for children are paid by the government, not by the family, and decisions regarding drug therapy or hospitalization depend strongly on social considerations.
There might be another pitfall with G-CSF treatment when used after transplantation, ie, the administration of G-CSF may further exaggerate cytokine dysregulation, which frequently occurs in the rapid recovery phase of hematopoiesis after PBSCT and is clinically recognized as engraftment syndrome. In addition, a sharp surge of granulocytes in the circulation may increase the trapping of cells in the microcirculation, particularly in pulmonary vessels. Consequently, the incidence and severity of pulmonary complications might be increased in adult patients, as has been reported after conventional chemotherapy.1-4 Considering these points, we agree that the routine application of this expensive strategy in patients undergoing PBSCT should be seriously reevaluated in adult patients. Reconsideration of this maneuver also has important implications for the cost-effectiveness of hematopoietic stem cell transplantation.
