Hematopoietic Growth Factors How Long After Neutrophil Recovery Should Myeloid Growth Factors Be Continued in Autologous Hematop

Hematopoietic growth factors How long after neutrophil recovery should myeloid growth factors be continued in autologous hematopoietic stem cell transplant recipients?

A Verma, J Pedicano, S Triﬁlio, S Singhal, M Tallman, J Winter, S Williams, L Gordon, J Monreal and J Mehta

Hematopoietic Stem Cell Transplant Program, Division of Hematology/Oncology, The Feinberg School of Medicine, The Robert H. Lurie Comprehensive Cancer Center, Northwestern University, Chicago, IL, USA

Summary: The traditional definition of myeloid engraftment has been the first of 3 consecutive days with an absolute Growth factors are routinely used after autotransplanta- neutrophil count (ANC) of 0.5 Â 109/l or more, to ensure tion to accelerate hematopoietic recovery, and are sustained recovery. Thus, growth factors are usually continued until the absolute neutrophil count (ANC) is continued until ANC is X0.5 Â 109/l for 3 consecutive X0.5 Â 109/l on 3 consecutive days. Since ANCoften days. We have shown that the neutrophil count almost increases to very high levels with this strategy, we never declines below 0.5 Â 109/l on the 2 consecutive days discontinued growth factor on the first day ANCreached after having reached that level in blood stem cell autograft8 0.5 Â 109/l in 45 patients (Study Group), and compared and allograft9 recipients. Based upon this, we modified our their subsequent ANCto 108 historic controls who definition of myeloid engraftment to the first day with ANC received growth factor longer. While ANCon the day X0.5 Â 109/l; a change that is associated with some logistic after reaching 0.5 Â 109/l was comparable between benefits.8 At the same time, since the ANC actually groups, ANCon the third day was significantly higher in increased several-fold over the 2–3 days following the day the Control Group (2.3 vs 4.9 Â 109/l; P ¼ 0.0003). When of engraftment after autograft,8 we changed our routine compared to the first day, ANCin the Study Group was clinical practice to discontinuing growth factor the day the higher by a median of 140% on the third day and by ANC reached 0.5 Â 109/l rather than continuing for an 450% in the Control Group (P ¼ 0.0002). A significantly additional 2 days. higher proportion of patients experienced a decline in This retrospective analysis was undertaken to compare ANCafter the first day in the Study Group. However, the course of neutrophil recovery in patients who discon- only one patient in the Study Group became neutropenic tinued growth factor, with those who continued growth transiently and ANCrecovered spontaneously the next factor after the first day with ANC X0.5 Â 109/l. day. The incidence of fever and hospitalization were comparable. We conclude that growth factors can be discontinued after autotransplantation the day the ANC Patients and methods reaches 0.5 Â 109/l, without compromising neutrophil recovery. A total of 153 adult patients with malignant diseases Bone Marrow Transplantation (2004) 33, 715–719. autografted in the Hematopoietic Stem Cell Transplant doi:10.1038/sj.bmt.1704415 Program of the Division of Hematology-Oncology of the Published online 26 January 2004 Feinberg School of Medicine, Northwestern University, Keywords: autologous hematopoietic stem cell transplan- and the Northwestern Memorial Hospital between Febru- tation; engraftment; filgrastim; neutropenia; sargramostim ary 1998 and January 2003 were studied. All patients provided informed consent for the transplant procedures, and all research protocols were approved by the Institu- Myeloid colony-stimulating factors such as filgrastim (G- tional Review Board. CSF) and molgramostim (GM-CSF) are routinely used Electronic and paper records were reviewed to obtain all after autotransplantation to hasten neutrophil recovery.1,2 blood counts, whether performed in the hospital laboratory Growth factors are typically started as early as day 0 or as (in-patient), the clinic laboratory (outpatient), or elsewhere late as day 7, and are continued until neutrophil recovery.3–7 (usually immediately after discharge through home health Therefore, the duration of growth factor administration is agency services). Whenever multiple blood count results usually related to the definition of myeloid engraftment. were available for the same day, the morning blood count was included in the analysis and the remaining counts were not studied. ANC included segmented neutrophils as well as bands. Correspondence: Dr J Mehta, 676 N. St Clair Street, Suite 850, Chicago, IL 60611, USA; E-mail: [email protected] Table 1 shows the patient characteristics. The condition- Received 09 July 2003; accepted 07October 2003 ing regimens employed were at standard doses, and Published online 26 January 2004 included high-dose melphalan (plasma cell dyscrasias), Growth factor duration A Verma et al 716 busulfan-etoposide (acute myeloid leukemia), busulfan- (1–4), and were compared to the Study Group (n ¼ 45), cyclophosphamide (lymphoma), carmustine-etoposide-cy- which received no growth factor after ANC reached tarabine-melphalan (BEAM) (lymphoma, Hodgkin’s dis- 0.5 Â 109/l. The two groups were compared using the ease), cyclophosphamide-carboplatin-etoposide with total- Wilcoxon rank-sum test for continuous variables, and the lymphoid irradiation (Hodgkin’s disease), or cyclopho- w2 test or the Fisher’s exact test for categoric variables. sphamide-carboplatin-thiotepa (breast cancer). CD34 þ cell doses were calculated on the basis of the ideal body weight (IBW).10 Patients did not receive trimethoprim- Results sulfamethoxazole routinely in the immediate post transplant period. As Table 1 shows, there were some differences between groups. There was a significantly higher proportion Growth factor administration of myeloma patients and GM-CSF recipients in those stopping growth factors early. A significantly higher After the transplant, 136 patients received filgrastim (G- proportion of Study Group patients started the growth CSF) and 17patients received sargramostim (GM-CSF) at factor later. The CD34 cell dose was slightly higher the dose of 5–10 mg/kg, subcutaneously daily, until the þ and engraftment slightly slower for those stopping ANC reached 0.5 Â 109/l, the first day with ANC growth factor early, but these differences were not X0.5 Â 109/l being considered the day of engraftment. statistically significant. The actual growth factor doses The type of growth factor used, unless dictated by a used were comparable for both groups. All GM-CSF particular protocol, was at the discretion of the physician. recipients received 500 mg of the drug. G-CSF recipients in The growth factor dose was usually rounded off to the both groups received a median of 480 mg G-CSF (range nearest vial size to avoid wastage. Growth factors were 300–780). usually administered in the late afternoon or evening. Table 2 and Figure 1 illustrate the patterns of neutrophil Growth factor was discontinued when the ANC reached recovery. ANC was higher in the Study Group on the day 0.5 Â 109/l in 45 patients, all treated since May 2002. The of engraftment, but higher in the Control Group on the last dose of growth factor in these patients was the evening subsequent 2 days. The difference was particularly marked before ANC reached this threshold, and they did on the third day. The magnitude of ANC increase was also not receive any growth factor once ANC reached the greater in the Control Group. A significantly greater threshold. The remaining patients continued growth factor proportion of Study Group patients experienced a decline until ANC was X0.5 Â 109/l on 2 consecutive days (an in ANC on the second and third days. However, only in additional day of growth factor; n ¼ 37), on 3 consecutive days (2 additional days of growth factor; n ¼ 30), or 4–5 consecutive days (3–4 additional days of growth factor; n ¼ 41). Most of these patients were treated prior Table 1 Patient characteristics to May 2002. No growth factor after Growth factor after P All patients other than the 45 stopping the growth factor ANC 40.5 Â 109/l ANC 40.5 Â 109/l early were to have received 2 additional days of growth (Study Group) (Control Group) factor after reaching ANC X0.5 Â 109/l, allowing for 9 n 45 108 documentation of ANC X0.5 Â 10 /l for 3 consecutive Age (years) 53 (27–69) 53 (22–70) 0.89 days. However, this was not the case for two reasons. Those Male sex 25 (55%) 50 (45%) 0.29 stopping the growth factor after 1 additional day (n ¼ 37) did so mostly because an additional intra-day blood count Diagnosis 0.006 9 Acute leukemia 6 (13%) 10 (9%) done on the day ANC reached 0.5 Â 10 /l showed a further Lymphoma 16 (36%) 54 (50%) increase in ANC. Thus, the growth factor was stopped the Plasma cell 23 (51%) 30 (28%) next day after seeing a third consecutive ANC count (rather disorders than the third consecutive day) X0.5 Â 109/l. Those Other 0 (0%) 14 (13%) continuing growth factor beyond the planned 2 additional CD34+ cell dose days (n ¼ 41) usually did so for one of two reasons. For 106/kg IBW 5.4 (1.8–22.9) 4.2 (1.0–27.8) 0.10 some hospitalized patients, the order to discontinue growth o2 Â 106/kg 2 (4%) 9 (8%) 0.51 factor automatically on the third day of ANC X0.5 Â 109/l, IBW written as part of the routine admission order set, did not Growth factor 0.001 get implemented on the correct day. For some discharged G-CSF 34 (76%) 102 (94%) patients, Home Health nurses or the patients themselves GM-CSF 11 (24%) 6 (6%) continued administering growth factor when it should have been stopped. Growth factor o0.0001 start day 0 4 (9%) 49 (45%) Statistical analysis 5 36 (80%) 44 (41%) Other 5 (11%) 15 (14%) Patients receiving growth factor after ANC reached (usually 1) 0.5 Â 109/l were grouped together (n ¼ 108; Control Days to ANC 11 (9–16) 10 (8–23) 0.11 X0.5 Â 109/l Group), irrespective of the number of additional doses

Bone Marrow Transplantation Table 2 Neutrophil recovery patterns by duration of growth factor administration

Days of n ANC (109/l) (median, range) ANC change (%) (median, range) Decline in ANC (n, %) growth factor after ANC X0.5 Â 109/l Day 1 Day 2 Day 3 Day 1–2 Day 2–3 Day 1–3 Day 1–2 Day 2–3 Day 1–3

All patients 0 45 (a) 1.0 (0.5–3.3) 2.2 (0.3–10.5) 2.3 (0.8–11.2) +120 (À60 to +662) +13 (À58 to +1000) +140 (À52 to +1000) 7(16%) 18 (44%) 5 (12%) X1 108 (b) 0.8 (0.5–2.2) 2.7(0.5–13.8) 4.9 (0.8–22.0) +211 ( À25 to +860) +71 (À69 to +417) +450 (À39 to +3433) 4 (4%) 11 (12%) 3 (3%)

P 0.015 0.43 0.0003 0.029 0.068 0.0002 0.016 o0.0001 0.046

G-CSF-treated patients 0 34 (c) 1.1 (0.5–3.3) 2.5 (0.3–10.5) 2.3 (0.8–11.2) +142 (À60 to +662) À3(À58 to +1000) +150 (À52 to +1000) 5 (15%) 16 (52%) 3 (10%) Verma A duration factor Growth X1 102 (d) 0.8 (0.5–2.2) 2.7(0.5–13.8) 5.2 (0.8–22.0) +211 ( À25 to +860) +76 (À69 to +417) +443 (À39 to +3433) 4 (4%) 6 (6%) 2 (2%) tal et P 0.02 0.62 0.0004 0.16 0.08 0.001 0.043 o0.0001 0.095

Growth factor started on day 5 0 36 (e) 1.0 (0.5–3.3) 1.2 (0.2–10.5) 2.2 (0.8–11.2) +87( À60 to +662) +13 (À58 to +1000) +129 (À52 to +1000) 7(19%) 14 (42%) 5 (15%) X1 44 (f) 0.85 (0.5–2.2) 2.35 (0.5–13.8) 4.4 (0.8–22.0) + 185 (À25 to +860) +48 (À69 to +417) +448 (À39 to +3433) 2 (5%) 6 (15%) 1 (3%)

P 0.0070.85 0.006 0.017 0.82 0.002 0.0710.009 0.085

Missing ANC values on day 3 (reflected in the numbers in the last two columns under ‘Decline in ANC’). a: 4, b: 8, c: 3, d: 7, e: 3, f:4. oeMro Transplantation Marrow Bone 717 Growth factor duration A Verma et al 718 6.0 stay. The cost savings for a program performing 100 autografts a year is likely to be approximately $50 000 5.0 to $80 000 a year, depending upon growth factor dose /L) 9 Control group 4.9 and acquisition cost. Other potential benefits may be 4.0 Continued growth factor decreased intensity and frequency of bone pain that some patients experience with growth factor-induced 2.7 3.0 myeloid recovery, avoidance of growth factor-induced 2.3 fever that is seen in a small proportion of patients, 2.0 2.2 and a decrease in the incidence of the so-called engraft- Median ANC (10 Study group 1.0 ment syndrome.11 However, it must be recognized that 1.0 No growth factor our study could not address these issues. There does 0.8 0.0 not appear to be any obvious disadvantage of stopping 123growth factor early in terms of longer hospital stay or increased infection rates. Day (1: First day with ANC ≥0.5 x 109/L) The practical growth factor questions that have been Figure 1 Median absolute neutrophil counts in both groups. Note the addressed previously include whether to use the growth steeper slope of recovery in patients continuing growth factor. factor,1,2 what growth factor to use,12 when to start growth factor,3,4,7 and what dose to use.5,6 However, how long to continue growth factor has not been studied. We believe one patient did the ANC drop below 0.5 Â 109/l, to our data help provide an answer to this question. 0.3 Â 109/l on the second day, but increased to 2 Â 109/l The limitations of this study are the heterogeneity on the third day without growth factor. The magnitude of between groups in terms of the growth factor used the rise from the second to the third day suggests that the (G-CSF vs GM-CSF) and the day growth factor was low count on the second day may simply have been an started (day 0 vs day 5), because of its retrospective erroneous count. However, in the absence of any other nature. The type of growth factor used would be expected evidence, it must be assumed that the patient did become to affect the time to neutrophil recovery rather than neutropenic transiently. Almost 90% of patients in the the behavior of the counts, once growth factor admini- Study Group had an ANC on the third day, which was stration was stopped. In keeping with this sup- higher than that on the first day. position, eliminating GM-CSF-treated patients from Since the number of patients who received GM-CSF the analysis did not change the findings. Similarly, the was significantly higher in the Study Group (Table 1), the day of growth factor commencement would also data were analyzed after removing these patients. be expected to affect the time to engraftment rather than The findings were identical (Table 2). Confining the the course after stopping growth factor. Confining analysis only to patients starting growth factor on day 5 the analysis to patients who started on growth factor post transplant for the sake of uniformity yielded similar on day 5 also did not change the findings. The hetero- results (Table 2). geneous distribution of the underlying diseases between Fever requiring institution of broad-spectrum intrave- the two groups is another shortcoming; however, it nous antibiotic therapy was seen in 40 of 45 patients would not be expected to affect the end point being stopping growth factor early compared with 100 of 108 studied here. receiving additional growth factor doses (P ¼ 0.45). Bacter- As Table 2 and Figure 1 show, the rate of rise of ANC is emia was seen in 12 patients (all gram-positive) stopping less steep if growth factor is discontinued, and, in those growth factor early compared with 28 patients (1 gram- continuing the growth factor, ANC levels can reach very negative, 27gram-positive) receiving additional growth high levels. There are no data to suggest that such high factor doses (P ¼ 0.92). ANC levels have any beneficial impact on outcome. This The duration of stay in the hospital measured from the suggests that the amount of growth factors administered day of transplant (to eliminate differences emanating from routinely may be in excess of what is required. differing durations of the conditioning regimens) was We have shown previously that myeloid recovery after comparable for the two groups: 9–22 days (median 14) autotransplants is usually robust, and ANC rarely drops for those stopping growth factor early and 8–36 days below 0.5 Â 109/l, after having reached that threshold in (median 14) for those receiving additional growth factor patients continuing the growth factor.8 The results of this doses (P ¼ 0.23). study confirm this observation in patients who discontinue growth factor. We suggest that growth factors be stopped in auto- Discussion transplant recipients once the neutrophil count recovers to 0.5 Â 109/l. Our data suggest that myeloid recovery is not compromised by discontinuing growth factor as soon as ANC recovers to 0.5 Â 109/l. Thus, it is not necessary to continue Acknowledgements growth factors for 2 additional days, as is routine practice. Stopping growth factor early decreases the This work was supported in part by the Auxiliary Board of the cost of the transplant modestly without increasing hospital Northwestern Memorial Hospital.

Bone Marrow Transplantation Growth factor duration A Verma et al 719 References 6 Bolwell B, Goormastic M, Dannley R et al. G-CSF post- autologous progenitor cell transplantation: a randomized 1 Schmitz N, Dreger P, Zander AR et al. Results of a study of 5, 10, and 16 micrograms/kg/day. Bone Marrow randomised, controlled, multicentre study of recombinant Transplant 1997; 19: 215–219. human granulocyte colony-stimulating factor (filgrastim) in 7Bolwell BJ, Pohlman B, Andresen S et al. Delayed G-CSF patients with Hodgkin’s disease and non-Hodgkin’s lymphoma after autologous progenitor cell transplantation: a pros- undergoing autologous bone marrow transplantation. Bone pective randomized trial. Bone Marrow Transplant 1998; 21: Marrow Transplant 1995; 15: 261–266. 369–373. 2 Gorin NC, Coiffier B, Hayat M et al. Recombinant human 8 Ali MY, Oyama Y, Monreal J et al. Reassessing the definition granulocyte-macrophage colony-stimulating factor after high- of myeloid engraftment after autotransplantation: it is not dose chemotherapy and autologous bone marrow transplanta- necessary to see 0.5 Â 109/L neutrophils on 3 consecutive days tion with unpurged and purged marrow in non-Hodgkin’s to define myeloid recovery. Bone Marrow Transplant 2002; 30: lymphoma: a double-blind placebo-controlled trial. Blood 749–752. 1992; 80: 1149–1157. 9 Rihn C, Cilley F, Monreal J et al. Definition of myeloid 3 Torres-Gomez A, Jimenez MA, Alvarez MA et al. Optimal engraftment after allogeneic hematopoietic stem cell trans- timing of granulocyte colony-stimulating factor (G-CSF) plantation. Blood 2002; 100: 417a–418a, (abstract #1618). administration after bone marrow transplantation. A prospec- 10 Ali MY, Oyama Y, Monreal J et al. Ideal or actual body tive randomized study. Ann Hematol 1995; 71: 65–70. weight to calculate CD34+ cell doses for autologous 4 Faucher C, Le Corroller AG, Chabannon C et al. Adminis- hematopoietic stem cell transplantation? Bone Marrow Trans- tration of G-CSF can be delayed after transplantation of plant 2003; 31: 861–864. autologous G-CSF-primed blood stem cells: a randomized 11 Spitzer TR. Engraftment syndrome following hematopoietic study. Bone Marrow Transplant 1996; 17: 533–536. stem cell transplantation. Bone Marrow Transplant 2001; 27: 5 Stahel RA, Jost LM, Honegger H et al. Randomized trial 893–898. showing equivalent efficacy of filgrastim 5 micrograms/kg/d 12 San Miguel JF, Hernandez MD, Gonzalez M et al. A and 10 micrograms/kg/d following high-dose chemotherapy randomized study comparing the effect of GM-CSF and G- and autologous bone marrow transplantation in high-risk CSF on immune reconstitution after autologous bone marrow lymphomas. J Clin Oncol 1997; 15: 1730–1735. transplantation. Br J Haematol 1996; 94: 140–147.

Bone Marrow Transplantation