Increases the Numbers of Megakaryocyte Progenitor

Home , Megakaryocyte

Leukemia (1998) 12, 907–911  1998 Stockton Press All rights reserved 0887-6924/98 $12.00 http://www.stockton-press.co.uk/leu Recombinant human megakaryocyte growth and development factor (MGDF) increases the numbers of megakaryocyte progenitor cells to normal values in long- term bone marrow cultures of patients with AML in ﬁrst remission C Kasper1, A Schwarzer1, EA De Wynter1, J Chang1, TM Dexter1, D Ryder2 and NG Testa1

1Cancer Research Campaign Department of Experimental Haematology, Paterson Institute for Cancer Research, and 2Department of Medical Statistics, Christie Hospital, Manchester, UK

The megakaryopoietic potential in the bone marrow (BM) of Materials and methods patients in first remission after treatment for acute myelogenous leukaemia (AML) was investigated using long-term bone marrow cultures (LTC) stimulated with megakaryocyte growth Patients and development factor (MGDF). The baseline number of megakaryocyte colony-forming cells (Meg-CFC) was very low. BM samples were collected with informed consent from 13 However, there was a 10 to 100-fold increase of Meg-CFC in patients with AML in first complete remission. Patients’ cultures treated with 10 ng/ml MGDF with mean numbers within characteristics are shown in Table 1. the normal range for the first 4 weeks of culture with a 24-fold As induction and consolidation treatment patients received increase in their cumulative numbers. Similarly, a 12-fold two to three cycles of DAT chemotherapy, containing dauno- increase in the numbers of megakaryocytes (MKs) was found 2 by CD61 immunostaining. These effects were lost at the dose rubicin (60 mg/m i.v. days 1–3), cytosine arabinoside of 100 ng/ml. In contrast, the cumulative mean numbers of Meg- (200 mg/m2 continuous infusion, days 1–7) and thioguanin CFC in the control cultures from normal bone marrow (NBM) (200 mg/m2, days 1–7), followed by up to two cycles of inter- were not significantly different from those in cultures treated mediate dose cytosine arabinoside (1.5 g/m2 at 12 h intervals, with 10 or 100 ng/ml MGDF. These results demonstrate that days 1–3). Two patients also received high-dose busulfan MGDF stimulates megakaryocytopoiesis in patients with AML in first remission, restoring the Meg-CFC compartment to (16 mg/kg body weight) with peripheral blood progenitor cell normal values, a result with potential clinical implications for rescue prior to the collection of the bone marrow sample, and their treatment with autologous transplantation. one patient an allogenic BM transplantation. Keywords: megakaryocyte growth and development factor (MGDF); thrombopoietin; colony assay; long-term bone marrow culture (LTC); AML Long-term cultures (LTC)

BM samples were obtained after informed consent from heal- Introduction thy donors (NBM) for BM transplantation or AML patients (AMLBM). Cells were separated from the red cells by gravity Megakaryocyte growth and development factor (MGDF) or sedimentation in 0.1% methylcellulose for 45–60 min at room thrombopoietin, the ligand of the c-mpl receptor, has been temperature, and washed twice in Iscove’s modified Dulbec- purified and cloned by several groups independently.1–6 co’s medium (IMDM, Gibco BRL) plus 2% fetal calf serum MGDF has potent effects on the generation, proliferation and (FCS, Gibco BRL, Life Technologies, Paisley, UK). 1.5–2 × 106 differentiation of megakaryocyte (MK) progenitor cells in vitro buffy coat cells/ml diluted in 10 ml LTC medium that con- and in vivo.7–10 tained 80% IMDM, 10% FCS, 10% horse serum (HS, Sera Lab- − Patients treated for acute myelogenous leukemia (AML) who oratories, Crawley Down, UK), and 5 × 10 7 M hydrocortisone achieved first remission show severe and persistent haemato- were inoculated into tissue culture flasks (Falcon, Becton poietic damage manifested by substantially decreased num- Dickinson, Franklin Lakes, NJ, USA), gassed with 5% CO2 in bers of clonogenic progenitor cells and of stem cells.11 AML air and incubated at 33°C.14 Rh MGDF (a gift from Amgen patients in first or subsequent remissions may be treated with Corporation, Thousand Oaks, CA, USA) was added to the LTC autologous transplantation following high-dose chemo- at two concentrations, 10 or 100 ng/ml. The cultures were therapy.12,13 Severe and prolonged thrombocytopenia is usu- maintained by weekly feeding replacing half of the growth ally observed after this treatment. medium with fresh medium and MGDF; control cultures with- The effect of MGDF on long-term cultures (LTC) derived out MGDF were performed concurrently. The non-adherent from patients with AML in first complete remission was inves- cells removed at feeding were used for the assays described tigated in order to assess their megakaryopoietic potential and below. response to stimulation. The results indicate that MGDF has the ability to increase the numbers of MK progenitors (Meg- + CFC) and CD61 cells from a very low baseline to within Clonogenic assays normal values. Cells (105) were plated in a mixture containing 30% FCS, 1% deionized bovine serum albumin (BSA), 10% conditioned medium from the bladder carcinoma cell line 5637 (as a source of growth factors), 2 U erythropoietin (Boehringer Mannheim, West Lothian, UK) and 1.35% methylcellulose Correspondence: C Kasper, Department of Internal Medicine (Cancer Research), West German Tumor Center, University of Essen, Hufe- (Sigma, St Louis, MO, USA). Cultures were plated in triplicate landstr. 55, 45122 Essen, Germany; Fax: 0201 723 5924 and incubated at 37°C for 14 days in a fully humidified atmo- Received 3 October 1997; accepted 30 January 1998 sphere of 5% CO2,5%O2 in nitrogen. Colonies consisting of MGDF stimulates megakaryocytopoiesis in first remission AML C Kasper et al 908 Table 1 Patients’ characteristics

No. Age Sex FAB Cytogenetics No. of chemotherapy Time between CR prior to BM sample and BM sample DAT ID Ara-C (month)

1 53 M M2 t(8;21) 3 2 6 2 42 M M5b ND 2 1 2 3 54 M M1 ND 2 1 6 4 36 F M2 normal 3 2 8 5 48 M M4 del (9) 1 0 0 662MM2−7205 7 63 M M2 normal 2 0 0 8 58 F M0 normal 3 2 8 9a 26 M M2 normal 2 2 8 10 32 M M4 t(8;21), t(4;10) 3 0 3 11 20 M M2 t(3;5), del (9) 3 0 3 12b 28 M M2 ND 2 0 106 13a 23 M M2 ND 2 1 56

aPatients studied two (No. 9) and 50 months (No. 13) after autologous transplantation with mobilized peripheral blood progenitor cells. bThis patient was studied 101 month after allogeneic bone marrow transplantation. FAB, French–American–British classiﬁcation; ID Ara-C, intermediate-dose cytosine arabinoside; CR, complete remission; BM, bone marrow; ND, not documented.

50 translucent cells or more were counted as GM-CFC- Results derived colonies using an inverted microscope. To assay Meg-CFC 1–2 × 105 cells were plated in a mixture In NBM cultures, Meg-CFC were detected for 7 weeks. How- − of 30% pretested human aplastic anaemia serum, 5 × 10 5 M ever, they were detected only until week 5 in the cultures 2-mercaptoethanol, 5% conditioned medium from phyto- from patients with AML (Figure 1). In those cultures only very haemagglutinin-stimulated leukocytes (PHA-LCM) and 1.35% methylcellulose. Cultures were plated and incubated as described above. We used the established criteria for defining MK colonies: colonies of at least five conspicuously large cells with translucent ‘crystal’ clear cytoplasm and a well-delin- eated cell border of high refractility were scored as Meg- CFC.14 Their megakaryocytic nature was confirmed by picking some colonies for CD61 immunostaining. All results were cal- culated as total colony-forming cells in the supernatant (SN) per culture; the data in Figures 1–3 are expressed as mean ± s.e.m.

Cell morphology

Cytospins were prepared with 5 × 104 cells for Pappenheim stain and CD61 immunostaining (DAKO, Glostrup, Denmark) as previously described,15 to determine the number of megakaryocytic cells.

Statistics

Formal statistical analysis was confined to the MK profiles. It was felt that the cumulative output in weeks 1–4 captured the aspect of primary interest as it encloses the time of acute regeneration after chemotherapy and also covers the time previously used for autologous transplants with cultured AMLBM cells.16 On occasion, however, when data were missing for some of the weeks, the observed cumulative output continued to be right censored for such individuals (ie it is assumed that the actual output is known only to be larger than that Figure 1 The upper panel shows the numbers of Meg-CFCs per observed). Right censored data may be analysed with ‘sur- í vival’ methods. The statistical analysis was done with a Cox culture in the supernatant of AML for untreated control cultures ( , n = 13), or treated with 10 ng/ml (b, n = 8) or 100 ng/ml MGDF (̆, regression model with appropriate adjustment for the corre- n = 7). The lower panel shows data from NBM cultures for untreated 17 lation of repeated observations for the same individual. All (n = 13), or treated with 10 ng/ml (n = 6) or 100 ng/ml MGDF (n = tests performed in this framework were Wald type tests. 9). Data are expressed as mean numbers of Meg-CFC ± s.e.m. MGDF stimulates megakaryocytopoiesis in first remission AML C Kasper et al 909 low levels of Meg-CFC were seen during the first 4 weeks of nificance was lost in the cultures treated with 100 ng/ml culture. The output of Meg-CFC in untreated AML cultures MGDF (P = 0.19). Treatment of NBM with MGDF had no sti- was at least one log lower than in NBM cultures (P = 0.001). mulating effect on the numbers of Meg-CFC (P = 0.21), CD61 immunostaining indicated the presence of MK until although there appeared to be a trend towards an increased week 3 for both NBM and AMLBM. However, treatment of yield (Figure 1). AML cultures with MGDF resulted in a two log increment in Although the number of experiments using immunostaining week 1 and a one log increase in week 2, reaching similar with CD61 was smaller similar results were found. In the NBM Meg-CFC numbers to those observed in NBM cultures. In (n = 8) there was a cumulative mean number of 353 positively these and following experiments, patient 12, who received an stained cells by week 4; this was not different in both treat- allogenic bone marrow transplantation, showed similar results ment groups (290 and 370 for 10 and 100 ng/ml MGDF, as the rest of the patients, so the results were analysed respectively). However, in AML cultures (n = 3) the corre- together. There were no MK colonies detected in clonogenic sponding number was 109 for the control with an increase assays performed from the adherent cells from NBM (n = 4) to 1252 in two experiments treated with 10 ng/ml MGDF. In and AMLBM treated with 10 ng/ml TPO (n = 4) after week 5 experiments with NBM (Figure 3a), or with AMLBM (Figure of culture. 3b), MGDF had no effect on the number of non-adherent Analysis of these results showed that the cumulative output cells, nor on the numbers of GM-CFC up to week 4 of culture. of Meg-CFC in weeks 1–4 (Figure 2) was significantly lower However, in agreement with our previously published data11 in cultures derived from AMLBM than from NBM (P = 0.01). from week 5 onwards the numbers of GM-CFC declined in However, treatment of AML cultures with 10 ng/ml MGDF the AML cultures compared to those derived from NBM. This resulted in a 24.3-fold increase in the cumulative mean num- was not changed by administration of MGDF. Treatment with ber of Meg-CFC at week 4 reaching similar numbers of Meg- 10 or 100 ng/ml MGDF had no effect on the erythroid colony CFC as those seen in the NBM cultures (P = 0.21). This formation (data not shown). enhancement was statistically significant (P Ͻ 0.001). The sig-

Discussion

Patients treated for AML develop severe neutropenia and thrombocytopenia which might cause treatment delay and could lead to fatal infections or bleeding. There is hope that MGDF may act in thrombocytopenia similarly to G-CSF in neutropenia, which decreases the median duration of neutropenia in elderly patients after treatment for AML.18 However, it is known that the haematopoietic status of AML patients in remission shows a marked and persistent deficiency in primitive haematopoietic cells, both at the level of committed progenitors, and of stem cells.11 It is therefore important to establish that there are enough residual primitive cells able to respond to megakaryopoietic stimulation in response to the appropriate cytokine. This is also relevant for patients who receive an allogenic bone marrow transplantation (like one of the patients in this series), since it is known that these patients have a long-term deficiency in the numbers of CFC in their BM.19 In these experiments LTC were used to investigate the effect of MGDF on BM from patients in first complete remission. The results demonstrate a several-fold enhancement of Meg- CFC to numbers within the normal range, as well as a similar increase in CD61-positive cells in LTC treated with MGDF at a dose of 10 ng/ml. However, there was no stimulation of Meg-CFC in NBM above normal ranges. Thrombopoietin is synthesized by BM stromal cells.20 Although the MGDF levels were not measured in these experiments it is known that the concentrations of MGDF used here were well above the levels measured in LTC20 and human serum.21 It appears to be likely that there is a deficiency in the microenvironment in AMLBM while the megakaryopoiesis in cultures derived from NBM may be already maximally stimulated by stromal cells. Figure 2 Percentage of Meg-CFC above an absolute number of MGDF effects appear not to be restricted to megakaryocyto- Meg-CFC (x). Data are expressed as cumulative numbers of Meg-CFCs poiesis. Recently, it has been shown that MGDF possesses the from weeks 1 to 4 in AML cultures (upper panel) and NBM (lower ability to enhance erythroid proliferation,22 and in combi- panel) for all experiments, discriminated between control (—), nation with other cytokines, stimulates the proliferation of 10 ng/ml MGDF (---), 100 ng/ml MGDF (– – –). Censored values are granulocytes, macrophages23 and primitive haematopoietic marked with ticks. The percentage of Meg-CFC in AMLBM using 24–26 10 ng/ml MGDF were significantly different from the values of control cells; these effects were not observed using MNC from cultures and cultures using 100 ng/ml MGDF but not significantly dif- umbilical cord blood (A Schwarzer, personal ferent from cultures derived from NBM. communication). Furthermore, in those experiments MGDF MGDF stimulates megakaryocytopoiesis in first remission AML C Kasper et al 910

Figure 3 The left panel shows the mean numbers (± s.e.m.) of nucleated cells (top) and of GM-CFCs (bottom) per culture in the supernatant of untreated control cultures of AML patients (í, n = 13), or treated with 10 ng/ml (b, n = 8) or 100 ng/ml MGDF (̆, n = 7). The right panel shows the corresponding data for cultures of NBM for untreated control (n = 13), or treated with 10 ng/m (n = 6) or 100 ng/ml MGDF (n = 9).

had no influence on the number of CFC or of granulopoietic only by the expected stimulation of platelet production, but or monocytic cells in the cultures. A similar effect has been also by correcting the serious defect in the numbers of MK noted previously using G-CSF which has only a small and progenitors in BM. transient stimulatory effect on granulopoiesis in LTC of human BM,27 perhaps because granulopoietic cell production is already adequately stimulated in this system. No attempts Acknowledgements were made in these experiments to optimize the system for erythroid growth. These results are in contrast with the CK is supported by the ‘Aktion Kampf dem Krebs’, Germany. 23 increase in granulopoiesis in LTC of murine BM seen This work was supported by CRC, UK. together with stimulation of megakaryopoiesis after treatment with thrombopoietin. The lack of influence of MGDF after 5 weeks of culture to generate Meg-CFC in LTC is likely to be References due to the poor status and short life of BM cultures of AML 11 in remission, reported previously and confirmed here. 1 Bartley TD, Bogenberge J, Hunt P, Li Y-S, Lu HS, Martin F, Chang It is known that AML cells may express the c-mpl proto- M-S, Samal B, Nichol JL, Swift S, Johnson MJ, Hsu R-Y, Parker VP, oncogene28 and that recombinant human thrombopoietin may Suggs S, Skine JD, Merewther LA, Clogston C, Hsu E, Hokom MM, induce proliferation of AML cells in vitro.29,30 A similar in vitro Hornkohl A, Choi E, Pangelinan M, Sun Y, Mar V, Mcnich J, stimulating effect of G-CSF (and also of other cytokines) on in Simonet L, Jacobsen F, Xie C, Shutter J, Chute H, Basu R, Selander vitro proliferation of AML cells has been reported.31 In spite L, Trollinger D, Sieu L, Padilla D, Trail G, Elliott G, Izumi R, Covey TT, Crouse J, Carcia A, Xu W, Del Castillo J, Biron J, Cole S, Hu of this, randomised clinical trials in patients with AML have Mc-T, Pacifici R, Ponting L, Saris C, Wen D, Yung YP, Lin H, not detected an effect of G-CSF administration on leukaemic Bosselman RA. Identification and cloning of a megakaryocyte cell growth.18 Thus, it remains to be determined whether the growth and development factor that is a ligand for the cytokine presence of the c-mpl receptor in AML cells will have any receptor MPL. Cell 1994; 77: 1117–1124. clinical relevance for the administration of MGDF to AML 2 Kaushansky K, Lok S, Holly RD, Broudy VC, Lin N, Bailey MC, patients. Forstrom JW, Buddle MM, Oort PJ, Hagen FS, Roth GJ, Papayan- nopaulou T, Foster DC. Promotion of megakaryocyte progenitor In summary, the present results indicate that the BM of AML expansion and differentiation by the c-Mpl ligand thrombopoietin. patients have enough potential reserve capacity to be able to Nature 1994; 369: 568–571. respond to stimulation of megakaryopoiesis by MGDF, not 3 Lok S, Kaushansky K, Holly RD, Kuijper JL, Lofton-Day CE, Oort MGDF stimulates megakaryocytopoiesis in first remission AML C Kasper et al 911 PJ, Grant FJ, Heipel MD, Burkhead SK, Kramer JM, Bell LA, 18 Dombret H, Chastang C, Fenaux P, Reiffers J, Bordessoule D, Bou- Sprecher CA, Blumberk H, Johnson R, Prunkard D, Ching AFT, abdallah R, Mandelli F, Frrant A, Auzanneau G, Tilly H, Yver A, Mathewes SL, Bailey MC, Forstrom JW, Buddle MM, Osborn SG, Degos L. A controlled study of recombinant human granulocyte Evans SJ, Sheppard PO, Presnell SR, O’Hara PJ, Hagen FS, Roth colony-stimulating factor in elderly patients after treatment for GJ, Foster DC. Cloning and expression of murine thrombopoietin acute myelogenous leukemia. AML cooperative study group. New and stimulation of platelet production in vivo. Nature 1994; 369: Engl J Med 1995; 332: 1678–1683. 565–568. 19 Messner HA, Curtis JE, Minden MD, Tritchler D, Lockwood G, 4 Miyazaki H, Kato T, Ogami K, Iwamatsu A, Shimada Y, Souma Y, Takahashi T, Lepine J, Jamal N, Tweeddale M, Wandl U. Clono- Akahori H, Horie K, Kokubo A, Kudo Y, Maeda E, Kawamura K, genic haematopoietic precursors in bone marrow transplantation. Sudo T. Isolation and cloning of a novel human thrombopoietic Blood 1987; 70: 1423–1432. factor. Exp Hematol 1994; 22: 838 (Abstr.). 20 Guerriero A, Worford l, Holland K, Guo G-R, Sheehan K, Waller 5 De Sauvage FJ, Haas PE, Spencer SD, Malloy BE, Gurney AL, K. Thrombopoietin is synthesized by bone marrow stromal cells. Spencer SA, Darbonne WC, Henzel WJ, Wong SC, Kuang W-J, Blood 1997; 90: 3444–3455. Oles KJ, Hultgren B, Solberg LA, Goeddel DV, Eaton DL. Stimu- 21 Cerutti A, Custodi P, Duranti M, Noris P, Balduini CL. lation of megakaryopoiesis and thrombopoiesis by the c-MPL Thrombopoietin levels in patients with primary and reactive ligand. Nature 1994; 369: 533–538. thrombocytosis. Br J Haematol 1997; 99: 281–284. 6 Wendling F, Maraskovsky E, Debili N, Florindo C, Teepe M, Titeux 22 Kobayashi M, Laver JH, Kato T, Miyazaki H, Ogawa M. Recombi- M, Methia N, Breton-Gorius J, Cosman D, Vainchenker W. c-MPL ligand is a humoral regulator of megakaryopoiesis. Nature 1994; nant human thrombopoietin (mpl ligand) enhances proliferation 369: 571–574. of erythroid progenitors. Blood 1995; 86: 2494–2499. 7 Debili N, Wendling F, Cosman D, Titeux M, Florindo C, Dusanter- 23 Ponting ILO, Hokorn M, Dwyer E, Covey T, Izumi R. Stimulatory Fourt L, Schooley K, Methia N, Charon M, Nador R, Bettaieb A, effects of recombinant human megakaryocyte growth and devel- Vainchenker W. The MPL-receptor is expressed in megakaryocytic opment factor (rHuMGDF/thrombopoietin) in long-term bone lineage from late progenitors to platelets. Blood 1995; 85: 391– marrow cultures. Exp Hematol 1995; 23: 829 (Abstr.). 401. 24 Kobayashi M, Laver JH, Kato T, Miyazaki H, Ogawa M. Throm- 8 Debili N, Wendling F, Katz A, Guichard J, Breton-Gorius J, Hunt bopoietin supports proliferation of human primitive hematopoietic P, Vainchenker W. The MPL-ligand or thrombopoietin or mega- cells in synergy with steel factor and/or interleukin-3. Blood 1996; karyocyte growth and development factor has both direct pro- 88: 429–436. liferative and differentiative activities on human megakaryocyte 25 Young JC, Bruno E, Luens KM, Wu S, Backer M, Murray LJ. Throm- progenitors. Blood 1995; 86: 2516–2525. bopoietin stimulates megakaryocytopoiesis, myelopoiesis, and 9 Broudy VC, Lin NL, Kaushansky K. Thrombopoietin (c-mpl ligand) expansion of CD34+ progenitor cells from single CD34+ Thy-1+ acts synergistically with erythropoietin, stem cell factor, and Lin− primitive progenitor cells. Blood 1996; 88: 1619–1631. interleukin-11 to enhance murine megakaryocyte ploidy in vitro. 26 Schwarzer A, De Wynter E, Kasper C, Chang J, Helbig W, Testa Blood 1995; 85: 1719–1726. NG. The generation of progenitor cells is increased by megakary- 10 Kaushansky K. Thrombopoietin: the primary regulator of platelet ocyte growth and development factor in stroma-supported cultures production. Blood 1995; 86: 419–431. of CD34+ cells from cord blood. Exp Hematol 1996; 24: (Abstr. 11 Chang J, Geary CG, Testa NG. Long-term bone marrow damage 406). ´ after chemotherapy for acute myeloid leukaemia does not improve 27 Coutinho LH, Will A, Radford J, Schiro R, Testa NG, Dexter TM. with time. Br J Haematol 1990; 75: 68–72. Effects of recombinant human granulocyte colony-stimulating fac- 12 Gale RP, Horowitz MM, Butturini A. Autotransplants in acute leu- tor (CSF), human granulocyte macrophage-CSF, and gibbon kaemia. Br J Haematol 1991; 78: 135–137. interleukin-3 on hematopoiesis in human long-term bone marrow 13 Imrie K, Dicke KA, Keating A. Autologous bone marrow trans- culture. Blood 1990; 75: 2118–2129. plants for acute myeloid leukaemia. Stem Cells 1996; 14: 69–78. 28 Vigon L, Dreyfus F, Melle J, Viguie F, Ribrag V, Cocault L, Souyri 14 Coutinho LH, Gilleece MH, De Wynter EA, Will A, Testa NG. M, Gisselbrecht S. Expression of the c-mpl proto oncogene in Clonal and long-term cultures using human bone marrow. In: human hematologic malignancies. Blood 1993; 82: 877–883. Testa NG, Molineux G (eds). Haemopoiesis: a Practical Approach. 29 Matsumara L, Kanakura Y, Kato T, Ikeda H, Ishikawa J, Horikawa Oxford University Press: Oxford 1993; pp 75–106. Y, Hashimoto K, Koriyama Y, Tsujimura T, Nishiura T, Miyazaki 15 Briddel RA, Brandt JE, Straneva JE, Srour EF, Hoffmann R. Charac- H, Matsuzawa Y. Growth of response of acute myeloblastic leuke- terization of the human burst-forming unit-megakaryocyte. Blood 1989; 74: 145–151. mia cells to recombinant human thrombopoietin. Blood 1995; 86: 16 Chang J, Morgenstern GR, Coutinho LH, Scarffe JH, Carr T, Deakin 703–709. DP, Testa NG, Dexter TM. The use of bone marrow cells grown 30 Motoji T, Takanashi M, Motomura S, Yan-Hua W, Shiozaki H, in long-term culture for autologous bone marrow transplantation Aoyama M, Mizoguchi H. Growth stimulatory effect of throm- in acute myeloid leukaemia: an update. Bone Marrow Transplant bopoietin on blast cells of acute myelogenous leukaemia. Br J 1989; 4: 5–9. Haematol 1996; 94: 513–516. 17 Lee EW, Wei LJ, Amato D. Cox-type regression analysis for large 31 Motoji T, Watanabe M, Uzumaki H, Kusaka M, Fukamachi H, numbers of small groups of correlated failure time observations. Shimosaka A, Oshimi K, Mizoguchi H. Granulocyte colony-stimu- In: Klein JP, Goe PK (eds). Survival Analysis, State of the Art. lating factor (G-CSF) receptors on acute myeloblastic leukaemia Kluwer Academic Publishers: Dordrecht, Netherlands 1992; cells and their relationship with the proliferative response to G- pp 237–247. CSF in colonogenic assay. Br J Haematol 1991; 77: 54–59.