Concise Review International Journal of Cell Cloning 10:262-268 (1992)

In Vitro Revelations of Aplastic Anemia Frances M. Gibson, Judith C. W. Marsh, Edward C. Gordon-Smith Division of Haematology, Department of Cellular and Molecular Sciences, St. George’s Hospital Medical School, London SW17 ORE, United Kingdom Key Words. Aplastic anemia Hemopoietic growth factors Antilymphocyte globulin GM-CSF G-CSF LTBMC Stem cell factor Abstract. Aplastic anemia (AA) is a most dimcult disease to study in vitro. By the time the disease presents, the marrow is already hypocellular and the peripheral blood shows pancytopenia, leaving little material remaining for study. However, an understanding of its pathogenesis could provide insight into the control of normal hemopoiesis since AA is an in vivo manifestation of failure of normal hemopoiesis and may provide a way of examining stromal cell-stem cell relationships. Recent interest in the pathogenesis of AA has resulted from a) new laboratory techniques, such as stem cell purification used with modificationsof the long-term bone marrow culture system and analysis of stem cells at the molecular level with X-linked DNA probes, and b) the availability of recombinant human hemopoietic growth factors (HGF) in large quantities. Consequently, analyses of the function of some of the individual components of stromal cell mediated hemopoiesis in AA patients have been performed. This has been paralleled, and in some instances preceded, by clinical trials of HGF in patients with AA.

Early In Vitro Studies in AA For many years the pathogenesis of AA has remained a mystery. Proposed mechanisms included an intrinsic deficiency or defective function of hemopoietic stem cells, a secondary stem cell defect consequent to abnormal humoral or cellular regulation, or a defective stromal cell microenvironment. Largely based on clinical observations, it had been assumed that an abnormality in the stem cell compartment rep resented an important pathogenetic mechanism in AA. Studies using clonogenic cultures from the mid1970s to early 1980s uniformly reported absent or Correspondence:Prof. Edward C. Gordon-Smith, Division of Hematology, Department of Cellular and Molecular Sciences, St. George’s Hospital Medical School, London SW17 ORE, United Kingdom. Received May 12, 1992; accepted for publication May 12, 1992. 0737- 1454/92/$2.0010 OAlphaMed Press

reduced numbers of marrow colonies in AA [l-31, but these results could not differentiate between stem cell or stromal cell abnormalities. Furthermore, only the concentration and not the absolute number of progenitors could be assessed, and no account was taken of marrow cellularity. Finally, interpretation is compounded by the very wide range of incidence of marrow progenitors seen in normal individuals. During the next five to ten years much interest centered around the idea of immune mediated suppression of hemopoiesis in AA [4, 51. Conflicting reports of the presence of an abnormal population of circulating activated T suppressor lymphocytes and elevated levels of gamma-interferon and/or interleukin-2 (IL-2) were reported [6, 71. These findings were used to explain the hematological response that occurred in a proportion of patients following treatment with antilymphocyte globulin (ALG). However, clinical response to ALG correlates poorly with these in vitro results, and response after ALG is rarely complete hematologically [8]. It is probable that ALG has other important effects in vivo, such as immunostimulation resulting in the release of HGF [9], as well as a lymphocytotoxic effect [ 101. Finally, many of these experiments were not controlled for transfusion-induced T and B cell sensitization, resulting in inhibition of marrow progenitors [ll]. Kurninsb and colleagues have recently shown that alloreactive CD8’ T lymphocytes are generated by blood transfusion in contrast to low numbers found in untransfused patients [12]. It is therefore probable that the above in vitro findings represent epiphenomena occurring secondary to an underlying stem cell defect in many patients with AA. From the 1970s onwards, allogeneic bone marrow transplantation (BMT) for severe AA was becoming a successful form of treatment with apparent cure of the aplasia [13, 141. Long-term trilineage engraftment in most cases implies that function of the recipient’s microenvironment is not damaged, and, since in vitro studies [15], including recent

263

long-term bone marrow culture (LTBMC) analysis [16], have confirmed that the stroma is host derived post-BMT, a stromal defect in these AA patients could not have been the cause of the original marrow failure.

Studies of LTBMC in Aplastic Anemia Despite the lack of an assay for the human pluripotent stem cell, the Dexter LTBMC system provides a physiological assessment of the primitive hemopoietic progenitor cells responsible for establishment and maintenance of hemopoiesis for up to three months. Generation of hemopoietic progenitors depends on an intact stromal cell layer with intimate contact between stromal and hemopoietic cells [17-191. Three studies in AA have consistently reported the absence of generation of hemopoietic progenitor cells in the presence of normal stroma formation within this system [20-221, even in patients who have shown hematological “response” after ALG therapy and have normal numbers of colonies grown in clonogenic culture [20]. Modification of the LTBMC system has permitted separate analysis of stromal and hemopoietic stem cell function. Once grown to confluence in LTBMC, the stroma can be irradiated to eliminate endogenous hemopoiesis without compromisingstromalfunction. Marrow cells of interest whether from autologous, allogeneic or HLA mismatched sources can then be seeded and grown on these stromal cell layers [23, 241. Using adherent cell depleted (ACD) cells or a purified population of progenitor cells expressing the CD34 antigen, two studies in AA [25, 261 have shown that 414 and 13/14 patients, respectively, with untreated or longstanding AA following ALG therapy had normal functioning stroma as assessed by the ability of AA stroma to support the growth of normal ACD or CD34’ marrow cells. Conversely, in all cases a severe defect of stem cell function was demonstrated when AA marrow cells (ACD or CD34’) were grown on normal stromas. An earlier study by Hortu and colleagues reported abnormal stromal function in 319 patients with acute severe AA. Stem cell function was not assessed in these patients [27]. It is important to note that these studies assess only the functional capacity of the stroma to support the generation of hemopoietic progenitor cells. The stroma is a functional, multicomponent unit comprised of fibroblasts, macrophages, fat cells and endothelial cells [ 181, which, along with the extracellular matrix, provide the microenvironment necessary for interactions of the stem cell with the

Aplastic Anemia In Vitro positive HGF and negative regulatory factors. It is theoretically possible for a defect in one component to be compensated by another component, or for a subtle abnormality not to compromise the functional property of support for hemopoiesis. For example, Gibson and coworkers [28], using stromal cells to stimulate colony formation of normal bone marrow target cells in semisolid culture in the absence of exogenous HGF, have demonstrated that stromal cells from some patients with AA produce greater than normal colony stimulating activity (CSA). This does not appear to be true for burst promoting activity (BPA).

Effect of Recombinant Human Hemopoietic Growth Factors on Hemopoiesis in Aplastic Anemia Such was the initial excitement in the efficacy of HGF such as granulocyte-macrophage colony stimulating factor (GM-CSF) and granulocyte CSF (G-CSF) to stimulate granulopoiesis in patients with acquired immunodeficiency syndrome 1291, with myelodysplastic syndrome (MDS) [30], or following chemotherapy [31], that the clinical use of HGF in AA preceded appropriate in vitro investigations. Nevertheless, recent studies of HGF in the laboratory have shown good correlation with results obtained from clinical trials. GM-CSF, IL-3 or a combination of both have been added to normal LTBMC and resulted in similar increases in total nonadherent cell layer counts between weeks 3 and 6 of culture compared with controls. However, IL-3 addition resulted in more than double the increase in GM colony forming unit (CFU-GM) numbers compared with GM-CSF, reaching a maximum at week 4. The combination of IL-3 and GM-CSF resulted in an earlier increase in CFU-GM in the cultures when it was additive for the effects of the growth factors alone. The longevity of the growth factor-treated cultures was not reduced (or prolonged), indicating that premature exhaustion of the hemopoietic system due to excess recruitment of stem cells did not occur. Conversely, HGF did not lead to increased stem cell renewal [32]. These results indicate that IL-3 and GM-CSF increase myeloid cell production by affecting different but overlapping populations of progenitor cells. It is not known, however, whether this is due to direct effects of the growth factors on target cells or indirect effects through stimulation of stromal cells to produce endogenous factors. Coutinho et al. [33] have also investigated the effect of exogenous GM-CSF and IL-3 in normal

Gibson/Marsh/Gordon-Srnith.

LTBMC. They reported comparable increases in total cell number and CFU-GM number, and that GM-CSF and IL-3 individually increased CFU-GM numbers in the adherent layer and markedly increased their cycling rates. The effect of a combination of GM-CSF and IL-3 was, however, not reported. In six patients with AA, the addition of GMCSF, IL-3 or a combination of both factors to LTBMC did not increase numbers of nonadherent cells or CFU-GM, except in one case studied during response to cyclosporin [34]. This would be compatible with a quantitative or qualitative progenitor cell deficiency and as a consequence raises the possibility that AA stroma endogenously generates a maximal proliferative signal. Several clinical trials have been conducted in AA patients using GM-CSF and show that, in cases of moderately severe disease, GM-CSF therapy may produce transient rises in peripheral neutrophil counts [35-371. However, in very severe cases the response is poor [38]. The use of G-CSF in general appears to be associated with fewer side effects and to produce a more pronounced increase in the neutrophil count compared with GM-CSF. Sonodu and colleagues have recently reported the long-term (2 to 11 months) administration of G-CSF to five patients with refractory AA, all demonstrating increases in neutrophil counts and no evidence of depletion of progenitor cells after one year of treatment [39]. IL-3 has also been used in the treatment of AA, but only a transient increase in neutrophils, eosinophils, monocytes and lymphocytes was seen, and the degree of increase in cell count was less than that seen with GM-CSF. There was no significant stimulation of erythropoiesis or megakaryocyte production [40]. IL-1 alone has no colony stimulating activity, but in combination with other HGFs such as G-CSF it acts synergistically to stimulate growth of multipotent progenitors [41]. Unlike all of the other HGFs in AA where serum levels are either increased or normal, IL-1 production by monocytes is reduced in patients with severe AA [42]. An increase in IL-1 levels following ALG therapy has been shown to be predictive of response to ALG [43], and this was the rationale for a recent trial of recombinant human IL-1 in four patients with refractory severe AA [44]. However, treatment with IL-1 resulted in no hematological response in any patient. On the contrary, there was a gradual fall in hemoglobin levels and platelet counts with increased transfusional requirements, in addition to considerable toxicity (notably hypotension, fever, rigors, headache, nausea). Although these patients also demonstrated increased numbers of circulating activated CD8' T lymphocytes before IL- 1 treatment

264

with reduction in numbers after treatment, the absence of hematological response suggests that the immunological abnormalities seen were not the primary cause of the marrow failure in these patients. The isolation and purification of the multipotent HGF, stem cell factor (SCF) [45], has provided an explanation for the hemopoietic defects seen in the genetically anemic WIW' and SllSld strains of mice, which may be applicable to some of the congenital bone marrow failure disorders in man. These mice have a macrocytic anemia, mast cell deficiency, abnormal megakaryocyte production, reduced granulopoiesis, coat color abnormalities and sterility [46]. The stem cell defect in the WIW' mouse is due to mutations at the W locus producing a defective receptor (c-kit) for which SCF is a ligand. The hemopoietic defect in the SWSP mouse is in the stroma, due to reduced production of SCF. Since SCF and the c-kit receptor in the mouse are important in the early regulation of hemopoiesis, it is perhaps relevant to examine the congenital bone marrow failure syndromes in man for analogies with the SlISP and WIW' mice [47]. In vitro culture of bone marrow from patients with Diamond Blackfan anemia (DBA, congenital red cell aplasia) in the presence of SCF resulted in marked increases in the numbers and sizes of erythroid bursts [48], implying that DBA is not analogous to the WIW' mouse. Since allogeneic BMT can correct the hematologic defect in DBA, a microenvironmental defect seen in the SUSP mouse is also excluded in the pathogenesis of this disorder [49]. Dyskeratosis congenita (DC), an inherited disorder characterized by leukoplakia, dystrophic nails and abnormal skin pigmentation results in AA in 50% of cases [50]. SCF has not been used in clinical studies, but recent LTBMC data have shown normal functioning stroma with severe stem ceI1 defects in all of three patients with DC, two of whom had AA, and one with a normal blood count (apart from mild macrocytosis) and normal bone marrow vephine biopsy [51]. These results again predict that the hemopoietic defect in DC would not be corrected by administration of SCF, perhaps analogous to the situation in the WIW' mouse, in that an intrinsic stem cell defect is present. There are no published reports on the effect of SCF in LTBMC in acquired AA. Indeed, one might predict that SCF would have no effect on hemopoiesis since the stroma in AA is functionally normal, and like all the other HGFs in AA, except IL-1, SCF levels may be elevated anyway. In this context, it is not uncommon to see increased mast cells on morphological examination of bone marrows of patients with AA, which may reflect increased SCF levels.

265

Analysis of Purified Progenitor Cells in Aplastic Anemia The crossover LTBMC experiments described earlier do not differentiate between an intrinsic stem cell defect or a defect secondary to abnormal regulation. Therefore, attempts have been made to isolate and characterize a more primitive population of hemopoietic progenitors in AA. CD34+marrow cells from AA patients have been isolated using immunomagnetic beads or fluorescent activated cell sorting (FACS) [26,52,53]. The population of CD34+cells in six patients with nonsevere AA who had responded to ALG was reduced compared with normal controls and generated fewer multipotent colonies (CFU-Mix), early erythroid cells (BFU-E) and granulocyte-macrophage (CFU-GM)colonies in clonogenic culture [26]. Novifsky and Jacobs also demonstrated reduced blast colony formation from AA hemopoietic cells grown over preformed normal stromal layers [52]. The absence of generation of hemopoietic progenitors from AA CD34+cells grown on normal stromas in LTBMC implies a deficiency in the early hemopoietic cells with marrow repopulating or long-term culture initiating (LTCI) ability [26]. These cells have been shown in normal bone marrow to express the phenotype CD34+CD33- [54]. Gibson et al. [53] have recently isolated by FACS the following subsets of CD34+cells from 14 AA patients with more severe disease: CD33CD34+,CD33+CD34+, and CD33+CD34-.These have been compared to similar cell populations from 11 normal controls. All three subsets were reduced by 50-70% in AA. lending furlher support for an intrinsic stem cell defect in this disorder, although the clonogenic potential of the AA CD34+CD33-cells has not yet been determined.

Clonal Analysis of Hemopoiesis in Aplastic Anemia It has long been recognized that AA may arise from or give rise to paroxysmal nocturnal hemoglobinuria (PNH) later in the course of the disease. PNH may rarely progress to acute myeloid leukemia (AML) [55]. Recent studies of long-term survivors of AA treated.with ALG [56] or androgens [57] have shown a high incidence of clonal disorders occurring as a late event. The risk of clonal evolution is around 50% at eight years following ALG therapy. Karyotype analysis had shown that clonal cytogenetic abnormalities in AA occur in only about 5% of cases [58]. A variety of X-linked DNA probes have recently become available and have been used to investigate whether hemopoiesis is monoclonal or

Aplastic Anemia In Vitro polyclonal in AA. Using the hypoxanthine phosphoribosyl transferase (HPRT) 800, phosphoglycerate kinase (PGK)and M27P probes, van Kamp [59] has reported that 10/16 AA patients have apparent monoclonal patterns of hemopoiesis on analysis of peripheral blood mononuclear cells (MNC), and myeloid and lymphoid cells in four of these ten patients. However, extreme lyonization (which can mimic a monoclonal pattern) was not excluded in these patients. Since it has recently been shown in two separate studies that 23% [60] and 16% (Janssen, unpublished observations), respectively, of normal females exhibit extreme lyonization using the PGK and/or HPRT probes, the lack of the appropriate controls in van Kamp's study renders an assessment of the real incidence of monoclonal hemopoiesis in their patients impossible. In contrast, Josten et al. reported that five out of six AA patients had polyclonal hemopoiesis (extreme lyonization was not excluded in the one patient with an apparent monoclonal pattern), and all of five patients with PNH (a preleukemic clonal disorder) exhibited monoclonal hemopoiesis, using the same DNA probes and flow cytometry analysis of decay accelerating factor (DAF) receptor expression [all. Results similar to those of Josten and colleagues have been obtained by Janssen (unpublished observations) on analysis of MNC from blood and marrow, showing a polyclonal pattern in 19/21 AA patients who were either untreated or treated with ALG, or following complete autologous reconstitution after allogeneic BMT. A true monoclonal pattern was only seen in two patients. However, subfractions of peripheral blood cells were not analyzed separately in this study, so a small monoclonal population could have been masked by a larger polyclonal population. Five patients in this study with polyclonal hemopoiesis were studied in LTBMC and showed severe stem cell defects, suggesting that recovery of hemopoiesis occurred with a population of functionally defective stem cells. The interpretation of analysis of clonality using X-linked DNA probes in AA is therefore not straightforward [62]. It is evident that further study of Xlinked DNA polymorphism with proper controls must be interpreted in conjunction with traditional marrow morphology, chromosome analysis and Hams test to provide further insight into the stem cell defect that is probably the major pathogenetic mechanism in AA.

Future Perspectives for In Vitro Study of Hemopoiesis in Aplastic Anemia It is now evident that the high incidence of clonal disease seen with long-term follow-up of AA

Gibson/Marsh/Gordon-Smith. patients treated with ALG or androgens is not seen following allogeneic BMT for AA [63,64], indicating that transplanted patients are truly cured of their original disease. ALG treated patients also have a high incidence of relapse [65] and rarely achieve complete normalization of peripheral blood counts [8],and results of LTBMC studies described earlier indicate a persistent stem cell defect despite hematological “recovery”. This defect appears to be intrinsic in origin and may predispose to late clonal disease. Studies to assess whether hemopoiesisin LTl3MC following BMT returns to complete normality have yet to be reported. There are two new techniques that have recently become established and which may circumvent some of the difficulty in obtaining sufficient cells for analyses from patients with AA. The polymerase chain reaction (PCR) has been applied to clonality analysis using X-linked methylation patterns with the PGK probe [66]. This allows analysis of a small number of cells requiring only 0.1 pg of DNA per sample (lo’ cells) instead of 50-100 pg of DNA per sample (lo6 cells) for Southern hybridization, and may allow clonality analysis on single colonies. Quantitation of the very primitive hemopoietic progenitors in normal bone marrow with long term culture initiating (LTCI) ability has been reported [67]. CD34’ cells are seeded at limiting dilution onto irradiated stromal layers in microwell plates. The frequency of LTCI cells in normal bone marrow is 1 per 50-100 CD34’ cells or 1 per 2 x 104cells in a buffy coat fraction of normal marrow. The application of this technique to the study of patients with AA would provide further quantification of stem cell reserve and could provide the most accurate assessment of disease severity and response to treatment. Newly cloned HGFs including IL-6 [68] and IL-11 [69], which stimulate megakaryopoiesis in vitro, are awaiting clinical trials in AA and the bone marrow failure syndromes characterized by thrombocytopenia. Although one might predict that SCF would not be expected to reverse the defect of hemopoiesis in acquired AA, it is very likely that clinical trials will be performed (as has been the case for G-CSF, GM-CSF and IL-3, where the effect on granulopoiesis is transient), to ascertain whether erythropoiesis andor platelet production can be increased in vivo.

Acknowledgments The author’s work quoted was supported by the Cancer Research Campaign, UK and the South Manchester Endowment Fund (JCWM), and the Marrow Environment Fund (FMG).

266

References 1 Kern P, Heimpel H, Heit W, Kubanek B. Granulocytic progenitor cells in aplastic anaemia. Brit J Haematol 1977;35:613-623. 2 Faille A, Barrett AJ, Balitrand N, Ketels F, Gluckman E, Najean Y. Effects of antilymphocyte globulin on granulocyte precursors in aplastic anaemia. Brit J Haematol 1979;42:371-380. 3 Hara H, Kai S, Fushimi M, et al. Pluripotent haemopoietic precursors in vitro (CFU-Mix) in aplastic anaemia. Exp Hematol 1980;8:1165-1171. 4 Zoumbos NC, Gascon BP, Djeu JY, Trost SR,Young

NS. Circulating activated suppressor T lymphocytes in aplastic anaemia. N Engl J Med 1985;312:257-265. 5 Bacigalupo A, Podesta M, Mingari M, et al. Immune

suppression of haematopoiesis in aplastic anaemia: activity of T-gamma lymphocytes. J Immunol 1980; 125:149-1453. 6 Torok-Storb B. Johnson GG,Bowden R, Storb R. Gamma-interferon in aplastic anaemia: inability to detect significant levels in sera or demonstrate haematopoietic suppressing activity. Blood 1987;69:629-633. 7 Sullivan R, Quesenberry PJ, Parkman R, et al. Aplastic anaemia: lack of inhibitory effect of bone marrow lymphocytes on in vitro granulopoiesis. Blood 1980; 561625632. 8 De Planque MM, Bacigalupo A, Wursh A, et al. Long-term follow-up of severe aplastic anaemia patients treated with antithymocyte globulin. Brit J Haematol 1989;73:121-126. 9 Tong J, Bacigalupo A, Piaggio G, Figari 0, Sogno G, Marmont A. In vitro response of T cells from aplastic anaemia patients to antilymphocyte globulin and phytohaemagglutinin: colony stimulating activity and lymphokine production. Exp Hematol 1991;19: 3 12-316. 10 Marsh JCW, Gordon-Smith EC. The role of antilymphocyte globulin in the treatment of chronic acquired bone marrow failure. Blood Rev 1988;2: 141-148. 11 Torok-Storb B, Sieff C, Storb R, Adamson J, Thomas ED. In vitro tests for distinguishing possible immunemediated aplastic anaemia from transfusion-induced sensitisation. Blood 1980;55:211-215. 12 Kaminski ER, Hows JM. Goldman JM, Batchelor JR. Pretransfused patients with severe aplastic anaemia exhibit high numbers of cytotoxic T lymphocyte precursors probably directed at non-HLA antigens. Brit J Haematol 1990;76:401-405. 13 Camitta BM, Thomas ED, Nathan DG,et al. A prospective study of androgens and bone marrow transplantation for treatment of severe aplastic anaemia. Blood 1979;53:504-514.

267 14 Storb R, Champlin RE. Bone marrow transplantation for severe aplastic anaemia. BMT 1991;8:69-72. 15 Simmons PJ, Przepiorka D, Thomas ED, TorokStorb B. Host origin of marrow stromal cells following allogeneic bone marrow transplantation. Nature 1987;328:429-432. 16 Agematsu K, Nakahori Y. Recipient origin of bone marrow-derived fibroblastic stromal cells during all periods following bone marrow transplantation in humans. Brit J Haematol 1991;79:359-365. 17 Dexter TM, Allen TD, Lajtha LG. Conditions controlling the proliferation of haemopoietic stem cells in vitro. J Cell Physiol 1977;91:335-344. 18 Dexter TM, Coutinho LH. Spooncer E, et al. Stroma1 cells in haemopoiesis. In: Molecular Control of Haemopoiesis. Ciba Foundation Symposium. Wily, Chichester, England 1990;148:76-95. 19 Gartner S, Kaplan HS. Long-term culture of human bone marrow cells. Proc Natl Acad Sci USA 1980; 77:4756-4759. 20 Marsh JCW, Chang J, Testa NG. Hows JM, Dexter TM. The haematopoietic defect in aplastic anaemia assessed by long-term marrow culture. Blood 1990; 76: 1748-1757. 21 Juneja HS, Lee S , Gardner FH. Human long-term bone marrow cultures in aplastic anaemia. Int J Cell Cloning 1989;7:129-135. 22 Gibson FM, Gordon-Smith EC. Long-term culture of aplastic anaemia bone marrow. Brit J Haematol 1990;75:421-427. 23 Bruhl P,Mergenthalev HG, Dormer P. Haemopoietic inductive capacity of irradiated stromal cell layers in human micro long-term bone marrow cultures. Cell Tissue Kinet 1988;21:411-417. 24 Dexter TM, Spooncer E. Loss of immunoreactivity in long term bone marrow culture. Nature 1978;275: 135-136. 25 Gibson FM, Scopes J, Laurie A, Gordon-Smith EC. Contribution of the marrow stroma to the pathogenesis of aplastic anaemia. Exp Hematol 1990;18:303a. 26 Marsh JCW, Chang J, Testa NG. Hows JM, Dexter TM. In vitro assessment of marrow “stem cell” and stromal cell function in aplastic anaemia. Brit J Haematol 1991;78:258-267. 27 Hotta T, Kato T, Maeda H, Yamao H, Yamada H, Saito H. Functional changes in marrow stromal cells in aplastic anaemia. Acta Haematol 1985;74:65-69.

Aplastic Anemia In Vitro cyte-macrophage colony-stimulating factor on myelopoiesis in acquired immunodeficiency syndrome. N Eng J Med 1987;317:593-598. 30 Vadhan-Raj S , Keating M. LeMaistre A, et al. Effect of recombinant human granulocyte-macrophage colony stimulating factor in patients with myelodysplastic syndromes. N Eng J Med 1987;317:1545-1552. 31 Antman KS, Griffin JD, Elias A, et al. Effect of recombinant human granulocyte-macrophagecolonystimulating factor on chemotherapy-induced myelosuppression. N Eng J Med 1988;319:593-598. 32 Gibson FM. Bagnara M, Ionnidou E, Gordon-Smith EC. Interaction of granulocyte-macrophage colony stimulating factor and interleukin-3 in human longterm bone marrow culture. Exp Hematol 1992;20: 235-240. 33 Coutinho LH, Will A, Radford J, Schiro R, Testa NG, Dexter TM. Effects of recombinant human granulocyte colony-stimulating factor (CSF), human granulocyte-macrophage-CSF and gibbon interleukin-3 on hematopoiesis in human long-term bone marrow cultures. Blood 1990;75:2118-2129. 34 Laurie AS, Gibson FM, Gordon-Smith EC. In longterm bone marrow culture (LTBMC) patients with aplastic anaemia show impaired responsiveness to GM-CSF and IL-3. Brit J Haematol 1991;77(suppl 1):V059a. 35 Vadhan-Raj S, Buescher S , Broxmeyer HE, et al. Stimulation of myelopoiesis in patients with aplastic anaemia by recombinant human granulocytemacrophage colony stimulating factor. N Eng J Med 1988;319: 1628-1634. 36 Champlin RE, Nimer SD, Ireland P, Oette DH, Golde DW. Treatment of refractory aplastic anaemia with recombinant human granulocyte macrophage colony stimulating factor. Blood 1989;73:694-699. 37 Guinan EC. Sieff CA, Oette DH, Nathan DG. A phase I/II trial of recombinant granulocyte-macrophage colony-stimulating factor for children with aplastic anaemia. Blood 1990;76: 1077-1082. 38 Nissen C, Tichelli A, Gratwohl A, et al. Failure of recombinant human granulocyte-macrophagecolonystimulating factor therapy in aplastic anaemia patients withvery severe neutropenia. Blood 1988;72:2045-2047. 39 Sonoda Y, Yashige H, Fuji H, et al. Bilineage response in refractory aplastic anaemia patients following long-term administration of recombinant human granulocyte colony-stimulating factor. Eur J Haematol 1992;48:41-48.

28 Gibson FM, Scopes J, Gordon-Smith EC. Aplastic anaemia stromal cells produce greater colony-stimulating activity than normal stromal cells. Exp Hemato1 1991;19:261a.

40 Ganser A, Lindemann A, Seipelt G, et al. Effects of recombinant human interleukin-3 in aplastic anaemia. Blood 1990;76:1287-1292.

29 Groopman JE, Mitsuyasu RT. Deleo MJ, Oette D, Golde DW. Effect of recombinant human granulo-

41 Heyworth CM, Ponting IL, Dexter TM. The response of haemopoietic cells to growth factors: develop-

Gibson/Marsh/Gordon-Smith. mental implications of synergistic interactions. J Cell Sci 1988;91:239-247. 42 Nakao S, Matshushima K. Young N. Decreased interleukin-1 production in aplastic anaemia. Brit J Haematol 1989;71:431-436. 43 Childs B, Chasseing NA, Tomelden C, O’Reilly RJ, Castro-Malaspina H. Hypoproduction of interleukin-1 beta (IL-1 beta) by activated monocytes is specific for severe aplastic anaemia. Exp Hematol 1991;19:260a.

44 Walsh CE, Liu JM, Stacie M, Anderson J, et al. A trial of recombinant human interleukin-1 in patients with severe refractory aplastic anaemia. Brit J Haematol 1992;80:106-110.

268 57 Najeen Y, Haguenauer 0, for the cooperative group for the study of aplastic and refractory anaemia. Long-term (5 to 20 years) evolution of nongrafted aplastic anaemias. Blood 1990;76:2222-2228.

58 Appelbaum FR, Barrall J, Storb R, et al. Clonal cytogenetic abnormalities in patients with otherwise typical aplastic anaemia. Exp Hematol 1987;15: 1134-1139. 59 Van Kamp H. Landegent JE,Jansen RPM, Willemze R, Fibbe WE. Clonal hematopoiesis in patients with acquired aplastic anaemia. Blood 1991;78:3209-3214.

45 Witte ON. Steel locus defines new multipotent growth factor. Cell 1990;63:5-6.

60 Gale RE, Wheadon H. Linch DC. X-chromosome inactivation patterns using HPRT and PGK polymorphisms in haematologically normal and postchemotherapy females. Brit J Haematol 1991;79: 193-197.

46 Bernstein A, Chabot B, Dubreuil P,et al. The mouse Wlc-kit locus. In: Molecular Control of Haemopoiesis 1990. Ciba Foundation Symposium, Wiley, Chichester England;148:158-172.

61 Josten KM, Tooze JA, Borthwick-Clarke C, Gordon-Smith EC, Rutherford TR. Acquired aplastic anaemia and paroxysmal nocturnal hemoglobinuria: studies on clonality. Blood 1991;78:3162-3167.

47 Moore MAS. Clinical implications of positive and negative haematopoietic stem cell regulators. Blood 1991;78:1-19.

62 Young NS. The problem of clonality in aplastic anaemia: Dr. Dameshek’s riddle, restated. Blood 1992;79:1385-1392.

48 Abkowitz JL, Sabo KM. Nakamoto B, et al. DiamondBlackfan anaemia: in vitro response of erythroid progenitors to the ligand for c-kit. Blood 1991;9: 2198-2202.

63 Speck B, Tichelli A, Nissen C, Stebler C, Gratwohl A. Prognostic factors for patients with severe aplastic anaemia (SAA) treated with antilymphocyte globulin (ALG). Exp Hematol 1990;18:310a.

49 O’Reilly RJ, Brochstein J, Dinsmore R, Kirkpatrick D. Marrow transplantation for congenital disorders. Semin Haematol 1984;21:188.

64 Witherspoon RP, Storb R, Pepe M, Longton G, Sullivan KM.Cumulative incidence of secondary solid malignant turnours in aplastic anaemia patients given marrow grafts after conditioning with chemotherapy alone. Blood 1992;79:289-292.

50 Sirinavin C, Trowbridge AA. Dyskeratosis congenita: clinical features and genetic aspects. J Med Genet 1975;12:339-354. 51 Marsh JCW, Will AJ, Hows JM, et al. “Stem cell” origin of the hematopoietic defect in dyskeratosis congenita. Blood 1992;79:3138-3144. 52 Novitzky N, Jacob P. Marrow stem cell and stromal cell function in aplastic anaemia. Brit J Haematol 1991;79:531-533. 53 Gibson FM, Bagnara M, Scopes J, Gordon-Smith EC. Immunophenotyping of progenitor subsets in normal (N) and aplastic anaemia (AA) bone marrow. Exp Hematol 1992;20:386a. 54 Andrews RG, Singer JW, Bernstein ID. Precursors of colony-forming cells in humans can be distinguished from colony forming cells by expression of the CD33 and CD34 antigens and light scatter properties. J Exp Med 1989;169:1721-1731. 55 Camitta BM, Storb R, Thomas ED. Aplastic anaemia (first of two parts). Pathogenesis,diagnosis, treatment and prognosis. N Eng J Med 1982;306:645-652. 56 Tichelli A, Gratwohl A, Wursh A, Nissen C, Speck B. Late haematological complications in severe aplastic anaemia. Brit J Haematol 1988;69:413-418.

65 Marin P, Schrezenmeier H. Bacigalupo A. Van Lint MT, for the EBMT SAA working party. Relapse of aplasia following ALG treatment: a report of the SAA working party. BMT 1990;28:48a. 66 Van Kamp H, Jansen R, Willemze R, Fibbe WE, Landegent JE. Studies on clonality by PCR analysis of the PGK-1 gene. Nucl Acid Res 1991;19:2794. 67 Sutherland HJ, Lansdorp PM, Henkelman DH, Eaves AC, Eaves CJ. Functional characterisation of individual human hematopoietic stem cells cultured at limiting dilution on supportive marrow stromal layers. Proc Natl Acad Sci USA 199087:3584-3588. 68 Navarro S, De Vili N, LeCouedie JP, et al. Interleukin 6 and its receptor are expressed by human megakaryocytes: in vitro effect on proliferation and endoreplication. Blood 1991;77:461-471. 69 Teramura M, Kobayashi S, Hoshino S, Oshimi K, Mizoguchi H. Interleukin-11 enhances human megakaryocytopoiesis in vitro. Blood 1992;79:332-337.