REVIEW a Hypothesis for the Pathogenesis of Myelodysplastic Syndromes: Implications for New Therapies

A hypothesis for the pathogenesis of myelodysplastic syndromes: implications for new therapies C Rosenfeld1 and A List2

1Texas Oncology, PA, Dallas, TX; and 2University of Arizona Cancer Center, Tucson, AZ, USA

To guide development of new clinical strategies, a review of MDS based on cell culture, cytokine, molecular and clinical recent investigations in the pathobiology of MDS was perfor- research is presented (Figure 1). med. Articles were identiﬁed through a Medline search. Stud- ies, including reviews, are cited in the references. A multistep pathogenesis is proposed. (1) Targeted injury or mutation within hemopoietic stem cells may be followed by an immuno- Early events in evolution of MDS logic response adversely affecting progenitor survival. (2) Accelerated proliferation and premature death of marrow cells Three large (Ͼ150 index cases) epidemiologic studies suggest ␣ is ampliﬁed by apoptogenic cytokines (TNF- , Fas ligand). (3) that radiation, smoking and occupational exposure to pesti- Establishment of an abnormal clone associated with telomere cides, organic chemicals and heavy metals are risk factors for shortening. (4) Disease progression associated with loss of 10–12 tumor suppressor activity. Opportunities for therapeutic inter- the development of MDS. Prevention of MDS will require ventions are possible at each step. Comparisons between the proposed pathogenesis of MDS and severe aplastic anemia (SAA) are also presented. Leukemia (2000) 14, 2–8. Keywords: myelodysplastic syndrome; acute nonlymphocytic leukemia; refractory anemia; preleukemia

Introduction

Three decades of investigations into the pathophysiology of the myelodysplastic syndromes (MDS) have confirmed the heterogenicity of MDS and highlighted the complexity in disease biology.1 Recent advances in technology have yielded provocative observations. The objective of this review is to integrate clinical and laboratory findings into a working hypothesis for the development of idiopathic MDS, differen- tiate idiopathic MDS from SAA, and suggest new therapeutic strategies. Interpretation of data from MDS studies remains problem- atic. Without a reliable disease marker, there can be questions regarding the accuracy of an MDS diagnosis.2 Additional problems arise when patients with disparate biologies are compared. For example, patients with idiopathic MDS and therapy-related MDS are sometimes included in the same data analyses. The same is true for FAB morphologic type and cyto- genetics. A potential source of ambiguity in laboratory studies derives from the mixture of normal and malignant progenitor cells which are known to coexist.3 The low number of polyclonal progenitor cells in most cases suggests that such studies are valid. However, patients in chemotherapy-induced remission may re-establish polyclonal hemopoiesis.4,5 Any hypothesis for the pathogenesis of MDS must support some long-standing clinical observations. Why are cytopenias present with hypercellular marrows? Why does MDS evolve more slowly than AML? Why is idiopathic MDS predominantly a disease of the elderly? Why does MDS sometimes respond to therapies for SAA? Proposals for the pathogenesis of MDS have been suggested previously.6–9 A specific multistep sequence for the development of adult-onset idiopathic Figure 1 A proposed multistep sequence for the development of idiopathic myelodysplastic syndrome. M-CSF, macrophage colony- Correspondence: C Rosenfeld, Texas Oncology, PA, 7777 Forest stimulating factor; GM-CSF, granulocyte–macrophage colony-stimul- Lane, Building D 400, Dallas, Texas 75230, USA; Fax: 972–566–5819 ating factor; G-CSF, granulocyte colony-stimulating factor; ATG, anti- Received 28 June 1999; accepted 2 September 1999 thymocyte globulin; TNF, tumor necrosis factor. Pathogenesis of MDS C Rosenfeld and A List 3 further delineation of disease-associated toxins and possible Furthermore, overexpression of TNF-␣ mRNA from marrow polymorphisms in toxin metabolism that may predispose to was detected in most cases of MDS, but not in normal controls a higher risk of MDS.13 Clearly, additional epidemiologic or AML patients.37 One probable source of TNF-␣ overpro- studies are needed. duction is marrow macrophages which are increased in One proposal is that progenitor cells damaged by toxin MDS.38,39 The increased density of marrow macrophages may exposure or spontaneous mutation evoke an immunologic occur in response to elevated serum levels of M-CSF.40 Point response that further compromises progenitor cell growth and mutations in c-fms, which encodes the M-CSF receptor, may maturation. What is the evidence for the existence of an also promote macrophage development in some cases.41 The aberrant immunologic response when over 25 years of studies physiological significance of TNF-␣ in MDS is supported by indicate a diminished immune state in MDS?14 Incubation of several lines of investigation: (1) enhanced in vitro formation marrow cells with cyclosporine or removal of T cells enhance of CFU-GM by antibody neutralization of TNF-␣ in MDS but colony formation in some patients.15,16 Studies reported by no effect on AML CFU;42 (2) inverse correlation between Molldrem and colleagues16 at the NIH indicate that sup- serum TNF-␣ concentration and hemoglobin in one study;34 pression of CFU-GM may be mediated by CD8+ cells directed (3) inverse correlation between clinical response to erythro- against MHC class I restricted antigens. The anti-CFU-GM poietin and TNF-␣ levels;35 (4) inverse correlation between a response does not appear to be mediated by MDS-derived platelet response to IL-3 therapy and TNF-␣ serum levels;13 immune cells since, most, but not all, studies indicate that T (5) positive correlation between TNF-␣ producing cells in the cells are not clonal in MDS.17–22 Clinical observations also marrow and apoptosis;44 and (6) correlation between plasma support the notion of immune suppression of progenitor cell TNF-␣ concentration with nucleotide oxidation in marrow growth in MDS. Treatment with anti-thymocyte globulin or MDS CD34+ cells.45 cyclosporine can improve cytopenias in select MDS This model suggests that secretion of TNF-␣ or other pro- patients.23–26 In contrast, attempts to augment an immunologic apoptotic cytokines plays a pivotal role in the ineffective hem- response with roquinimex or IL-2 have met with very limited atopoiesis of MDS, but the relationship to disease progression suceess indicating that the diminished immune response may remains ill defined. This implies that strategies which effec- be the result rather than the cause of MDS.27,28 tively neutralize TNF may inprove hematopoiesis. Pentoxifyl- Non-clonal lymphopoiesis provides indirect evidence for a line at micromolar concentrations suppresses TNF-␣ mRNA lack of stem cell involvement in MDS. Using precursors sorted transcription. Combination therapy with pentoxifylline + cip- by flow cytometry and subsequent FISH to define clonal hem- rofloxacin yielded no hematologic benefit in one study but a opoiesis, primitive progenitors (CD34+, Thy1+) lacked the triple drug regimen of pentoxifylline + ciprofloxacin + cytogenetic marker whereas more committed progenitors dexamethasone produced hemopoietic responses in 35% (CD34+, CD33+) display a clonal chromosome abnormality.29 (18/51) of patients and 28% (5/18) of responders demonstrated Conceivably, these non-clonal primitive stem cells could be a cytogenetic response.46,47 An alternative approach is to neu- utilized as a stem cell graft for autologous transplantation. tralize circulating TNF-␣ by administration of soluble TNF The growth and differentiation of the progeny of clonal pro- receptors. In vitro, incubation of MDS marrow with TNFR:Fc genitors is further compromised by an accelerated rate of enhanced CFU-GM formation.42 Strategies which reduce the apoptotic cell death. In cell culture, cytokines such as TNF-␣ impact of multiple soluble mediators of progenitor cell and IFN-␥ can suppress the growth of hemopoietic progenitors apoptosis or increase the threshold for apoptosis induction and induce Fas expression on CD34 cells.30–32 What is the may be more effective than single agent therapy. For instance, evidence for a functional role of apoptogenic cytokines in interruption of progenitor cell apoptosis could be attempted MDS? Elevated serum levels of TNF-␣ in patients with MDS with soluble TNF receptor (to inhibit the initiation of is well documented (see Table 1).33–35 Increased TNF-␣ pro- apoptosis) plus amifostine (to raise the threshold for duction by blood mononuclear cells in one study was restric- apoptosis). Another possibility for combined therapy is simul- ted to patients with RA and RARS, but not RAEB or RAEBt.36 taneous inactivation of more than one inducer of apoptosis.

Table 1 Cytokine levels in MDS compared to normal controls

Colony-stimulating factors Pro-apoptotic factors

Serum Blood cells Marrow cells Serum Blood cells Marrow cells

G-CSF ↑34 ↓79 U51 GM-CSF ↑76,U34 ↓80,82,U51,N39 M-CSF ↑40 IL-3 ↑34,U76 IL-6 ↑34 ↑36* ↑51,N80 SCF ↓75,83 BPA ↓84 FLT3 ligand ↑85* TNF-␣ ↑33–35, ↑42* ↑36* ↑37,39,N80,82,U51 Fas ligand ↑42, ↑59‡ IL-1␤ ↑35 ↑36* ↑39,51 TGF-␤ ↑39,N51 IFN-␥ ↑37,U49

↑, increased; ↓, decreased; N, not different from normal controls; U, undetectable; *increased predominantly in patients with refractory anemia; ‡levels higher in RAEB/RAEBt than in RA/RARS.

Leukemia Pathogenesis of MDS C Rosenfeld and A List 4 One potential approach includes simultaneous blockade of tigate the cause(s) of cytopenia in established MDS. Numerous the activity of Fas ligand and TNF-␣.48 studies indicate deficient growth of myeloid, erythroid and IFN-␥ or IL-1␤ are apoptogenic cytokines that could con- megakaryocytic colonies.69 Similar to AML, blast cell colonies tribute to ineffective hematopoiesis in MDS. In two series, (CFU-L) can be detected in MDS.70 One cause for cytopenias IFN-␥ gene overexpression was detected in only 5/12 and may be related to diminished capacity for differentiation.71 0/11 MDS cases.37,49 Increased production of IL-1␤ by blood Therapeutic trials of differentiating agents have not demon- and marrow cells has been reported.36,50,51 Increased pro- strated consistent efficacy in ameliorating cytopenia.72–74 duction of IL-1␤ by cultured marrow mononuclear cells was Interpretation of the results from investigations of colony- detected in 13/32 MDS patients.50 Patients with RA tended to stimulating factor levels in MDS patients is difficult to discern. have the highest IL-1␤ production.36,50 IL-1␤ levels have been Results appear inconsistent (Table 1) and CSF levels may be correlated with the extent of apoptosis, but not proliferation.50 altered by clinical events (ie infections). Furthermore, in most Deficient production of the IL-1␤ receptor antagonist by MDS studies, serum CSF levels did not correlate with clinical para- stromal cells may give rise to unopposed apoptotic activity meters.34,75,76 Given these limitations, the decreased cellular from IL-1␤.52 These studies suggest that IFN-␥ and IL-1␤ do production of G-CSF, GM-CSF and burst-promoting activity not contribute to apoptosis in the majority of patients with (BPA) plus low serum levels of SCF suggest that low CSF levels MDS. may be related to cytopenias and provide a basis for CSF ther- Several investigators have reported increased marrow cell apy of MDS. Diminished elaboration of CSF may arise from apoptosis in MDS and have implicated the potential role of apoptosis and functional abnormalities of stromal cells.44,77,78 the Fas/Fas ligand system.53–58 Increased Fas expression was Studies indicate decreased production of GM-CSF and G-CSF detected on marrow CD34 cells from MDS patients.42,55,56 by MDS monocytes.79,80 The use of colony-stimulating factors Lack of correlation between Fas expression and apoptosis sug- for the treatment of MDS is beyond the scope of this paper.81 gests that multiple mediators of cell death are operational.55 59 Gupta and coinvestigators have examined Fas ligand MDS disease progression expression in marrows of MDS patients. The mean percentage of Fas ligand expressing cell was higher in MDS (17%) than The next proposed step is reduction of telomere length. Tel- from normal controls (6%). In contrast to normals where Fas omeres are located at the ends of eukaryotic chromosomes, ligand was detected mostly in lymphocytes, Fas ligand was and function to stabilize chromosomes. Telomere length is expressed in erythroblasts, myeloblasts, megakaryocytes, mat- progressively reduced by cell divisions. Accelerated apoptosis uring myeloid cells and dysplastic cells. In another study, Fas and proliferation may lead to the reduction in telomere length ligand expression in MDS patients was localized to marrow observed in MDS.53,86,87 Reduction in telomere length is prob- macrophages.60 Further investigations revealed more Fas ably not directly responsible for disease progression since only ligand positive marrow cells in RA/RARS (9%) than in 42% of patients with RA have telomere length reduction.87 RAEB/RAEBt (20%).59 However, this appears to be inconsist- Instead, telomere shortening may give rise to genomic insta- ent with the finding that the extent of marrow cell apoptosis bility that leads to cytogenetic evolution and disease pro- inversely correlates with clinical stage of MDS. A higher gression. This assertion is supported by the finding that telo- degree of apoptosis was seen in early FAB classes for both mere shortening is associated with advanced MDS, CD34+ cells or marrow aspirates in most studies.53–56 Further- cytogenetic abnormalities, percentage of marrow blasts, leu- more, as MDS clinically progresses, apoptotic signals decrease kemic transformation and poor prognosis.86,87 Since telomer- (Fas antigen, c-Myc oncoprotein) whereas anti-apoptotic sig- ase activity (a DNA polymerase that can synthesize the telo- nals increase (bcl-2 oncoprotein).54,55,61 Fas-associated phos- meric sequence) is normal to low in most patients, telomere phatase-1 (fap-1) is a negative regulator of fas. Recent studies length may be a better reflection of the pathophysiology in have shown that fap-1 expression is reduced in MDS marrow MDS.86 Progenitor cells with shortened telomeres may be cells compared to marrow cells from either normals or AML more susceptible to elimination by telomerase inhibitors. than has progressed from MDS.62 Several investigators have Telomerase inhibitors are currently in development.88 suggested that the clinical sequelae of apoptosis is cytopenia. Progression to advanced MDS and AML has been linked to Since MDS is usually detected by cytopenia and apoptotic inactivation of the tumor suppressor genes, p15INK4b and to a activity is most pronounced in the early phases of MDS, it is lesser extent, p53, and thereby contribute to clonal expansion possible that apoptosis precedes the clinical recognition of (see Figure 1). p53 is a tumor suppressor gene which serves 89 MDS. as a major control of the G1 checkpoint. Studies indicate that Amifostine is a phosphorylated aminothiol with dual bio- mutations of p53 occur in less than 20% of MDS cases.90–93 At logic activities including free radical scavenging, by addition the molecular level, p53 mutations in MDS are usually point of reducing equivalents, and inhibition of TNF-␣ and other or missense mutations of one allele, associated in some cases inflammatory cytokine elaboration.63 In this way, amifostine with 17 p deletion of the alternante allele.94 Evidence that sug- may protect against TNF-induced apoptosis in MDS.43 In one gests mutation of p53 is related to progression of MDS trial, amifostine responses were noted in 83% (15/18) includes the limitation to advanced MDS (RAEB, RAEBt), patients.64 In another study with 12 patients, none satisfied association with complex cytogenetic abnormalities, and risk the criteria for a partial or complete response.65 Theoretically, of secondary leukemia.89–91 In view of the low frequency of amifostine activity should be more pronounced in early, rather mutations, introduction of wild-type p53 into MDS cells by than, late MDS. However, the potential application of this gene therapy offers limited therapeutic potential.95 Bishop and agent in MDS awaits further investigation. Inhibition of coinvestigators96 investigated whether a rebound in p53 after apoptosis by agents that decrease the c-Myc/Bcl-2 ratio could brief inactivation of p53 RNA by antisense oligonucleotide also be attempted. There are candidate agents to decrease could inhibit clonal proliferation. In a phase I trial that

c-Myc (Vitamin D3, agents that potentiate intracellular included 10 high risk MDS patients (three RAEB, seven RAEB- cAMPcontent and antisence oligonucleotide).66–68 t), there was one transient response detected after intravenous Extensive cell culture studies have been performed to inves- infusion OL(1)p 53.

Leukemia Pathogenesis of MDS C Rosenfeld and A List 5 As noted earlier, Fas expression decreases as the blast per- pathophysiology but the disease expression is dependent on centage increases.55 The inhibitory effect of TGF␤ on leu- patient age. kemic colony formation in early MDS is diminished after There are a few other points suggested by the model. Mar- transformation to AML.97 Transcription of p15INK4b, a cyclin- row failure has a multifactorial etiology including decreased dependent kinase inhibitor represents one mechanism by marrow production of colony-stimulating factors, increased which TGF␤ exerts its inhibitory effect. Hypermethylation of elaboration of hemopoietic inhibitors and accelerated the p 15INK4b gene occurs in 38–50% of MDS patients and apoptosis of progenitor cells. As proposed previously, cyto- may contribute to loss of proliferative regulation.98,99 Evidence penia with a hypercellular marrow can be attributed to simul- implicating a causal role for inactivation of p 15INK4b with dis- taneous progenitor cell apoptosis, increased proliferation with ease progression is provided by the increased frequency of a differentiation block.44,53,71 hypermethylation observed in advanced MDS and secondary The sequential development of MDS does not explain all AML compared to early MDS. Agents which promote hypo- observations. This proposal provides a basis for designing new methylation of DNA may impact the evolution of the leukemic studies for what remains an enigmatic disease. clone by derepression of p 15 transcription.100 5-Aza-2Ј- deoxycytidine and 5-azacytidine promote DNA hypomethyl- ation by inhibition of DNA methyltransferase.101 However, References most of the activity of 5-azacytidine, a drug demonstrated to have clinical efficacy, is through mechanisms other than DNA 1 Greenberg PL, Nichols WC, Schrier SL. Granulopoiesis in acute demethylation.101,102 Incubation of cell lines harboring myeloid leukemia and preleukemia. New Engl J Med 1971; 22: methylation silenced p 15 with 5-Aza-2Ј-deoxycytidine leads 1225–1232. 103 2 Rosati S, Anastasi J, Vardiman J. Recurring diagnostic problems in to re-expression of p 15. In one study, a 72 h infusion of the pathology of the myelodysplastic syndromes. Semin Hematol 5-Aza-2Ј-deoxycytidine to 29 high risk MDS patients yielded 1996; 33: 111–126. a 29% complete response rate and 24% partial responses.104 3 Asano H, Ohashi H, Ichihara M, Kinoshita T, Murate T, Kobayashi Prolonged myelosuppression was the sole cause for a 17% M, Saito H, Hotta T. Evidence for nonclonal hematopoietic pro- drug-related mortality. Using a lower dose of 5-Aza-2Ј-deoxy- genitor cell populations in bone marrow of patients with myelo- cytidine in 61 patients produced similar clinical effects with- dysplastic syndromes. Blood 1994; 84: 588–594. 105 4 Delforge M, Demuynck H, Vandenberghe P, Verhoef G, Zacheé out marked myelosuppression. Additional studies with 5- P, Van Duppen V, Marijnen P, Van den Berghe H, Boogaerts M. Aza-2Ј-deoxycytidine should be performed. Polyclonal primitive hematopoietic progenitors can be detected in mobilized peripheral blood from patients with high-risk myelodysplastic syndromes. Blood 1995; 86: 3660–3667. Clinical relevance of model 5 Delforge M, Demuynck H, Verhoef G, Vandenberghe P, Zacheé P, Maertens J, Duppen VV, Boogaerts MA. Patients with high risk How does this model account for MDS being a disease of myelodysplastic syndrome can have polyclonal or clonal haemo- the elderly? Accumulated environmental exposures provide a poiesis in complete haematological remission. Br J Haematol cumulative probability of mutational events that increases 1998; 102: 486–494. with time. Age-related telomere shortening fosters heightened 6 List AF, Jacobs A. Pathogenesis and biology of myelodysplasia. susceptibility to genomic instability that can lead to emerg- Semin Oncol 1992; 29: 14–24. 7 Sanz GF, Sanz MA, Vallespi T. Etiopathogeny, prognosis and ther- ence of clonal disease. A similar explanation can account for apy of myelodysplastic syndromes. Hematol Cell Ther 1997; 39: the differences between MDS and SAA. The younger age of 277–294. SAA patients may reflect exposure to a progenitor cell toxin 8 Gallagher A, Darley RL, Padua R. The molecular basis of myelo- or a genetic event in a patient not susceptible to genomic dysplastic syndromes. Haematologica 1997; 82: 191–204. instability from age-related telomere loss. This proposal is sup- 9 Aul C, Bowen DT, Yoshida Y. Pathogenesis, etiology and epidemi- ported by the finding of shorter telomere lengths in MDS than ology of myelodysplastic syndromes. Haematologica 1998; 83: 106 71–86. SAA. Another difference between MDS and SAA may be 10 West RR, Stafford DA, Farrow A, Jacobs A. Occupational and the milieu of hemopoietic inhibitory cytokines that drive environmental exposures and myelodysplasia: a case–control apoptosis. In SAA, IFN-␥ gene overexpression was detectable study. Leukemia Res 1995; 19: 127–139. in most patients.49,107 As described earlier, IFN-␥ mRNA was 11 Nisse C, Grandbastien B, Brizard A, Hebbar M, Haguenoer JM, usually not detected in MDS patients, but TNF-␣ was over- Fenaux P. Myelodysplastic syndromes (MDS) and exposure to expressed in 11/14 cases.37,49 These data suggest a tendency occupational or environmental hazards: a case–control study. ␥ ␣ Blood 1997; 90 (Suppl. 1): 518a (Abstr.). for IFN- -mediated apoptosis in SAA and TNF- -driven 12 Rigolin GM, Cuneo A, Roberti MG, Bardi A, Bigoni R, Piva N, apoptosis in MDS. Some features of MDS and SAA are shown Minotto C, Agostini P, Angeli CD, Senno LD. Exposure to myelo- in Table 2. We propose that MDS and SAA may have a similar toxic agents and myelodysplasia: case–control study and correlation with clinobiologic findings. Br J Haematol 1998; 103: Table 2 Pathophysiologic factors in MDS and SAA 189–197. 13 Larson LA, Wang Y, Banerjee M, Wiemels J, Hartford C, Le Beau 609 Ͼ MDSa SAA MM, Smith MT. Prevalence of the inactivating C- T polymor- phism in the NAD(P)H: quinone oxidoreductase (NQO1) gene in patients with primary and therapy-related myeloid leukemia. ↑ 36,49,108 Hemopoietic inhibitors Yes Yes Blood 1999; 94: 803–807. ↓Progenitor cells Yes Yes109 110,111 14 Takaku S, Takaku F. Natural killer cell activity and preleukaemia. Apoptosis of marrow cells Yes Yes Lancet 1981; 2: 1178. Telomere shortening Yesb Yes106 112 15 List AF, Glinsmann-Gibson B, Spier C, Taetle R. In vitro and in Clonal hemopoiesis Yes Rare vivo response to cyclosporin-A in myelodysplastic syndromes ↓ ↑113 Production of G-CSF and GM- Often Normal or (MDS): identification of a hypocellular subset responsive to CSF immune suppression. Blood 1992; 80 (Suppl. 1): 28a (Abstr.). 16 Molldrem JJ, Jiang YZ, Stetler-Stevenson M, Mavroudis D, Hensel aSee text for details. N, Barrett AJ. Hematological response of patients with myelo- bMore pronounced telomere shortening in MDS than SAA (see text). dysplastic syndrome to antithymocyte globulin is associated with

Leukemia Pathogenesis of MDS C Rosenfeld and A List 6 a loss of lymphocyte-mediated inhibition of CFU-GM and alter- 35 Stasi R, Brunetti M, Bussa S, Conforti M, Martin LS, Presa ML, ations in T-cell receptor V␤ profiles. Br J Haematol 1998; 102: Bianchi M, Parma A, Pagano A. Serum levels of tumour necrosis 1314–1322. factor-␣ predict response to recombinant human erythropoietin in 17 Van Kamp H, Fibbe WE, Jansen RPM, van der Keur M, de Graaff patients with myelodysplastic syndrome. Clin Lab Haematol 1997; E, Willemze R, Landegent JE. Clonal involvement of granulocytes 19: 197–201. and monocytes, but not of T and B lymphocytes and natural killer 36 Koike M, Ishiyama T, Tomoyasu S, Tsuruoka N. Spontaneous cyto- cells in patients with myelodysplasia: analysis by X-linked restric- kine overproduction by peripheral blood mononuclear cells from tion fragment length polymorphisms and polymerase chain reac- patients with myelodysplastic syndromes and aplastic anemia. tion of the phosphoglycerate kinase gene. Blood 1992; 80: Leukemia Res 1995; 19: 639–644. 1774–1780. 37 Kitagawa M, Saito I, Kuwata T, Yoshida S, Yamaguchi S, Takahaski 18 Hamblin TJ. Immunological abnormalities in myelodysplastic syn- M, Tanizawa T, Kamiyama R, Hirokawa K. Overexpression of dromes. Semin Hematol 1996; 33: 150–162. tumor necrosis factor (TNF)-␣ and interferon (IFN)-␥ by bone mar- 19 Tefferi A, Thibodeau SN, Solberg LA. Clonal studies in the myelo- row cells from patients with myelodysplastic syndromes. Leuke- dysplastic syndrome using X-linked restriction fragment length mia 1997; 11: 2049–2054. polymorphisms. Blood 1990; 75: 1770–1773. 38 Kitagawa M, Kamiyama R, Kasuga T. Increase in number of bone 20 Kibbelaar RE, van Kamp H, Dreef EJ, de Groot-Swings G, Kluin- marrow macrophages in patients with myelodysplastic syndromes. Nelemans JC, Beverstock GC, Fibbe WE, Kluin M. Combined Eur J Haematol 1993; 51: 56–58. immunophenotyping and DNA in situ hybridization to study lin- 39 Shetty V, Mundle S, Alvi S, Showel M, Broady-Robinson L, Dar eage involvement in patients with myelodysplastic syndromes. S, Borok R, Showel J, Gregory S, Rifkin S, Gezer S, Parcharidou Blood 1992; 79: 1823–1828. A, Venugopal P, Shah R, Hernandez B, Klein M, Alston D, Robin 21 Anastasi J, Feng J, Le Beau M, Larson RA, Rowley JD, Vardiman E, Dominquez C, Raza A. Measurement of apoptosis, proliferation JW. Cytogenetic clonality in myelodysplastic syndromes studied and three cytokines in 46 patients with myelodysplastic syn- with fluorescence in situ hybridization: lineage, response to dromes. Leukemia Res 1996; 20: 891–900. growth factor therapy, and clone expansion. Blood 1993; 81: 40 Janowska-Wieczorek A, Belch AR, Jacobs A, Bowen D, Padua RA, 1580–1585. Paietta E, Stanley ER. Increased circulating colony-stimulating fac- 22 Kroef MJPL, Fibbe WE, Mout R, Jansen RPM, Haak HL, Wessels tor-1 in patients with preleukemia, leukemia, and lymphoid malig- JW, Van Kamp H, Willemze R, Landegent JE. Myeloid but not nancies. Blood 1991; 77: 1796–1803. lymphoid cells carry the 5 q deletion: polymerase chain reaction 41 Tobal K, Pagliuca A, Bhatt B, Bailey N, Layton DM, Mufti GJ. analysis of loss of heterozygosity using mini-repeat sequences on Mutation of the human FMS gene (M-CSF receptor) in myelodys- highly purified cell fractions. Blood 1993; 81: 1849–1854. plastic syndromes and acute myeloid leukemia. Leukemia 1990; 23 Sulecki M, Shadduck RK, Zeigler Z. Anti-thymocyte globulin for 4: 486–489. hypoplastic myelodysplastic syndrome. Blood 1988; 72 (Suppl. 1): 42 Gersuk GM, Beckham C, Loken MR, Kiener P, Anderson JE, Far- 229a (Abstr.). rand A, Trout AB, Ledbetter JA, Deeg HJ. A role for tumour 24 Biesma DH, van den Tweel JG, Verdonck LF. Immunosuppressive necrosis factor-alpha, Fas, and Fas-Ligand in marrow failure asso- therapy for hypoplastic myelodysplastic syndrome. Cancer 1997; ciated with myelodysplastic syndrome. Br J Haematol 1998; 103: 79: 1548–1551. 176–188. 25 Molldrem J, Caples M, Mavroudis D, Plante M, Young NS, Barrett 43 Ganser A, Ottmann OG, Seipelt G, Lindemann A, Hess U, Geissler AJ. Antithymocyte globulin for patients with myelodysplastic syn- G, Maurer A, Frisch J, Schulz G, Mertelsmann R, Hoelzer D. Effect drome. Br J Haematol 1997; 99: 699–705. of long-term treatment with recombinant human interleukin-3 in 26 Jona´s¯ova´ A, Neuwirtova´ R, C¯ erma´k J, Vozobulova´ V, Mocikova´ patients with myelodysplastic syndromes. Leukemia 1993; 7: K, S˘is¯kova´ M, Hochova´ I. Cyclosporin A therapy in hypoplastic 696–701. MDS patients and certain refractory anaemias without hypoplastic 44 Raza A, Mundle S, Shetty V, Alvi S, Chopra H, Span L, Parchari- bone marrow. Br J Haematol 1998; 100: 304–309. dou A, Dar S, Venugopal P, Borok R, Gezer S, Showel J, Loew J, 27 Rosenfeld CS, Zeigler ZR, Shadduck RK, Nilsson B. Phase II study Robin E, Rifkin S, Alston D, Hernandez B, Shah R, Kaizer H, Gre- of roquinimex in myelodysplastic syndrome. Am J Clin Oncol gory S. Novel insights into the biology of myelodysplastic syn- 1997; 20: 189–192. dromes: excessive apoptosis and the role of cytokines. Int J Hema- 28 Nand S, Stock W, Stiff P, Sosman J, Martone B, Radvany R. A tol 1996; 63: 265–278. phase II trial of interleukin-2 in myelodysplastic syndromes. Br J 45 Peddie CM, Wolf CR, McLellan LI, Collins AR, Bowen DT. Oxidat- Haematol 1998; 101: 205–207. ive DNA damage in CD34+ myelodysplastic cells is associated 29 Saitoh K, Miura I, Takahashi N, Miura AB. Fluorescence in situ with intracellular redox changes and elevated plasma tumour hybridization of progenitor cells obtained by fluorescence-acti- necrosis factor-␣ concentration. Br J Haematol 1997; 99: 625– vated cell sorting for the detection of cells affected by chromo- 631. some abnormality trisomy 8 in patients with myelodysplastic syn- 46 Nemunaitis J, Rosenfeld C, Getty L, Boegel F, Meyer W, Jennings dromes. Blood 1998; 92: 2886–2892. LW, Zeigler Z, Shadduck R. Pentoxifylline and ciprofloxacin in 30 Selleri C, Sato T, Anderson S, Young NS, Maciejewski JP. Inter- patients with myelodysplastic syndrome. Am J Clin Oncol 1995; feron-gamma and tumor necrosis factor-alpha suppress both early 18: 189–193. and late stages of hematopoiesis and induce programmed cell 47 Raza A, Gezer S, Venugopal P, Kaizer H, Hines C, Thomas R, Alvi death. J Cell Physiol 1995; 165: 538–546. S, Mundle S, Shetty V, Borok R, Lowe J, Reza S, Robin EL, Rifkin 31 Nagafuji K, Shibuya T, Harada M, Mizuno S, Takenaka K, Miya- SD, Alston D, Hernandez B, Shah R, Hsu WT, Dar S, Gregory SA. moto T, Okamura T, Gondo H, Niho Y. Functional expression of Hematopoietic and cytogenetic responses to novel anti-cytokine Fas antigen (CD95) on hematopoietic progenitor cells. Blood therapy in myelodysplastic syndromes (MDS). Proc ASCO 1997; 1995; 86: 883–889. 16: 7a (Abstr.). 32 Maciejewski J, Selleri C, Anderson S, Young NS. Fas antigen 48 Hattori K, Hirano T, Miyajima H, Yamakawa N, Tateno M, Oshimi expression on CD34+ human marrow cells is induced by inter- K, Kayagaki N, Yagita H, Okumura K. Differential effects of anti- feron ␥ and tumor necrosis factor ␣ and potentiates cytokine- Fas ligand and anti-tumor necrosis factor ␣ antibodies on acute mediated hematopoietic suppression in vitro. Blood 1995; 85: graft-versus-host disease pathologies. Blood 1998; 91: 4051– 3183–3190. 4055. 33 Zoumbos N, Symeonidis A, Kourakli A, Katevas P, Matsouka P, 49 Nistico` A, Young NS. Gamma-interferon gene expression in the Perraki M, Georgoulias V. Increased levels of soluble interleukin- bone marrow of patients with aplastic anemia. Ann Intern Med 2 receptors and tumor necrosis factor in serum of patients with 1994; 120: 463–469. myelodysplastic syndromes. Blood 1991; 77: 413–414. 50 Mundle SD, Venugopal P, Cartlidge JD, Pandav DV, Broday-Rob- 34 Verhoef GEG, De Schouwer P, Ceuppens JL, Van Damme J, Goos- inson L, Gezer S, Robin EL, Rifkin SR, Klein M, Alston DE, Hernan- sens W, Boogaerts MA. Measurement of serum cytokine levels in dez BM, Rosi D, Alvi S, Shetty VT, Gregory SA, Raza A. Indication patients with myelodysplastic syndromes. Leukemia 1992; 12: of an involvement of interleukin-1␤ converting enzyme-like pro- 1268–1272. tease in intramedullary apoptotic cell death in the bone marrow

Leukemia Pathogenesis of MDS C Rosenfeld and A List 7 of patients with myelodysplastic syndromes. Blood 1996; 88: myelodysplastic syndromes and acute myeloid leukemia. Blood 2640–2647. 1992; 80: 625–633. 51 Wetzler M, Kurzrock R, Estrov Z, Estey E, Talpaz M. Cytokine 71 Sawada K, Sato N, Notoya A, Tarumi T, Hirayama S, Takano H, expression in adherent layers from patients with myelodysplastic Koizumi K, Yasukouchi T, Yamaguchi M, Koike T. Proliferation syndrome and acute myelogenous leukemia. Leukemia Res 1995; and differentiation of myelodysplastic CD34+ cells: phenotypic 19: 23–34. subpopulations of marrow CD34+ cells. Blood 1995; 85: 194– 52 Holmberg L, Seidel K, Leisenring W, Torok-Storb B. Cytokine pro- 202. duction by long-term marrow cultures established from myelo- 72 Santini V, Ferrini PR. Differentiation therapy of myelodysplastic dysplastic patients. Exp Hematol 1994; 22: 741 (Abstr.). syndromes: fact or fiction? Br J Haematol 1998; 102: 1124–1138. 53 Raza A, Gezer S, Mundle S, Gao XZ, Alvi S, Borok R, Rifkin S, 73 Mellibovsky L, Diez A, Pe´rez-Vila E, Serrano S, Nacher M, Aubiá Iftikhar A, Shetty V, Parcharidou A, Loew J, Marcus B, Khan Z, J, Supervıá A, Recker RR. Vitamin D treatment in myelodysplastic Chaney C, Showel J, Gregory S, Preisler H. Apoptosis in bone syndromes. Br J Haematol 1998; 100: 516–520. marrow biopsy samples involving stromal and hematopoietic cells 74 Howard MR, Bond LR. Calcitriol in myelodysplastic syndromes. in 50 patients with myelodysplastic syndromes. Blood 1995; 86: Br J Haematol 1999; 106: 569–574. 268–276. 75 Bowen D, Yancik S, Bennett L, Culligan D, Resser K. Serum stem 54 Rajapaksa R, Ginzton N, Rott LS, Greenberg PL. Altered oncopro- cell factor concentration in patients with myelodysplastic syn- tein expression and apoptosis in myelodysplastic syndrome mar- dromes. Br J Haematol 1993; 85: 63–66. row cells. Blood 1996; 88: 4275–4287. 76 Zwierzina H, Scho¨llenberger S, Herold M, Schmalzl F, Besemer 55 Bouscary D, De Vos J, Guesnu M, Jondeau K, Viguier F, Melle J, J. Endogenous serum levels and surface receptor expression of Picard F, Dreyfus F, Fontenay-Roupie M. Fas/Apo-1(CD95) GM-CSF and IL-3 in patients with myelodysplastic syndromes. expression and apoptosis in patients with myelodysplastic syn- Leukemia Res 1992; 16: 1181–1186. dromes. Leukemia 1997; 11: 839–845. 77 Gida´li J, Fehe´r I, Hollań S. Blast colony forming cell-binding 56 Parker JE, Mufti GJ. Ineffective haemopoiesis and apoptosis in capacity of bone marrow stroma from myelodysplastic patients. myelodysplastic syndromes. Br J Haematol 1998; 101: 220–230. Stem Cells 1996; 14: 577–583. 57 Irvine AE, Magill MK, Somerville LE, McMullin MF. Spontaneous 78 Tennant GB, Truran LN, Kell WJ, Burnett AK. Myelodysplastic intramedullary apoptosis is present in disorders other than myelo- bone marrow stroma is defective in its ability to sustain normal dysplasia. Exp Hematol 1998; 26: 435–439. stem cell growth. Blood 1997; 90 (Suppl. 1): 582a (Abstr.). 58 Bogdanovic´ AD, Trpinac DP, Janokovic´ GM, Bumbas¯irevic´VZ¯ , 79 Greenberg PL, Mackichan ML, Negrin R, Renick M, Ginzton N. ¯ Obradovic´ M, Colovic´ MD. Incidence and role of apoptosis in Production of granulocyte colony stimulating factor (G-CSF) by myelodysplastic syndrome: morphological and ultrastructural normal and myelodysplastic syndrome (MDS) peripheral blood assessment. Leukemia 1997; 11: 656–659. (PB) cells. Blood 1990; 76: (Suppl. 1) 146a (Abstr.). 59 Gupta P, Niehans GA, LeRoy SC, Gupta K, Morrison VA, Schultz 80 Visani G, Zauli G, Tosi P, Fogli M, Cenacchi A, Finelli C, Gamberi C, Knapp DJ, Kratzke RA. Fas ligand expression in the bone mar- B, Ottaviani E, Manfoi S, Tura S. Impairment of GM-CSF pro- row in myelodysplastic syndromes correlates with FAB subtype duction in myelodysplastic syndromes. Br J Haematol 1993; 84: and anemia, and predicts survival. Leukemia 1999; 13: 44–53. 227–231. 60 Kitagawa M, Yamaguchi S, Takahaski M, Tanizawa T, Hirokawa 81 Ganser A, Karthaus M. Clinical use of hematopoietic growth fac- K, Kamiyama R. Localization of Fas and Fas ligand in bone mar- tors in the myelodysplastic syndromes. Leuk Lymphoma 1997; 26 row cells demonstrating myelodysplasia. Leukemia 1998; 12: (Suppl. 1): 13–29. 486–492. 82 Mencoboni M, Castello G, Lerza R, Haupt E, Ballarine P, Cerruti 61 Davis RE, Greenberg PL. Bcl-2 expression by myeloid precursors A, Bogliolo G, Pannacciulli I. Production of tumor necrosis factor in myelodysplastic syndrome: relation to progression. Leukemia and granulocyte colony stimulating factor by bone marrow access- Res 1998; 22: 767–777. ory cells in myelodysplastic patients. Eur J Haematol 1996; 56: 62 Mundle S, Mativi BY, Bagai K, Feldman G, Cheema G, Gautam 148–152. U, Reza S, Cartlidge J, Venugopal P, Shetty V, Gregory SA, Raza 83 Cortelezzi A, Cattaneo C, Cristiani S, Sarina B, Pomati M, Silvestris A. A possible link of tumor necrosis factor a (TNFa) with down- I, Ibatici A, Maiolo AT. Low plasma stem cell factor levels corre- regulation of Fas-associated phosphatase-1 (fap-1) and activation late with in vitro leukemic growth in myelodysplastic syndromes. of Fas-mediated apoptosis in myelodysplasia. J Clin Oncol 1999; Leukemia Res 1999; 23: 271–275. 18: 22a (Abstr.). 84 Merchav S, Nagler A, Fleischer-Kurtz G, Tatarsky I. Regulatory 63 Klimecki W, Heaton R, Gibson BG, List AF. Amifostine suppresses abnormalities in the marrow of patients with myelodysplastic syn- apoptosis in myelodysplastic CD34+ cells and promotes progeni- dromes. Br J Haematol 1989; 73: 158–164. tor cell growth via polyamine-like effects. Blood 1997; 90 (Suppl. 85 Zwierzina H, Anderson JE, Rollinger-Holzinger I, Torok-Storb B, 1): 520a (Abstr.). 64 List AF, Brasfield F, Heaton R, Glinsmann-Gibson B, Crook L, Nuessler V, Lyman SD. Endogenous FLT-3 ligand serum levels are Taetle R, Capizzi R. Stimulation of hematopoiesis by amifostine associated with disease stage in patients with myelodysplastic syn- in patients with myelodysplastic syndrome. Blood 1997; 90: dromes. Leukemia 1999; 13: 553–557. 3364–3369. 86 Ohyashiki JH, Iwama H, Yahata N, Ando K, Hayashi, Shay JW, 65 Bowen DT, Denzlinger C, Brugger W, Culligan D, Gelly K, Ohyashiki K. Telomere stability is frequently impaired in high-risk Adlakha S, Groves M, Hepurn M, Kanz L. Poor response rate to groups of patients with myelodysplastic syndromes. Clin Cancer a continuous schedule of amifostine therapy for ‘low/intermediate Res 1999; 5: 1155–1160. risk’ myelodysplastic patients. Br J Haematol 1998; 103: 785–787. 87 Boultwood J, Fidler C, Kusec R, Rack K, Elliott JW, Atoyebi O, 66 Garcia-Bermejo L, Pe´rez C, Vilaboa N, de Blas E, Aller P. cAMP Chapman R, Oscier DG, Wainscoat JS. Telomere length in myelo- increasing agents attenuate the generation of apoptosis by etopo- dysplastic syndromes. Am J Hematol 1997; 56: 266–271. side in promonocytic leukemia cells. J Cell Sci 1998; 111: 637– 88 Sharma S, Raymond E, Soda H, Sun D, Hilsenbeck SG, Sharma 644. A, Izbicka E, Windle B, Von Hoff DD. Preclinical and clinical 67 Caligo MA, Cipollini G, Petrini M, Valentini P, Bevilacqua G. strategies for development of telomerase and telomere inhibitors. Down regulation of NM23.H1, NM23.H2 and c-myc genes during Ann Oncol 1997; 8: 1063–1074. differentiation induced by 1,25 dihydroxyvitamin D3. Leukemia 89 Kirsch DG, Kastan MB. Tumor-suppressor p53: implications for Res 1996; 20: 161–167. tumor development and prognosis. J Clin Oncol 1998; 16: 68 Calabretta B, Skorski T. Targeting c-myc in leukemia. Anticancer 3158–3168. Drug Des 1997; 12: 373–381. 90 Sugimoto K, Hirano N, Toyoshima H, Chiba S, Mano H, Takaku F, 69 Greenberg PL. Biologic and clinical implications of marrow cul- Yazaki Y, Hirai H. Mutations of the p 53 gene in myelodysplastic ture studies in the myelodysplastic syndromes. Semin Hematol syndrome (MDS) and MDS-derived leukemia. Blood 1993; 81: 1996; 33: 163–175. 3022–3026. 70 Aul C, Gattermann N, Schneider W. Comparison of in vitro growth 91 Wattel E, Preudhomme C, Hecquet B, Vanrumbeke M, Quesnel characteristics of blast cell progenitors (CFU-L) in patients with B, Dervite I, Morel P, Fenaux. p 53 mutations are associated with

Leukemia Pathogenesis of MDS C Rosenfeld and A List 8 resistance to chemotherapy and short survival in hematologic 103 Herman JG, Jen J, Merlo A, Baylin SB. Hypermethylation-associa- malignancies. Blood 1994; 84: 3148–3157. ted inactivation indicates a tumor suppressor role for p15INK4B1. 92 Mitani K, Hangaishi A, Imamura N, Miyagawa K, Ogawa S, Kanda Cancer Res 1996; 56: 722–727. Y, Yazaki Y, Hirai H. No concomitant occurrence of the N-ras 104 Wijermans PW, Krulder JWM, Huijgens PC, Neve P. Continuous and p 53 gene mutations in myelodysplastic syndromes. Leukemia infusion of low-dose 5-Aza-2Ј-deoxycytidine in elderly patients 1997; 11: 863–865. with high-risk myelodysplastic syndrome. Leukemia 1997; 11: 93 Adamson DJA, Dawson AA, Bennett B, King DJ, Haites NE. p 53 1–5. mutation in the myelodysplastic syndromes. Br J Haematol 1995; 105 Lu¨bbert M, Haak HL, Kunzmann R, Danzer-Driessen S, Verhoef 89: 61–66. G, Mertelsmann R, Wijermans P. Cytogenetic response to low- 94 Preudhomme C, Fenaux P. The clinical significance of mutations dose 5-Aza-2Ј-deoxycytidine (DAC) in poor-risk myelodysplastic of the p 53 tumour suppressor gene in haematological malig- syndromes (MDS)–phase II study results. Blood 1997; 90: 582a nancies. Br J Haematol 1997; 98: 502–511. (Abstr.) 95 Nielsen LL, Maneval DC. p 53 tumor suppressor gene therapy for 106 Ball SE, Gibson FM, Rizzo S, Tooze JA, March JCW, Gordon- cancer. Cancer Gene Ther 1998; 5: 52–63. Smith EC. Progressive telomere shortening in aplastic anemia. 96 Bishop MR, Iversen PL, Bayever E, Sharp JG, Greiner TC, Copple Blood 1998; 91: 3582–3592. BL, Ruddon R, Zon G, Spinolo J, Arneson M, Armitage JO, Kess- 107 Nakao S, Yamaguchi M, Shiobara S, Yokoi T, Miyawaki T, Tani- inger A. Phase I trial of an antisense oligonucleotide OL(1) p 53 guichi T, Matsuda T. Interferon-␥ gene expression in unstimu- in hematologic malignancies. J Clin Oncol 1996; 14: 1320–1326. lated bone marrow mononuclear cells predicts a good response 97 Masuya M, Kita K, Shimizu N, Ohishi K, Kataya N, Sekine T, to cyclosporine therapy in aplastic anaemia. Blood 1992; 79: Otsuji A, Miwa H, Shirakaw S. Biologic characteristics of acute 2532–2535. leukemia after myelodysplastic syndrome. Blood 1993; 81: 108 Hinterberger W, Adolf G, Bettelheim P, Geissier K, Huber C, 3388–3394. Irschick E, Kalhs P, Ko¨ller U, Lechner K, Meister B, Woloszcszuk 98 Uchida T, Kinoshita T, Nagai H, Nakahara Y, Saito H, Hotta T, W. Lymphokine overproduction in severe aplastic anemia is not Murate T. Hypermethylation of the p15INK4B gene in myelodys- related to blood transfusions. Blood 1989; 74: 2713–2717. plastic syndromes. Blood 1997; 90: 1403–1409. 109 Young NS, Maciejewski J. The pathophysiology of acquired 99 Quesnel B, Guillerm G, Vereecque R, Wattel E, Preudhomme C, Bauters F, Vanrumbeke M, Fenaux P. Methylation of the p15INK4B aplastic anemia. New Engl J Med 1997; 336: 1365–1372. gene in myelodysplastic syndromes is frequent and acquired dur- 110 Philpott NJ, Scopes J, Marsh JC, Gordon-Smith EC, Gibson FM. ing disease progression. Blood 1998; 91: 2985–2990. Increased apoptosis in aplastic anemia bone marrow progenitor 100 Bender CM, Zingg JM, Jones PA. DNA methylation as a target cells: possible pathophysiologic significance. Exp Hematol 1995; for drug design. Pharm Res 1998; 15: 175–187. 23: 1642–1648. 101 Momparler RL, Cote S, Eliopoulos N. Pharmacological approach 111 Callera F, Falcao RP. Increased apoptotic cells in bone marrow for optimization of the dose schedule of 5-Aza-2Ј-deoxycytidine biopsies from patients with aplastic anaemia. Br J Haematol (Decitabine) for the therapy of leukemia. Leukemia 1997; 11: 1997; 98: 18–20. 175–180. 112 Raghavachar A, Janssen JWG, Schrezenmeier H, Wagner B, Bar- 102 Silverman LR, Demakos EP, Peterson B, Odchimar-Reissig R, tram CR, Schulz AS, Hein C, Cowling G, Mubarik A, Testa NG, Nelson D, Kornblith AB, Stone R, Holland JC, Powell BL, DeCa- Dexter TM, Hows JM, Marsh JCW. Clonal hematopoiesis as tro C, Ellerton J, Larson RA, Schiffer CA, Holland JF. A ran- defined by polymorphic X-linked loci occurs infrequently in domized controlled trial of subcutaneous azacitidine (AZA C) in aplastic anemia. Blood 1995; 86: 2938–2947. patients with the myelodysplastic syndrome (MDS): a study of 113 Kojima S, Matsuyama T, Kodera Y. Hematopoietic growth factors the Cancer and Leukemia Group B (CALGB). Proc ASCO 1998; released by marrow stromal cells from patients with aplastic 17: 14a (Abstr.). anemia. Blood 1992; 79: 2256–2261.

Leukemia