Document o9vXJgL45DQXKKy8mnZwqoNxD

392 Thepathogenesisofaplasticanaemia c C G G a r y , Royal Infirmary and Christie Hospital, Manchester ~ The association of fatty atrophy of haemopoietic although there is no doubt as to the association marrow with multiple cytopenias in the blood was fkst between this drug and marrow aplasia. Drugs known noted by Ehrlich in 1888. He described the case of a to cause panhypoplasia may also cause selective young pregnant woman whose disease ran a hypoplasia .of one or more cell lines, but the latter is particularly fulminating course, buf the term aplastic more often reversible when treatment is stopped. anaemia was probably not introduced until 1904 On the other hand there are undoubted when Chauffard used it to describe the syndrome. similarities between the two syndromes. The cultural Although these earlv accounts. based on autousv find- characteristicsof marrow from both groups of patients ings, eLphasized h-ypoplasia of the marrow &-a car- are often indistinguishable (see below). The same dinal diagnostic feature, the terminology later became qualitative abnormalities of erythropoiesis and . confused because, for some cime, "aplastic anaemia" granulopoiesis are found in patients with was virtually synonymouswith blood pancytopenia. predominantly hypoplastic mamws as in those with Benzene, the first substance found to cause marrow hyperplastic ineffective marrows. Occasionally damage, is indeed unusual in producing pancytopenia patients presenting with isolated cytopenias (such as with both hypocellular and hypercellular marrow red cell aplasia) later develop panhypoplasia, and pictures, but it was not until marrow aspiration children with constitutional marrow failure (such as became a routine clinical investigation that the Fanconi's anaemia) are sometimes found to have morphology of the marrow in pancytopenic patients cellular mamws at diagnosis but later progress to was studied in detail. Patients whose marrow was profound hypoplada. cellular or hypercellular were subsequently recognized as cases of hypersplenism, or classified somewhat Normal haemopoiesis unsatisfactorily as having achrestic anaemia, The daily production of differentiated blood cells is refractory anaemia with cellular mamw. or (in kinetic maintained by the proliferation of stem cells in the terms) functional aplasia. The paper by Bomford and, marrow. Theseoccur in all rapidly dividing tissuesand Rhoads (1941) which correlated the histological are defined as cells that both maintain their own appearances of the marrow with blood film numbers and give rise to differentiated cells. The appearances was a notable landmark. Later the haemopoietic system is in fact a cell renewal system in importance of trephine biopsy for assessing marrow which a continuous supply of highly specialized non- cellularity was realized, while the recent introduction dividing blood cells is sustained by the proliferation of plastic embedding for processing these histological and differentiation of such ancestral pluripotent cells. sections permits detailed scrutiny of residual The amibute of "sternness" is not lost abruptly and haemopoietic cells and their relationship to the there is at least one other major stratum of stem cells stroma. that fulfils the criteria defined above but is committed It is now redized that patients with aplastic to a particular line of differentiation (for example. 'maemia frequently show islands of cellular. dthoueh abnormal. haemopoiesis k n o m as "hot pockets" (Kansu and Erslev. 1976): indeed complete aplasia would be incompatible with life. However. there still seem to be reasonable clinical grounds erythroidstem cellsproduce only erythrocytesunder the influence of erythropoietin). Subsequent re rateddivisions in the more differentiated progeny o these -committed cells sometimes called the transit corn- -partment produces an amplifying effect, so that for distinguishing between cytopenic patients with except under ccnditions of stress, stem cells need onl: predominantly hypocellular and those with divide intermittently. Moreover it CM be shown hypercellular marrows. Thus. isolated cytopenk are experimentally that comparatively normal production more frequent in the latter group and the natural of mature blood cells can be achieved with a stem cel: history is different: leukaemia may be a commoner compartment that has undergone serious depletion. sequel (Vilter et al. 1967) while others respond to The long resting phase that characterizes the kinetic: splenectomy or immunosuppressants. Also, the of the individual stem cell has biological advantages: i aetiology is probably different. Apart from benzene permits a period for "genetic housekeeping" (Lajtha and possibly X-irradiation, no substance is known to 1978) and reduces the risks of p- rop.ag-ation of abnm cause chronic marrow failure with both hypercellular' mal clones. and hypocellular m m w pictures. Butatdidin has Stem cells or their progeny can be assayed both ir &n suspected of causing a picture indistinguishable vivo and in vitro. m e mouse spleen a s k y uses fi: L-- - - x l - * * l - - - - i - hlqt t h i c rPni:!ins unDmvetl. irradiated animal (Td and blcCulloch. 1961) a i I .-' 394. detects cells known as colony-forming units (CFU-S) haemopoiesisin m m w and Spleen appearsto depend which have the properties outlined above. They are on the presence of 2 specialized microenvironment that capable of self-renewal and give rise to clonal colonies determines the growth and differentiation of brood that contain cells of the granulocytic, erytiiroid, and cells; in this animal at least, bone marrow appears to megakaryocytic series (Siinovitch et al, 1963). For be a highly organized tissue in which the distribution obvious reasons this assay system cannot be used in of granulocytic (CFU-C) and erythroid (CFU-E) humans, but methods have been developed for precursor cells is not random (Lord et al, 1975). growing haemopoietic cells in soft agar in the presence there is evidence that environmental factors exert 9 of a suitable conditioning medium. In the m o w it is different regulatory influences on ancestral stem Ti now possible to grow mixed colonies of unicellular origin that give rise to all three cell types, as in the and may induce them to develop along one or other line of differentiation. r. -> +spleen assay (Johnson and Metcalf, 1977), but in man Although erythropoietin is the best understood of it has so far only been possible to estimate the the humoral factors affecting blood cell development, - 2 pluripotent stem cell population indirectly by studying granulopoiesis proceeds under the influence of an the numbers of their descendants. analogous glycoprotein, colony-stimulating factor, and - -These progeny of the CFU-S produce, only one thrombopoiesis and lymphopoiesis are +so regulated type of differentiated ceI1 line erythroid, gran- by specificfactors. Kegenerating m m w rn the m o w ulccytidrnonoqdc, or megakaryocytic. There are produces a stimulatory factor that increases stem cell two types of erythroid colony: the burst-form turnover, while normal marrow (in which the CFU-S unit (BFU-E) which is only slightly sensitive to are dividing only very slowly) specifically inhibits the erythropoietin (although it is probably under some proliferation of CFU-S obtained from regenerating other humoral control) and the CFU-E which is m m w (Lord et al, 1977). It is possible that, in the exquisitely sensitive to erythropoietin in culture. The intact animal, the rate of pluripotent stem cell CFU-C produce coloniescontaining cells exclusively of turnover is adapted to satisfy differentrequirementsby the granulocytic and monocytic series and are formed balancing the stimulatory and inhibitory substances. only when colony-stimulating activity is present in the Although little is known about the identity of the conditioning medium. Assays of such colonies have marrow stromal cells that regulate haemopoiesis, it is shown that bone marrow from patients with aplastic certain that macrophages and "epithelioid" cdls anaemia possesses reduced or undetectable numbers (which may really be of endothelial origin) are of progenitor cells, both of CFU-C (Greenberg essential for the propagation of CFU-S and CFU-C in and Schrier, 1973; Kern et al, 1977) and of CFU-E culture (Allen and Dexter, 1976; Dexter et al, 1978) (Hansi et al, 1977)lies. while the cooperation of a population of T lymphocytes Other in-vitro colony assays that detect progenitors is required to establish eqrthroid celllines in the mwse of megakaryocytes, B lymphocytes, and T lymphocytes (Jedrzejczak et al, 1977). have been developed recently. Studies of genetic It is clear therefore that the control of n m a l markers in patients with certain types of haemopoiesis is extremely complex. Haemopoietic myeloproliferative disorder have ,provided tentative failure, that is aplastic anaemia, can arise as the result evidence that B lymphocytes. at least, are derived from of a number of different lesions. In addition to direct an ancestral cell common to other haemopoietic cell damage to stem cells (Heimpel and Kubanek, 1975) lines. This may explain why some patients with Severe defects in the marrow stroma (Knospe and Crosby, aplastic anaemia have lymphopenia involving B-cell 1971), regulatory disturbances (Stohlman 1972), and populations and sometimes hypogammaglobulin- autoimmune damage to the marrow (Ascensao et al, aemia (Morley et al, 1974; Mir et al, 1977). 1976) have been postulated. The subject has recently been d e w e d by Alter et al (1978) and Gary and The marrow microenvironment Testa (1979) among others. The possible pathogenetic Haemopoietic stem cells do not exist in isolation and mechanisms are summarized in Table I. numerous observations testify to the importance of both cellular and humoral influences in the regulation Table i. Possible lesions in aplastic anaemia of replication and differentiation. In a series of elegant Pluripotent stem cell reduced numbers experiments using graduated doses of irradiation defective (inuinsri function Knospe and Crosby (1971) showed that an intact Committed stem cells' marrow vasculature is essential for haemopoiesis: stem Emuonrncntrl influences delecfr in baicular or other `stromal * cc& cells could not repopulate fatty marrow unless the specialized sinusoidal matrix was present. and marrow ahnorm.litier nf shori.ranuc .x Ion*-ranhumoral h n n n influencine stem CCII ym.Th transplantation was not possible in a heavily irradiated area unless this was curetted to promote vascularization. Perhaps the most striking example of the M i i t i o n and darsification of causes of aplastic importance of the marrow microenvironment is anaemin. provided by the genetically ylaemic S1/Sld (Steel) Aplastic anaemia can be defined as "a syndrome of mouse. This animal possesses normal stem cells, as peripheral blood .pancytopenia without dominant shown by the spleen colony assay and by the capacity peripheral blood CRU destruction, associated with of its marrow to reconstitute haemopoiesis in lethally hrpocellularity of haemopoietic tissue in both irradiated mice, but has a defect in its haemopoietic intramedullary and extramedullary sites, and without environment that impairs the proliferation of normal bone marrow fibrosis or invasion by malignant cells" stem cells. However, the anaemia can be cured by an (Benestad, 1974; Heimpel and Kubanek, 1975). The implant of spleen tissue which provides a normal absence of extramedullary haemopoiesis in aplastic environment into which the animal's own stem cells anaemia is particularly significant and indicates an BHroirtpisirhilloMurendaicl ionIe \ " m l 11)-9 can mimate and differentiate (Russell and Ekrnstein, inability of surviving stem cells in the marrow to v 1966). migrat; and populvate potentially fertile sites for I n the normal adult mouse the distribution of haemopoiesis in the spleen and the liver. ,399 . c Bntlsh Journal of HorpitalMcdrlne It is possible for aplastic anaemia to occur as a between 1966-70 the incidence of marrow apla result of an agent damaging simultaneously all cell caused by b u t a n e s was almost three times grea lines in the transit compartment. However, this than that for chloramphenicol due to the m( implies that the pathogenetic process is acting frequent use of the former drugs by that time (Bottii continuously, since the remarkable regenerative and Westerholm, 1973). Penicillamine and Brufen ; capacity of the pluripotent stem cell compartment among the newer drugs believed to cause aplu would be expected to replenish the differentiating anaemia compartments following transient inhibition. Damage at this level results more frequently in isolated Table 3. Some of the currently most important cytopenias such as red cell aplasia or agranulocytosis &ugs associated with m a m w aplasia although, as already emphasized, these can occasionally progress into panhypoplasia. In such cases there is presumably less severe damage to stem cells and the production of one or more mature cell lines is maintained by physiological regulating Antibvticr and ebemothcraputie drugs Anti-inflammatory and antirhcumrlk drugs &!orunphenieol subhonamlds phaylbutazaw oxmhmbuluoo. gold compounds * mdometh8d. ~ mechanisms in the early stages of the disease. Patients Annepdcpnc drugs presenting with incomplete pancytopenias may have a dipbcaylhydantoin troxidonc better prognosis than those who show complete Anadiabetic drugs chbrpropamide pancytopenia from the outset (Heimpel, 1979). Annthyrad drugs. potassium pcrchloratc Paroxysmal nocturnal haemoglobinuna and, less Ihiourxclll commonly, acute leukaemia occur as late Tranqudlizcrs chlorpromazine manifestations of clonal instability in a proportion of cases, presumably as a result of the initial damage to All cytotoxic drugs used in the treatment the marrow. Those with acute leukaemia must be malignant disease, whether phase-specific ( distinguished from patients (usually elderly) who have example cytosine arabinoside and methotrexate) preleukaemia or smouldering leukaemia but in whom cycle-active (alkylating agents and nitrosourea the marrow is also hypoplastic from the outset (Milner damage the rapidly dividing cellsin the marrow tran and Gary, 1979). compartment. However, resting pluripotent stem ct No classification of aplastic anaemia is completely are largely unscathed by these drugs, so that recovc satisfactory because the way in which bone marrow occurs rapidly unless the patient has an intrin damage occurs is often uncertain, but that shown in marrow disorder such as leukaemia. Nevertheless Table 2 attempts a correlation of the known causes can be shown in both animals and man that repeal with possible pathophysiological mechanisms. Drugs doses of cycle-active drugs (such as nitrogen must: and chemicals are the agents most commonly and its analogues) eventually produce depletion of - incriminated; although in most series 50 per cent of stem cell compartment and perhaps also damage cases are labelled idiopathic, it is likely that many of microenvironment. This may sometimes be of clini these are due to exposure to environmental agents, importance. such as small doses of benzene, or antibiotics in food, This type of inevitable, dose-related, and usu; which may be difficult to identify. reversible marrow depression following the use cytotoxic agents must be sharply distinguished frc Table 2. Classificution of aplastic anaemia (data fmm Benestad. 1974) that occurring as a result of exposure to drugs chemicals that are noninjurious or slightly toxic lonizine radiation Cytoto& chemicals dose-deucndent the majority of individuals. The fader characteristically unpredictable, not related to do . -Chemicals and drugs (conditional or immunoloyical) largely dose-independent and often irreversible; it appears to be related tc single episode of cataclysmic damage. Although u d Viral cytotoxicity (direct or immunologicnl) Autotmmunecytotoxicily leukaemia aplastic anaemia is not a progresz disorder, it appears that neither the microenvironm nor the damaged stem cells can be repairrd. The rar Hereditary defect of this type of damage is illusuated chloramphenicol with which it occurs in only ab Drug mechanisms: Only a few agents such as one in 30000 people taking the drug. while H benzene, cytotoxic drugs, and X-irradiation have been .Butamlidin the incidence may be as low as one clearly established as causes of marrow aplasia in 100 OOO (Benestad. 1979). experimental animals. In clinical practice the evidence Drug reactions producing marrow d m a g e IT of drug or chemical toxicity is usually circumstantial thus be dose-dependent. conditional (or idiosyxrat and there are no laboratory tests to prove a cause and or immunological. Conditional mechanisms im effect relationship. Many patients are subjected to that either the target cells or drug metabolism multiple exposures of several potentially toxic agents abnormal; such a defect is most likely to be heredit and a reliable exposure history is often difficult to and may render marrow cells susceptible to damage obtain. the drug or its metabolites. Although these conditio The incidence of drug-induced marrow aplasia reactions are unexpected, increased doses usually clearly depends on the properties of the drugs as well more marked effectswhich are closely related to int as the frequency with which they are used. Table 3 of the drug. In reactions having .immunolog shows the drugs most commonly incriminated in causes, however, the dose and time relationships aplastic anaemia, but it must be remembered that as much more erratic. Morley et a1 (1976)suggested t fashions in drugs change 50 do the patterns of the apparently unpredictable and capricious ONG exposure that may produce marrow damage. aplasia after the use of chloramphenicol, and poss Chloramphenicol heads the list of kno\vn causes in other agents that are normally only slightly rnyelotc most reviews but the incidence of cases due to this micht be exdicable if the rnarrow stem 400 * (lorazepam) * P (oxazepam) ensure excellent relief of anxiety Ativan for the majority of your anxious patients Serenid-D for arlier insult which was unrecognized. It is also -mssible that a toxic agent kills only sensitive stem cells some resistant cells survive but can only replicate more slowly. If there were a genetic predisposition to aplasia, the distinction between idiosyncratic and Lmmune mechanisms would not be easier because the immune response to specific antigens is also influenced by genetic factors. Fig. I shows the stage in the haemopoietic cell system at which some of the dr:.: known to cause aplasia in man may act. It has tc -. admitted. however. that the evidence is often fragmentary and in some cases even conflicting (Benestad. 1979). Some drugs that produce irreversible hypoplasia by damaging the stem cell compartment can also inhibit the transit compartment. presumably by a different mechanism. Thus in many individuals chloramphenicol also produces a temporary dose-related erythroid depression and sometimes a reversible par cytopenia, probably because the drug inhibits prate:synthesis in mitochondria (Yunis, 1973). Ditantin causes reversible red cell aplasia in certain individuals as well as an idiosyncratic marrow hypoplasia. In one patient studied by Yunis et al (1969) the former syndrome was found to be due to an exaggerated depression of DNA synthesis in erythroid cells by the drug. Gold may also produce marrow damage by different mechanisms. In some patients a progressive fall in blood leukocyte and platelet counts precedes marro: failure and this seems to be a dose-related effect (Kay. 1973). On the other hand, in a small group of individuals marrow aplasia occurs without warning after very small doses of gold salts (McCarty et al, 1962). Selectivethrombocytopenia, agranulocytosis, and even red cell aplasia (Reid and Patterson, 1977)may occur as a result of damage to the transit compartment. Since gold persists for long periods in the body, these potentially reversible syndromes may remain after administration of the metal has ceased. The bone marrow hypoplasia caused by chlorpromazine is fairly common in psychiatric patients (Pisciotta, 1971). It is a dose-dependent but unpredictable disorder, probably reflecting a constitutional defect in marrow cells which renders them F7g.i. Possible pathogenetic mechanisms involved in damage to bone marrow cells by certain a h g s (@-omBenestad. 1979). Predictable patientresponse and quick relief ot symptoms . Little or no sedation or other unwanted effects Little risk of accumulation Full prescribinginformationavailable on request Wyeth Laboretorier, John Wyeth L BrotherLimited,Taplow. Maidenhead, Berks. At 24 wade marks I ielsllos . ate 1 to md the ced the ugs be ten ing by ibit ent .mted anein Itin ais me ier led the :nt in OW aY. \viter 12). fen :ur nt. se ter ornts ,ut onem - 3 4 3 2 3 9 ,.`S-A 9 2 -- `401 unable to compensate for the depressing effect that marrow lymphocytosis and plasmacytosis &en this drug exerts on many tissues. It is usually rapidly patients with aplastic anaemia resembles a &on reversible and the likelihood is that the marrow transit inflammatory disorder, possibly of an autoimmur compartment, rather than the pluripofent stem cell, is nature, and this may be triggered by unusual reactiol affected. Other drugs that may act in this way include to drugs such as chloramphenimi. On the other hanc thiouracil and methicillin. aplastic anaemia is rarely seen in the clinical setting Benzene, which can be shown in rabbits to produce disorders known to have an autoimmune basis such . a dose-dependent pancytopenia with aplastic and disseminated lupus erythematosus or Sjiigren hyperplastic manuw pictures and with disorderly syndrome, whereas red cell aplasia and immur maturation (Speck, 1975), is still used as a solvent in granulocytopenia are commonly found in such glue, cement, grease and insecticides, for example, context. Moreover with the exception of posthepatit P and as a dduent for paint, varnish, and lacquer. It is aplasia and those rare cases associated with thymom interesting that toluene and xylol are free from autoantibodies to blood cells are not seen in aplast myelotoxic activity. In an animal model the anaemia. However, it has been suggested th. pluripotent stem cells (CFU-S) were unaffected by autoantibodies in aplastic anaemia might be dkectc exposure to benzene and it is therefore possible that against a "private" antigen expressed only on fl clinical aplasia in man is the result of an idiosyncratic haemopoietic stem cell and not on its diffmntiatc reaction. The facts that chromosomal abnormalities progeny (Benestad, 1974). Alternatively they may 1 have been seen in marrow cells after exposure to directed against vascular cells in the marrow. benzene and that it is undoubtedly leukaemogenic An autoimmune basis for some cases of aplas suggest a direct action on haemopoieticstem cells. received support from the recent work of Ascensao Virus-induced aplasia: Although it is possible that al (1976) who showed that some patients hi viral damage to the marrow accounts for a significant lymphocytes in m m w or blood that appeared proportion of cases at present labelled idiopathic, the suppress CFU-C or CFU-E growth from norm only well defined viral aplasia syndrome is that follow- marow. When these lymphocytes were! removed fro ing infectious hepatitis. Transient inhibition of the the patient's marmw colony growth improve marrow transit compartment Seemsto be common, for Although it is not yet certain whether the presence ` example leukopenia is frequently found during the these suppressor cells is related to the cause of tl illness but the association with aplastic anaemia was aplasia or represents an epiphenomenon, perha1 not recognized until 1955 (hrenz and Quaiser, 1955). following blood transfusions, it is worth recalling th In one series of 193 patients with aplastic anaemia thymic lymphocytes are essential cooperator cells f following hepatitis 18 per cent had received chloram- establishing erythropoiesisin mice (see above). phenicol (Hagler et al, 1975)and it has been suggested The most compelling evidence for disturbt that in some patients the aplasia is really due to immunity as a cause of aplastic anaemia derives fro chloramphenicol or other drugs, impaired the response of both drugrelated and idiopathic cas detoxification of which might predispose to m m w to immunosuppressive therapy. Sometimes chem damage. However. aplastic anaemia is not common in therapy has been given to prepare the patient for : individuals with chronic liver disease and there are allogeneic m a m w graft when, unexpectedly,regrow now sufficient recorded cases of aplastic anaemia of his own m m w has been observed. Although 01 resulting from hepatitis, with no known drug cannot exclude a transient "permissive" role of t! exposures, for a typical clinical picture to have transfused marrow which might supply cells emerged. humoral factors essential for haemopoiesis, -such + The age distribution is similar to that of explanation would not be valid in cases whe uncomplicated infectious hepatitis and men are remission has occurred after the use of antilymphoc! more commonly affected than women. The onset of serum alone (Speck et al, 1978). A particular aplasia is usually within 10 weeks of the hepatitis, instructive case of this type was that described by t: which is often mild. However. the aplastic anaemia is Royal Marsden Hospital Bone M m w Trvl particularly severe, especially in women: in one series plantation Team (1977). A 15-year-old girl develop the mortality rate was 88 per cent and the median posthepatitis aplasia and rejected bone marrow mar survival only 10 weeks (Ajlouni and Doeblin. 1974). plants from her identical twin. but after cyclophc Patients with posthepatitis aplastic anaemia thus fall phamide treatment a third graft was succe~sfulE. kfc into that group with severe disease in whom bone immunosuppressive therapy the patient's seru marrow transplantation should be considered at an inhibited CFU-C growth in the tuin's mmow. The early stage. Williams et al (1973) suggested that was also evidence of a cellular inhibitor that w hepatitis could cause aplasia by several mechanisms: removed by coculture uith mtith:.moqie globulin. direct viral damage to the mama. damage to the Constitutionalaplasticanaemia: Thiscategoryinclud microcirculation. an autoimmune response to the virus patients uith congenital. genetic, or familial dise: (see below). or failure by the liver to eliminate endo- who have an inherent disposition to bone marrc genous or exognous bone marrow toxins. failure. Sometimes the development of marrow hyF Autoimmunity: Immune mechanisms involving both plasia appears to follow exposure to toxic extrin: humoral antibodies and probably cell-mediated agents (already noted) SO that patients with toxicity can injure proliferating haemopoietic cells in unrecognized propensity for aplastic anaemia wiU the marrow. Damage may be restricted to a single cell listed wrongly as having acquired disease. Howevl type or involve several simultaneously. The best the constitutional nature of the disease is more eas recognized clinical syndrome of this type is pure red diagnosed when the child shows other stigmas. cell aplasia in which autoantibodies to erythroid Thus in Fmconi's anaemia, which is the best defin precursor cellsor erythropoietin are sometimespresent, of these clinical syndromes and is inherited in an aul but autoantibodies to CFU-C as well as more differ- somal recessive pattern, the child is often of shc entiated myeloid precursor cells have been demon- stature and may show skin hyperpigmentatic -..-_strated in certain granulocytopenic patients (for 3 skeletal or rend defects. and sometimes men comnrr:ipnciIp y p L - i p \ ~ -c - dins- qnr4 C , \ l r l ~ 107Rl T h . 2 ..-*--,a C . : A - - . I . . .L- I.-- f1 ' 402 ' one manifestation of a generalized cellular defect. Most of these children show chromosomal abnormalities (cytogenetic analysis is an essential investigation in any child with bone marrow failure) and it has been Dosik. H. Skier. W. Lubiniecki (1979) British Journal Of Haernarology. 41. 77 Ehrlich. P(1888) (7harire-Annalen. 13,300 Freedman, M H.Saundets. E F (1975)in Abstracts of the 18th Annual Meeting of the American Societyof Haematology. Dallas. USA postulated that the underlying biochemical lesion is a defect in DNA repair mechanisms. It is possible that some patients with constitutional marrow failure-who do not show the other nonhaematological features of Geary. C G. Testa. N G (1979)in Aplastic Anaemia (edited by Gcary, C G).Bailli6reTiidall. London. p.1 Greenberg. P L. Schrier. S L(1973)Blood.41.753 Hagler, L. Pastore. R A. Bergin. J J (1975)Medicine (Balrimore). 54, 139 Fanconi's syndrome represent formes frustes. In a recent review of 40 patients with constitutional aplastic anaemia diagnosed at the Children's Hospital Medical Center in Boston between 1958-77,26 had the classic featups of Fanconi's anaemia while 10 had bone marrow failure Without other stigmas and four had amegakarpcytic thrombocytopenia which eventually Hansi. W, Rich, I. Heimpel. H. Heit. W. Kubanek. B (1977)Brituh Journalof Haematology. 37.483 Heimpel. H (1979) in Aplastic Anaemia (edited by Geary. C G). -.Bailliirc Tindall. London. p. 65 Kubanek. B (1975)Britlsh Journal of Haematolo~v3. 1, Suppl. 57 J e d t q a a k . W W.Sharkis. s. Ahmed. A. Sell. K W (1977)S c k n n . 196, 313 Johnson. C R. Metcdf. D (1977)hoceedinj?sof the NationaIAcademy of Scunces of :he USA. 74,3879 evolved into panhpplasia (Alter et al. 1978). The incidence of leukaemia in these patients is much Kansu. E. Erslcv. A J (1976)Scandinavian Journal of Haemarology. 17.326 Kay. A (1973)AnnaLcofrhe Rheumatic Diseases, 33.277 higher than in other forms of aplastic anaemia and an Kern. P. Heimpel. H. Heit. W.Kubanek. B (1977)British Journal of interesting feature held to be diagnostic of Fanconi's anaemia is the increased expression of the S V 4 T antigen when the skin fibroblasts from these children are incubated uith thisoncogenicvirus in culture. This feature also occurs in first-degree relatives who do not show marrow failure(Dosik et al, 1979). The defect in the marrow of children with Fanconi's -anaemia is at the level of the pluripotent stem ceU Haemamloo. U.613 Knospe. W H. Crorby. W H (1971)Loncer.i. 20 Lajtha. L G (1978)in Proceedings ofthe 17th Congress of the Inter- national Society of Hatmatology. Pans. p.286 -.Lord, B I . Testa, N G. Hendry. J H (1975)Blood. 46,65 Mori, K J. Wright. E G (1977)Biomedicine. 2l. 223 Lorenz. E. Quaker. K (1955) Wiener medizinische Wochenschrifi, 105. 19 McCarty. D 1. Brill. J M, Harmp. D (1962)Journal of !he American Medical Association. 179.655 both CFU-C and CFU-E are decreased (Saunders et Milner. G R. Geary. C G (1979)in Aplastic Anaemia (edited by Gury. al, 1974; Freedman and Saunders, 1975). By contrast, in congenital red cell aplasia (Diamond-Blackfan C G). Baillidre Tiidall. London. p. 230 MU. M A. G a y . C G. Delarnorc. 1 W (1977)ScandinavianJournalof Haemardogy. 19.225 -syndrome)the defectis at the level of the BFU-E which Morky, A. Holmes. K, Forks. I (1974)Austrdian and New Zealand appears to be inhibited by abnormal T lymphocytes this may be a disorder of cellular immunity (Steinberg Journal of Medicine. 4.538 -. Trainor. K, Remes. J (1976)British Journal of Haematdogy. 32. 525 et al, 1979). Psciona. A V (1971)CIinicalPhannacology and Therapeutics. 12, 13 Reid, G. Patterson. A C (1977)British MedicalJournd. ii. 1457 Summary PA la& anaemia is thus a syndrome with a number o different causes. The fact that bone marrow transplantation restores haemopoietic function in a substantial number of cases (although there are still formidable immunological problems to be overcome) Royal Marsden Hospital Bone Marrow Transplantation Team (1977) Lancet, ii. 742 Russell. E S. Banstein. S E (1966) in Biology of the Labdratory Mouse, 2nd edn (edited by Green. E L). McGraw HiU. New York. p. 351 Saunders. E F.Amato. D.Freedman, M H (1974)Blood. 44.913 Simonovitch. L. McCuloch. E A. Til. S E (1963)Journal cellular and Comparotiw Physiolw. 62.327 suggests that the disease is indeed often due to defec- Smith, L W (1919)AmericanJournalofDLeasesofChildren. 17. 174 -.tive or deficient stem cells. Howeverthere is evidence to suggest that in some cases at least, stem cells are Speck. B (1975)Haemat.Blurtransfusion, 16. 235 Gluckman. E. Hark. H L. van Rood, J J (1978) CIinics in Haemardogy. 7,611 L- inhibited rather than completely destroyed. It must not be forgotten that only one in four patients has a. suitable prospective donor, so the identification of a group in whom, for exampie, growth of haemopoietic Steinberg. M H. Coleman, M F, Pennebaker, J B (1979) British lournal 4Haemarologv. 41.57 Stohlman. F (1972)Blood. 40.282 Til. J E. McCulioch, E A (1961)Radiation Research, 14, 213 Viter. R W, Will, J J. Janold. T (1967)Seminars in Haemarology, 4. cells was depressed by an autoimmune process would have important therapeutic implications. Although our understanding of the pathogenesis of this mysterious disease has advanced since 1919, when 175 Williams. D M. Lynch, R E , Cartmight, G E (1973)ibid. 10. 195 -Yunis. A A (1969)Advances in Internal Medicine, 15,357 (1973)Seminanin Haematology. 10.225 Smith concludedthat the aetiology of aplastic anaemia was "nothing but speculation", much still remains to be learnt. - l r i g ~ m1 u rcptvdvad by kind permisston of the publishers Baifliire Thdd Ajouni. K, Dccblin. T D (1974)British Journal of Hacmarofvgy, 27. 345 Allen. fD. D u t a . T M (1976)Differentiation. 6 , 191 Alter. B P. Potta. N U. Li. F P (1978)CIinicsin Haemardogy. 7.431 Arensao. J. Kagan. W, Moore. M. Pahwa. R, Hanxn. K. Good. R (1976)bncct. i. 669 -Benestad. H B (1974)AcramedicaScandinavica. 196.255 (1979) in Aplastic Anaemia (edited by G&y; C G). Baillike T m d d , London. p.26 Bomford. R R, Rhoads, C P (1941)Quam+ Journal ofMedicine. 10. 175 Bottiger. LE. Wertaholrn. B (1973)EritW MedicalJwmd. iii. 339 Chauffud. M (1%) Bulletin et mimories de la SocidtC Mddicale des H6pitawdc Paris. 21.313 Cline. M 1, Golde. D W (1978)Americ.n Journal ofMcdicine. 64.501 Dena. 7 M. Spooom. 1. Heorkey, J H, Ljthe. L G (1978) in Hacmopoictic Cell Dillerentiation (edlted by Gdde. D W , Cline, M I). Acadcmr h,New York. p.163 .i