The single centre experience with allogeneic bone marrow transplantation over 25 years is reviewed. 473 allogeneic stem cell transplants have been performed since 1973. Sixty-eight (14%) were done between 1973-80, 216 (46%) between 1981 and 1990 and 189 (40%) between 1991 and 1998. 437 were first trans plants, 36 were second or subsequent transplants. Bone marrow stem cells were given in 421 cases (89%), and peripheral blood stem cells have been used since 1994 in 52 (11%) transplants. Of 437 patients, 101 (23%) had acute lymphoblastic leukaemia, 131 (30%) acute myeloid leukaemia, 101 (23%) chronic myeloid leukaemia, 24 (5%) a myelodysplastic or myeloproliferative disorder, 23 (5%) lymphoproliferative disease and 53 (12%) severe aplastic anaemia. Of patients with leukaemia, 187 (52%) were transplanted at an early disease stage, i.e. first complete remission or first chronic phase, and 48% with advanced disease. Median age was 27 years (range: 1 to 58), 55% were male. Pretransplant conditioning was with cyclophosphamide (Cy) and total body irradiation (TBI) in 192 patients (41%), Cy + TBI + VP-16 in 182 (39%), Cy alone in 53 (11%), busulfan + Cy in 24 (5%) and other regimens in 22 (5%). Graft-versus-host disease prophylaxis was with methotrexate (MTX) in 41 cases (9%) in the seventies, in the eighties and nineties cyclosporine (CSA) was used in 202 patients (43%) or CSA + MTX in 161 (34%). 51 (11%) received T-cell depleted transplants and 18 were transplanted without GVHD prophylaxis. Donors were HLA-identical siblings for 405 transplants (86%), identical twins for 18 (4%), volunteer unrelated donors for 32 (7%) and HLA-mismatched relatives for 18 (4%). Changes over time have affected not only transplantation practices but also patient selection. 31% of leukaemia patients in the seventies but 58% in the eighties and 55% in the nineties were transplanted in an early disease stage (p <0.01). The 15-year survival probability for all patients is 38 +/- 6%. The 5-year survival probability increased from 35 +/- 12% in the seventies to 46 +/- 7% in the eighties and 59 +/- 8% in the nineties (p <0.006). In the last 25 years, allogeneic stem cell transplantation has become accepted treatment for a number of malignant and non-malignant haematological diseases. Over a third of transplant patients become long-term survivors. Transplant outcome is improving probably due to patient selection and improved transplant technology.

Fifty-three patients with severe aplastic anaemia were admitted to this hospital between January 1976 and June 1980, of whom three arrived in terminal condition and died before treatment for their basic disease could be given. Thus 50 patients were treated and evaluated in a prospective study according to one protocol. Eighteen patients with an HLA-identical sibling underwent bone-marrow transplantation with the aim of achieving haematopoietic chimerism. Thirty-two patients without an HLA-identical sibling were given antilymphocyte globulin with or without an infusion of HLA-haplotype-identical marrow. All these 32 patients received low-dose androgens after the procedure. In the first group eight patients (44%) survived. In the two other groups, 22 patients survived (69%), of whom 20 were completely self-sustaining (63%). Engraftment and graft-versus-host disease did not occur in the group who received antilymphocyte globulin and haploidentical marrow, and the haematopoietic reconstitutions in these patients were all autologous. These results confirm the efficacy of antilymphocyte globulin in the treatment of severe aplastic anaemia and show that such treatment is at least as good as bone-marrow transplantation. Its mechanism of action remains unknown, but most patients with aplastic anaemia have a pool of haematopoietic stem cells able to repopulate the marrow after this type of treatment.

Experience with a total of 49 plasma exchanges in 14 patients is reported. All exchanges were performed with the Aminco cellseparator in an attempt to remove documented or suspected antibodies completely. Indications were ABO-barrier before bone marrow-transplantation (6), acute rejection following renal transplantation (3), autoimmune diseases (3), Moschowitz disease, antithrombocytic antibodies and hepatic coma (1 each). The technique was modified to make it possible routinely to exchange 121 of plasma in 4 hours with no major untoward side effects even in anuric or thrombopenic patients. Complete removal of circulating antibodies was possible and often followed by impressive clinical benefit.

Summary In 12 patients with severe aplastic anaemia who had achieved self‐sustaining autologous bone marrow function after treatment with antilymphocyte globulin, or with cyclophosphamide given for attempted bone marrow transplantation, colony formation by all haemopoietic precursors remained far below normal. Precursors from peripheral blood, erythroid precursors in particular, failed to form a normal number of colonies. This paucity of colony formation does not reflect a true lack of precursor cells but is due to circulating cells which impair maturation. Addition of low density peripheral blood cells to autologous bone marrow cultures diminished colony formation by granulocyte‐macrophage precursors (GM‐CFC) and abolished ‘burst’ formation by BFU‐E. Strong auto‐inhibition preceded relapse in four of eight patients. The phenomenon was not observed in five normals, in five aplastic anaemia patients with stable haemopoietic grafts and in three polytransfused control patients.The T‐cell poor subpopulation of peripheral blood cells, containing mainly B‐cells and macrophages, was especially inhibitory. Accordingly, removal of plastic adherent cells from bone marrow cell suspensions improved plating efficiency in aplastic anaemia patients, but not in normals. Isolated E‐rosette positive cells had no negative effect.Inhibition could only be demonstrated in the autologous situation. Colony formation by normal allogeneic peripheral blood precursors was not impaired by patient cells. The phenomenon is likely to reflect residual disease activity which is compensated in vivo but can be demonstrated in vitro. It may be of help in early recognition of patients who are at risk of relapse after autologous bone marrow reconstitution.

Non-clonal growth of macroscopic cell aggregates in methylcellulose cultures of abnormal marrow is described.

Experience with 17 allogeneic bone marrow transplantation (BMT) patients with refractory acute leukemia is reported. The authors show that it is extremely difficult to cure patients with end stage disease by this procedure. Based on experience that, if performed early, BMT can bring about over 50% potential cures, it is postulated that marrow grafts should be performed much earlier than is done at present. From the results in 45 patients with severe aplastic anemia it is concluded that the majority do not have a defect at the pluripotent stem cell itself. There is disturbed maturation to functional end cells which can be corrected in most instances by ALG, infusion of allogeneic marrow and low dose androgens. BMT between HLA-identical siblings need not cause major problems, but it involves a considerable risk of fatal complications such as irreversible marrow rejection and GvH.

We are reporting our experience with 13 patients suffering from end stage acute leukemia that were prepared for allogeneic bone marrow transplantation by combined chemotherapy followed by high dose cyclophosphamide (Cy) and total body irradiation (TBI). Only one patient became a long term survivor. Of the evaluable 12 patients, 6 died of interstitial pneumonia, 4 of GvH and 1 of recurrent leukemia. We conclude that adding combined chemotherapy to the standard conditioning program with Cy and TBI probably increases the risk of developing fatal interstitial pneumonia without eliminating the risk of recurrent leukemia. We suggest that allogenic marrow grafts should be performed earlier in the course of refractory acute leukemias, because in patients with end stage disease its chances of being curative are small.

18 patients with severe aplastic anemia (SAA) but without an HLA-identical sibling were treated by antilymphocyte globulin (ALG) followed by infusion of marrow cells from a semi-compatible family donor. 13 of these received low dose androgens after ALG: 11 (85%) achieved stable remission without transfusion requirement. One patient relapsed after 4 months, one patient with only partial remission died from infection. None of the 4 patients who did not receive androgens after ALG achieved remission. ALG, marrow and low-dose androgens represent a promising therapy for SAA and can be favorably compared with allogeneic bone marrow transplantation.

A 56-year-old patient with severe acquired myasthenia gravis refractory to conventional therapy underwent a series of plasma exchanges. Three single exchanges did not improve her status. A series of intensive plasma exchanges of 6 liters each on 4 consecutive days reduced her anti-acetylcholine receptor antibodies to one-tenth of the initial value. The reduction was accompanied by an impressive clinical improvement. This study confirms that antibodies can be removed by plasma exchange. It further shows that the clinical improvement appears to be closely related to the quantity of antibodies removed.

Immunosuppression with ALG has become a new alternative in the treatment of SAA [19]. In the original reports ALG was followed by a BM infusion from a family donor [12, 15]. In subsequent experimental and clinical trials ALG has been effective with and without BM infusion [16, 17, 19, 20, 21]. This modality is applicable in the HLA-nonidentical setting; it is not followed by GvHD and leads to autologous BM recovery in more than 50% of patients with SAA. It is still an unsolved problem whether ALG followed by BM is more or less effective than ALG alone.

Much progress has been made in allogeneic bone marrow transplantation for severe aplastic anemia (SAA) and acute leukemia (AL). In SAA it was shown that hemopoietic chimerism and apparently permanent cures can be achieved in the majority of patients by conditioning with cyclophosphamide followed by bone marrow transplantation (BMT) from an HLA-identical sibling. The previous transfusion history is crucial for failure or success: untransfused patients do very well while graft rejection is an enormous problem in most polytransfused ones. We have shown that most patients without HLA-identical sibling donors can be adequately helped as well. After conditioning with ALG followed by transfusion of haploidentical marrow and low dose androgens there is partial to complete autologous hemopoietic reconstitution in virtually all patients. This points to the fact that most of these patients have pluripotent hemopoietic stem cells that are intact, but apparently unable to differentiate to mature cells, because they are inhibited by autoimmune mechanisms. The results of BMT in patients with endstage leukemia are modest. New pilotstudies with early marrow grafts, i.e. for ANLL in first remission and for ALL in second remission indicate that with this type of approach potentially over 50% of all patients with HLA-identical siblings can be cured. We recommend that HLA-typing should be performed early in families with SAA and AL and that the possibility of a marrow graft should be seriously considered before the patients have endstage disease. Marrow grafts are technically simple but they may pose enormous problems such as graft versus host reaction (GvH), interstitial pneumonia, graft rejection and leukemic recurrence. Therefore, the procedure should only be performed in highly specialized centers with much knowledge and experience in the immunobiology of bone marrow transplantation.

Peripheral blood precursor cells from all of 12 patients with polycythaemia vera (PV) formed hemoglobinized colonies in vitro without addition os was strictly epo-dependent in 30 normals and 8 patients with erythrocytosis of other origin. By addition of 1 U epo/ml to the cultures, colony formation was increased up to 5-fold in PV. Untreated patients with a short history had the highest numbers of such epo-responsive precursor cells. In 3 patients with concomitant myelofibrosis, erythroid in vitro growth was abundant, but scattered, single colonies could hardly be identified and their hemoglobinization was poor. This picture did not change with addition of epo, and hemoglobinization did not improve. In vitro grown colonies in PV had morphological abnormalities as compared to colonies grown from normal precursor cells: all stages of erythroid differentiation and up to 50% necrotic cells were found within single colonies. Scattered colonies contained an excess of large, immature, vacuolated erythroblasts. It is concluded that these functionaland morphological abnormalities of in-vitro erythropoiesis are a reliable indicator of PV. Further, transition to myelofibrosis is recognized by a characteristic growth pattern with decreased epo-responsiveness.