Using data from centers that are collaborating in the European Group for Blood and Marrow Transplantation (EBMT), we conducted a retrospective, descriptive study on late death, health, and social reintegration of patients who had received allogeneic and syngeneic bone marrow transplants for hematologic disorders before 31 December 1985 and who had survived for at least 5 years. We evaluated variables associated with late illness and death and reduced participation in school or work. The patient's primary disease, age, and sex and the availability of a donor were predetermined variables. However, evaluation of the variables related to the treatment strategy, such as the selection of the optimal time in the course of the disease for transplantation, the choice of the conditioning treatment (that is, with or without radiation), and the regimen used for prophylaxis of graft-versus-host disease, is of particular interest.

Methods

Top Methods Results Discussion Author & Article Info References

All European transplantation centers cooperating in the EBMT were asked to provide information on patients who had transplantation before 31 December 1985 and had survived for at least 5 years after receiving the graft. Data on 798 consecutive transplant recipients were reported from 43 centers in 13 European countries. The mean percentage (±SD) of long-term survivors was 39.2% ± 8.8%. Selective reporting was excluded by scoring unique patient numbers that were previously reported to the EBMT registry in Leiden, the Netherlands. The data were validated at meetings of the EBMT Late Effects Working Party and through personal contact with the responsible persons at the centers. Clinical performance was assessed according to the latest Karnofsky score, and social activity was evaluated according to the patient's ability to attend work or school full-time or part-time. Karnofsky scores were assigned in increments of 10%. Scores of 90% and 100% are compatible with normal activity; a score of 80% reflects special efforts to carry on normal activity; and scores of 70% or less reflect varying need for assistance with normal activity. For patients who died more than 5 years after transplantation, the causes of illness and death were reported. We considered the following factors as potential risk factors for impaired clinical and social performance and late death and entered them into a bivariate analysis: age and sex of the patient and donor; histocompatibility of the donor; primary disease and status at the time of transplantation; conditioning regimen, including details on radiation; method of prophylaxis of graft-versus-host disease; occurrence of acute and chronic graft-versus-host disease; development of secondary cancer; and recurrence of original disease.

Patient Data

Patient characteristics are summarized in Table 1. Patients younger than 18 years of age were classified as children (321 patients), and patients 18 years of age or older were classified as adults (477 patients). Most patients (n = 652) had transplantation for the treatment of leukemia. Early phases of disease were defined as first remission of acute leukemia or the chronic phase of chronic myelogenous leukemia; intermediate phases were acute leukemia in second remission or chronic myelogenous leukemia in the accelerated phase; and advanced phases were later remissions of or active acute leukemia and blastic transformation of chronic myelogenous leukemia. Donors were HLA-identical siblings for 775 patients, HLA-mismatched family members for 7 patients, and syngeneic twins for 16 patients.

A total of 645 patients (80%) received total-body radiation; 434 patients were treated with single-dose total-body irradiation, and 211 patients received fractionated total-body irradiation. Thirty-four patients were conditioned by total lymphoid or thoracoabdominal radiation, and 89 patients were conditioned by chemotherapy only. For total lymphoid irradiation, the radiation field involved an inverted Y for the lower half of the body and a mantle field with shielding of the oropharynx. When thoracoabdominal irradiation was given to the trunk, the head and the limbs remained outside the field. Chemotherapy most frequently consisted of cyclophosphamide, either alone (50 mg/kg of body weight per day for 4 days) or in combination with radiation (60 mg/kg per day for 2 days or 50 mg/kg per day for 4 days). Other chemotherapy consisted of combination regimens, including cytosine-arabinoside (38 patients), melphalan (16 patients), busulfan (8 patients), daunomycin (13 patients), and the nitrosoureas bischlorethylnitrosourea or cyclohexylchlorethylnitrosourea (11 patients). In 30 patients, the type of conditioning procedure was not reported.

The most frequently used methods of prophylaxis of graft-versus-host disease were methotrexate given after transplantation, cyclosporine, or the combination of cyclosporine and methotrexate. Eighty-three patients received prophylaxis with T-cell-depleted bone marrow or antithymocyte globulin.

Data on the occurrence of graft-versus-host disease were reported for 655 recipients of allogeneic bone marrow. Acute graft-versus-host disease of grade II or higher developed in 23% of patients. Limited chronic graft-versus-host disease was seen in 182 patients (28%), and extended chronic graft-versus-host disease was seen in 92 patients (14%). Graft-versus-host disease had not developed in 381 patients (58%).

Statistical Analysis

Age and sex of the patients and donors, diagnosis and stage of the disease at the time of transplantation, conditioning regimen, method of prophylaxis of graft-versus-host disease, occurrence of acute and chronic graft-versus-host disease, development of secondary cancer, and recurrence of the original disease were assessed for their influence on late death, impaired clinical performance (Karnofsky score 80%), and return to social activity (full-time attendance at work or school). In a bivariate analysis, we used the Fisher exact test to compare proportions. Mortality was evaluated by comparing duration of survival using the log-rank test. A P value of 0.05 or less was considered statistically significant. Analyses were done using the Number Cruncher Statistical Systems (NCSS) statistical package (Dr. J.L. Hintze, Kaysville, Utah).

Results

Top Methods Results Discussion Author & Article Info References

Survival and Late Mortality

The median duration of observation was 8.4 years (range, 5 to 19 years). Figure 1 shows the number of patients as a function of time since treatment; the cumulative overall mortality of the observed patients; and the expected mortality of a similar, unselected group of patients. Fifty-five patients died more than 5 years after bone marrow transplantation, and 743 patients are alive and evaluable for clinical performance and social activity. For 5-year survivors, the actuarial risk for death is 8% (95% CI, 6% to 11%) at 10 years (216 patients at risk) and 14% (CI, 8.1% to 19.9%) at 15 years (16 patients at risk).

View larger version (33K): [in this window] [in a new window]   Figure 1. Mortality more than 5 years after bone marrow transplantation. The curve marked by diagonal lines gives the number of observed patients as a function of time since treatment (person-years at risk). The step function with the hatched range of 95% Cls represents the probability of treatment failure before the specified time, with all observed deaths considered. The lower curve gives the expected mortality according to mortality rates of an age-adjusted normal population in western Europe. The mortality rate at 17 years was 21.6%; six patients were at risk at 17 years. The longest duration of observation was 19 years.

The causes of death occurring more than 5 years after bone marrow transplantation are listed in Table 2. Leukemic relapse was the most common single cause of death and most frequently occurred in patients with chronic myelogenous leukemia (14%) and lymphoma (25%); it was also seen in 4% of patients with acute myelogenous leukemia and 4% of patients with acute lymphoblastic leukemia. Chronic graft-versus-host disease with and without infection and chronic lung disease were the second most frequent causes of death. Secondary cancer and transfusion-associated acquired immunodeficiency syndrome (AIDS) also contributed to mortality.

The most important risk factor for survival after more than 5 years is recurrence of leukemia and lymphoma. Other statistically significant factors in the bivariate analysis were the development of secondary cancer, chronic graft-versus-host disease, stage of disease at the time of transplantation, and the use of radiation for conditioning treatment (Table 3). After patients with recurrent disease were excluded, extended chronic graft-versus-host disease, the development of secondary cancer, female sex of the donor, male sex of the patient, and the use of methotrexate were statistically significantly associated with late mortality (Table 3).

Clinical Performance

Information on Karnofsky scores was available for 647 patients. Patients with a Karnofsky score of 80% or less are considered to be unable to perform normally without special effort.

Factors associated with clinical performance reduced to less than 100% were reported for 125 patients. After 15 patients who were living with recurrent disease were excluded, the major cause of illness was chronic graft-versus-host disease (Table 4). Pulmonary changes consisted of obliterative bronchiolitis, lung fibrosis, and recurrent infections, which were probably the sequelae of chronic graft-versus-host disease. Similarly, aseptic osteonecrosis and osteoporosis were associated with chronic graft-versus-host disease and its treatment with corticosteroids.

Neuropsychological changes included seizures, depression, and leukoencephalopathy. Cataracts contribute to illness, but surgery can restore clinical performance. In one patient, the human immunodeficiency virus (HIV) was transmitted by blood transfusions. Five patients had preexisting conditions, such as paraplegia, diabetes, and injuries, that caused reduced performance status. In the bivariate analysis (Table 5), chronic graft-versus-host disease and recurrent leukemia were significant risk factors for reduced performance status.

Social Activity

Attendance at school and work correlated well with clinical performance status (Table 6). Most transplant recipients (89%) have returned to full-time work or school; only 6% do not go to work or school. Factors associated with absence from work and school are patient age and sex, diagnosis of chronic myelogenous leukemia, chronic graft-versus-host disease, occurrence of secondary cancer, and recurrence of leukemia (Table 5). Fewer patients with extended chronic graft-versus-host disease than patients with limited disease have returned to work or school. Age at the time of transplantation also affected social activity. The adults in our study engage in full-time activities less frequently than do the children, and the female patients engage in full-time activities less frequently than do the male patients. Patients with secondary cancer are less likely to return to work or school than are patients without secondary cancer.

Discussion

Top Methods Results Discussion Author & Article Info References

The number of patients having bone marrow transplantation has increased considerably in the past 20 years [1, 2]. In 1973, 16 bone marrow transplantations were done in Europe; the number increased to more than 7700 in 1993 and to more than 10 400 in 1994. The survival of patients with acute complications has improved, and an increasing proportion of patients survive more than 5 years after transplantation. It is important to assess the risk for late death and chronic illness in these patients and to identify risk factors. We have done a retrospective study in European centers that did bone marrow transplantation more than 10 years ago. Only patients with hematologic diseases who received allogeneic bone marrow grafts were included in our study. To ensure homogeneity, patients with solid tumors and patients receiving autografts were excluded. Forty-three centers cooperated in the study, and data on 798 patients who survived 5 years were reported. The original source of data was the central registry of the EBMT in Leiden, the Netherlands, where reports on consecutive transplant recipients were collected until 1988. Further information was obtained from the centers by direct contact. We sent out questionnaires that solicited detailed information on conditioning regimens, including radiation; type of prophylaxis of graft-versus-host disease; acute and chronic graft-versus-host disease; secondary cancer; survival; clinical performance; and social activity. We also determined causes of illness and death. Data were discussed with representatives of the centers at regular meetings of the EBMT Late Effects Working Party and were validated by several checkbacks on computer printouts for correction. Inconsistencies were clarified by direct telephone contact with the representative of the center. Because members of the study group were deeply interested in this matter, selective reporting is very unlikely.

Our principal finding is that most patients who survived more than 5 years after bone marrow transplantation are in good health and are socially reintegrated. The clinical performance status of more than 93% of patients is compatible with normal activity (Karnofsky scores of 90% to 100%). More than 89% of patients have returned to full-time work or school (Table 6). This result is encouraging for patients with hematologic diseases, and it supports expectations for cure in many patients who have survived 5 years after transplantation. Several studies had smaller patient cohorts and shorter observation times [6, 7-9]. In these studies, the proportion of patients with Karnofsky scores of 90% and 100% was smaller, but the median observation time was only 2 to 4 years. Improvement in quality of life during the first 3 years after transplantation was documented in one study [9].

However, the risk for illness and death remains increased beyond 5 years after transplantation (Figure 1). The major causes of death are late recurrence of malignant disease [10], development of secondary cancer, and chronic graft-versus-host disease. The risk for recurrence of the original disease is particularly high in patients with chronic myelogenous leukemia and lymphoma (actual risk, 14% to 25%), but disease also recurs in patients with acute leukemia (actual risk, 3% to 4%). The risk for recurrence of leukemia is higher in patients who have transplantation at a more advanced stage of disease (P = 0.04) and in patients given T-cell-depleted transplants (P = 0.001) (Table 3). In an analysis of persons who survived 2 years, the International Bone Marrow Transplant Registry (IBMTR) found an actuarial risk for relapse at 5 years of 5% to 13%, depending on the form of leukemia and the stage of disease at the time of transplantation [11]. However, little is known about the risk for relapse after more than 5 years.

Another cause of late death in bone marrow transplant recipients is secondary cancer [12]. Fortunately, most secondary malignant conditions are cancers of the skin that can be cured by surgical excision of the lesions. Brain tumors and carcinoma of the oropharynx and esophagus occur less frequently but are life-threatening (Table 2). The incidence of glioblastomas is increased in children who receive cranial radiation as prophylaxis of leukemic meningosis [13]. The risk factors for secondary cancer must still be defined. Ionizing radiation is known to induce cancer in animals and humans. In our group of patients, an association of secondary malignant conditions with the use of radiation for conditioning treatment could not be shown, but deaths from secondary cancer were only seen in eight patients conditioned with radiation. Radiation has been reported as a risk factor for secondary cancer in patients with severe aplastic anemia [14], but such factors as chronic graft-versus-host disease and its treatment may also be important. Immunosuppressive treatment is a known risk factor for secondary cancer in organ transplant recipients. Other investigators [11, 12, 15, 16] have reported that patients who receive bone marrow transplants and are treated with antithymocyte globulin and T-cell-depleted transplants have a high incidence of lymphomas. Many lymphomas are associated with Epstein-Barr virus infection in donor cells and develop soon after transplantation. We did not observe any lymphomas in our study because we excluded patients who survived less than 5 years after transplantation.

In patients whose late death was caused by AIDS, the syndrome resulted from transfusions with HIV-contaminated blood. Infection with HIV had occurred at a time when current screening tests for HIV antibodies in blood donors were not available. Recurrent infections and obstructive and restrictive lung diseases are common, often life-threatening complications of chronic graft-versus-host disease. In some patients, fatal infections and lung disease occur even without other signs of chronic graft-versus-host disease.

A morbid condition has been defined as an illness that distinctly reduces clinical performance status (Karnofsky scores 80%). Social reintegration was assessed by asking whether the patient was "back to work or school full time, part time or no time." Information on the Karnofsky score and attendance at work or school was available from 562 patients. In accordance with health status, 491 patients with a Karnofsky score of 90% or 100% had returned to full-time work or school; 28 patients with a Karnofsky score of 80% or less had not returned to full-time work or school. Significantly fewer older patients and female patients attended work or school after transplantation (Table 5). Despite having Karnofsky scores of 90% or 100%, 25 women and 7 men had not returned to full-time work. Conversely, of 9 adults with a Karnofsky score of 80%, 7 men and only 2 women worked full time. One patient in each performance group was still a child at the time of evaluation. These differences may simply reflect problems other than health status. Although the question we asked concerned the resumption of previous activities at work or at home before transplantation, it may have been misinterpreted as a question about employment. Evaluation of the proportion of employed patients as criteria of successful treatment may be difficult in multicenter studies because employment rates differ among countries and among cultural and economic backgrounds. Such variations may be most prominent in women. Wingard and colleagues [6] evaluated employment before and after transplantation and found that loss of employment was associated with female sex, chronic graft-versus-host disease, and lower socioeconomic status.

Particular problems may be presented by patients who went through puberty after transplantation. One hundred twenty-five boys and 99 girls had transplantation as children and were adults at the time of the last evaluation. In these patients, no difference was seen between men and women with regard to clinical performance and ability to attend work or school. Several problems associated with puberty may not, however, be reflected by the Karnofsky score or assessment of social activities. Because of primary ovary failure, most girls conditioned with total-body radiation need supplementation of sex hormones in order to mature at puberty [17, 18]. In most boys, testicular failure is limited to aspermia and testosterone production is not irreversibly impaired [19]. Repeated courses of cranial radiation, as used to treat recurrent lymphoblastic leukemia, resulted in impaired cognitive function that may be greater in girls than in boys [20].

There is little doubt that growth retardation and disturbances of gonadal development are sequelae of intensive conditioning treatment, including radiation [18, 21]. Infertility, endocrinologic disturbances, and cataracts are commonly associated with radiation in adults [18, 22]. However, most patients with acute leukemia and lymphoma receive intensive therapy before conditioning treatment and transplantation. Proper evaluation of the cause of late effects therefore requires comparison with a group of patients who have the same diagnosis and have not had transplantation. In a study that compared children who had acute myelogenous leukemia and received intensive chemotherapy alone with children who received radiation and had transplantation, the transplantation group showed the well-known late effects of radiation. However, the chemotherapy group also showed late sequelae [21]. Disturbances of growth and development are evident, but they may be corrected by hormonal supplementation. Cataracts may be removed, but infertility may be a problem in certain phases of life.

Our evaluation does not consider psychosocial factors associated with bone marrow transplantation. These factors can only be evaluated with special questionnaires [3-6, 23] that give patients the opportunity to score their own performance. A great deal of quality of life may be influenced by self-esteem and the continued role of the patient in a family or partnership. Special psychological attention may be necessary for improvement of psychosocial activity. Our study did not include questions on self-assessment of quality of life; more detailed questionnaires and well-defined controls are necessary to assess quality of life.

We limited our evaluation to medical problems assessed by physicians at the transplantation centers. Extended chronic graft-versus-host disease is the most significant risk factor for reduced clinical performance and social activities. Chronic graft-versus-host disease is the major cause of late complications that ultimately determine the success of bone marrow transplantation [24-26]. Chronic pulmonary disease and recurrent infections are sequelae of extensive chronic graft-versus-host disease. Avascular necrosis and osteoporosis are known complications of allogeneic bone marrow transplantation. In a previous study of patients having this type of transplantation, avascular necrosis was associated with chronic graft-versus-host disease and treatment with steroids [22]. In a study from Hopital St. Louis in Paris [27], predisposing factors included male sex, age older than 16 years, and steroid therapy for acute graft-versus-host disease. The occurrence of avascular necrosis may be related to treatment with large daily doses of corticosteroids rather than long-term therapy with smaller doses [28]. Despite the possibility of correction by hip replacement, avascular necrosis results in substantial illness [29]. The same is true for severe osteoporosis. Neuropsychological abnormalities include polyneuropathy, leukoencephalopathy, seizures, and depression. These changes occur predominantly after allogeneic bone marrow transplantation, and their relation to chronic graft-versus-host disease has not yet been defined [30]. Chronic graft-versus-host disease may aggravate the toxicity of chemotherapy and radiation. Endothelial cells have a long life span and presumably are responsible for late vascular effects. At the same time, endothelial cells are a target of the graft-versus-host reaction. Better prophylaxis of acute graft-versus-host disease has been achieved by depletion of T cells from the bone marrow graft, but in one study this advantage was compromised by an increased rate of leukemia recurrence [31]. A possible solution to this dilemma may be to transfuse donor lymphocytes after the establishment of tolerance with a T-cell-depleted bone marrow graft [32].

Treatment of chronic graft-versus-host disease has improved the prognosis of patients with this disease. The combination of prednisolone and cyclosporine has been successful in patients with standard-risk and high-risk chronic graft-versus-host disease [26]. Other forms of treatment, such as thalidomide, phototherapy, and FK506 (tacrolimus), are still to be evaluated.

Most patients regain normal or minimally impaired clinical performance and full social activity in work and school more than 5 years after transplantation. Some patients, however, develop chronic graft-versus-host disease and its sequelae, recurrence of the primary malignant condition, new malignant conditions, and other organ dysfunctions that reduce quality of life and social function. These late effects must be considered when a decision is made about the choice of time, donor, and treatment regimen for bone marrow transplantation.

Appendix

This study was a collaboration of the following physicians and centers participating in the EBMT (numbers in parentheses are numbers of patients reported): R. de Bock, MD, PhD: AZ Middelheim, Antwerp, Belgium⁵; E. Carreras, MD: Hospital Clinic, Postgraduate School of Haematology, Barcelona, Spain (25); A. Tichelli, MD, PhD: Kantonsspital, Basel, Switzerland (64); Y. Cahn, MD, PhD: Hopital Jean Mijoz, Besancon, France (13); G. Bandini, MD, PhD: Ospedale San Orsola, Bologna, Italy (19); A. Ferrant, MD, PhD: Clinique Universitaire St. Luc, Brussels, Belgium¹⁸; N. Jacobsen, MD, PhD: Rigshospitalet, Copenhagen, Denmark (9); J.P. Vernant, MD, PhD: Hopital Henri Mondor, Creteil, France (33); U.W. Schaefer, MD, PhD: University Hospital, Essen, Germany (19); M.T. van Lint, MD: Ospedale San Martino, Genova, Italy (74); C. Chabannon, MD: Hopital A. Michallon, Grenoble, France (9); T. Ruutu, MD, PhD: University Hospital, Helsinki, Finland (18); U. Saarinen, MD: Children's Hospital, University Helsinki, Helsinki, Finland (16); P. Ljungman, MD, PhD: Huddinge Hospital, Huddinge, Sweden (64); D. Niederwieser, MD, PhD: University Hospital, Innsbruck, Austria (2); M. Suttorp, MD, PhD: University Hospital, Kiel, Germany (3); M. van Weel, MD: University Hospital, Leiden, the Netherlands (26); J.H.F. Falkenburg, MD, PhD: University Hospital, Leiden, the Netherlands (17); M. Boogaerts, MD, PhD: University Hospital, Leuven, Belgium¹⁴; D. Samson, MD: Charing Cross Hospital, London, Great Britain (11); J.M. Goldman, MD, PhD: Hammersmith Hospital, London, Great Britain (40); H.G. Prentice, MD, PhD: Royal Free Hospital, Hampstead-London, Great Britain (34); A.H. Goldstone, MD, PhD: University College Hospital, London, Great Britain (6); D. Maraninchi, MD, PhD: Institute Paoli L. Calmettes, Marseilles, France (8); G. Lamberttenghi-Deliliers, MD, PhD: University of Milan, Milan, Italy (1); H.J. Kolb, MD, PhD: University Hospital, Munich, Germany (46); P. Bordigoni, MD: University Hospital, Nancy, France (2); N. Milpied, MD, PhD: Hotel Dieu, Nantes, France (19); T. de Witte, MD, PhD: University Hospital, Nijmegen, the Netherlands (14); A. Poros, MD: University Hospital, Budapest, Hungary (1); G. Socie, MD: Hopital St. Louis, Paris, France (49); C. Bernasconi, MD, PhD: Policlinico San Matteo, Pavia, Italy (2); G. Lucarelli, MD, PhD: Pesaro Hospital, Pesaro, Italy (21); W. Arcese, MD, PhD: Universita degli Studi La Sapienza, Rome, Italy (10); J.J. Cornelissen, MD: Dr. Daniel de Hoed Cancer Center, Rotterdam, the Netherlands (6); A. Iriondo, MD, PhD: Hospital Universitario Marques de Valdecilla, Santander, Spain (13); F. Freycon, MD, PhD: Hopital Nord, St. Etienne, France (2); H.H. Einsele, MD, PhD: Medizinische Universitatklinik II, Tuebingen, Germany; T. Klingebiel, MD, PhD: Kinderklinik II, Tuebingen, Germany (8 [Drs. Einsele and Klingebiel combined]); J. Nikoskelaninen, MD, PhD: Turku University, Turku, Finland (9); W. Friederich, MD, PhD: University of Ulm, Ulm, Germany (5); D. Bunjes, MD, PhD: University of Ulm, Ulm, Germany (35); H. Greinix, MD: University Hospital, Vienna, Austria (7); J. Gmuer, MD, PhD: University Hospital, Zurich, Switzerland (4).

From the University of Munich, Munich, Germany; Ospedale San Martino and Istituto Giannina Gaslini, Genova, Italy; Huddinge Hospital, Huddinge, Sweden; Kantonsspital Basel, Basel, Switzerland; Hopital St. Louis, Paris, France; Hammersmith Hospital, London, United Kingdom; and University of Ulm, Ulm, Germany.

Dr. van Lint: Department of Hematology, Ospedale San Martino, Viale Benedetto XV, 16132 Genova, Italy.

Dr. Ljungman: Department of Medicine, Huddinge Hospital, 14186 Huddinge, Sweden.

Dr. Tichelli: Department of Hematology, Kantonsspital, Petersgraben 4, CH-4031, Basel, Switzerland.

Dr. Socie: Bone Marrow Transplant Unit, Hopital St. Louis, 1 Avenue Claude Vellefaux, 75475 Paris Cedex 10, France.

Dr. Apperley: Department of Haematology, Royal Postgraduate Medical School, Hammersmith Hospital, DuCane Road, London W12 ONN, United Kingdom.

Dr. Weiss: Institut fur Klinische Physiologie und Sozialmedizin, Universitat Ulm, Frauensteige 10, Ulm, Germany.

Dr. Cohen: University Department of Pediatrics, Gaslini Institute, Largo G. Gaslini 5, 16147 Genova, Italy.

Dr. Nekolla: Stahlenbiologisches Institut, Ludwig-Maximilians-Universitat, Schillerstrasse 42, D-80336 Munchen, Germany.

Dr. Kolb: Medizinische Klinik III, Knochenmarktransplantation, Klinikum Gro ß hadern, Universitat Munchen, Marchioninistrasse 15, 81377 Munchen, Germany.