|
|
| |||||||||||||||||||||||||||||||
Background Patients with homozygous beta-thalassemia, who have a good prognosis during treatment with conventional therapy, appear to have an especially high probability of hematologic cure with bone marrow transplantation, although the morbidity and mortality associated with such treatment are not established.
Methods The records of all patients with thalassemia who received bone marrow transplants from HLA-identical donors in Pesaro, Italy, were examined from October 1982 through May 1992. Detailed evaluation of the outcome was conducted in the 89 patients identified as being in class 1 according to the Pesaro classification, in which hepatomegaly, portal fibrosis, and the inadequacy of iron chelation therapy are considered independent risk factors, and the patients are classified as being in class 1 if none of these factors are present, class 2 if one or two of the factors are present, and class 3 if all three factors are present. Sixty-four of the patients had been prepared for transplantation with a drug regimen in current use that includes busulfan and cyclophosphamide followed by cyclosporine as prophylaxis against acute graft-versus-host disease (protocol 6).
Results There were seven deaths, all within 101 days of transplantation. Two of the 64 patients treated according to protocol 6 died. The probabilities of survival, rejection-free survival, death from causes unrelated to rejection, and rejection were 0.92, 0.85, 0.06, and 0.08, respectively, in the total group and 0.97, 0.93, 0.03, and 0.04 in the 64 patients treated according to protocol 6. Preliminary evidence suggests that there was useful unloading of tissue iron deposits.
Conclusions The high probability of cure with little early or late morbidity and mortality suggests that patients with class 1 thalassemia who have HLA-identical donors available should be treated by bone marrow transplantation. However, this was not a controlled trial, so we cannot directly compare the outcome with that of conventional treatment.
We recently described the results of bone marrow transplantation in adults whose conditioning regimen was assigned on the basis of the prognostic class at the time of transplantation3. All patients were in class 2 or 3, and marrow transplantation was undertaken because the disease progressed despite conventional therapy. The probabilities of survival, disease-free survival, and rejection were 0.85, 0.80, and 0.05, respectively; the survival rate reached a plateau after six months and remained at that level for three years. We concluded that marrow transplantation is effective and should be offered to adults with progressive thalassemia who have suitable donors.
Marrow transplantation in patients in class 1 has been more controversial, since these patients by definition do not have signs of serious deterioration with conventional treatment, and they may have prolonged and relatively uncomplicated survival if such treatment is continued. Marrow transplantation offers a high probability of cure for these patients, but the morbidity and mortality associated with this approach are important factors in the decision to proceed with transplantation in such a setting.
We report the results of bone marrow transplantation in patients with class 1 thalassemia in Pesaro, Italy.
Methods
Between October 21, 1982, and June 4, 1992, 537 marrow transplantations for homozygous beta-thalassemia were performed in the Division of Hematology and Marrow Transplantation of the Hospital of Pesaro in Pesaro, Italy. Five hundred eighteen of these transplants were from HLA-identical siblings or parents. At the time of transplantation, 89 patients, 6 of whom had parents as donors, were in class 1. A complete absence of globin beta-chain synthesis was demonstrated in 52 of these patients.
The HLA-A, B, and C loci were matched for each patient and donor, and samples from each donor-recipient pair were mutually unresponsive in the mixed-leukocyte test.
We employed a system for assigning patients undergoing marrow transplantation for thalassemia to prognostically useful categories1. Risk factors considered included the presence of hepatomegaly (enlargement of more than 2 cm below the intercostal margin), the presence of portal fibrosis in the pretransplantation liver-biopsy specimen, and the quality of iron chelation in the years before transplantation. The quality of chelation was characterized as adequate when deferoxamine therapy was initiated within 18 months of the first transfusion and administered subcutaneously for 8 to 10 hours continuously for at least 5 days each week. Chelation was defined as inadequate if there was any deviation from this requirement. All the patients in this study were in class 1. Table 1 lists some characteristics of the patients at the time of transplantation.
|
Seven different protocols were used, but most patients were treated with protocol 3 or 6, which were modifications of the regimen described by Santos5. All patients received busulfan orally three times daily for four days (total dose, 14 mg per kilogram of body weight), followed by intravenous cyclophosphamide once daily for four days (total dose, 200 mg per kilogram). Marrow was infused 36 hours after the last dose of cyclophosphamide. The day of marrow infusion was designated day 0. Fifteen patients received 7.5 mg of cyclophosphamide per kilogram on day 1 and 10 mg of methotrexate per square meter of body-surface area on days 3, 6, and 11 and weekly thereafter for 102 days (protocol 3) as prophylaxis against acute graft-versus-host disease (GVHD). Sixty-four patients received 5 mg of cyclosporine per kilogram daily intravenously from day -2 through day 5, followed by a daily dose of 3 mg per kilogram intravenously until they could tolerate oral administration at a daily dose of 12.5 mg per kilogram (protocol 6). The dose of cyclosporine was tapered from day 60 until the drug was discontinued after one year.
All patients were treated in positive-pressure isolation rooms with oral nonabsorbable antibiotics. The first six patients received prophylaxis against systemic infection with amikacin and piperacillin beginning the day before transplantation and continuing until the granulocyte count exceeded 500 per cubic millimeter and the patient was afebrile; persistent unexplained fever was treated with amphotericin B as soon as the fever occurred. Subsequent patients received ceftazidime instead of piperacillin, together with amphotericin B prophylactically beginning on day 8. All blood products given after transplantation were irradiated with a 30-Gy dose. Acute and chronic GVHD were graded according to the Seattle criteria6,7. The drug of first choice for the treatment of acute GVHD was prednisolone given in escalating doses up to 10 mg per kilogram.
Cytogenetic analyses were performed on unstimulated marrow and phytohemagglutinin-stimulated peripheral blood if the donor and recipient were of opposite sexes. Globin-chain synthesis of marrow and peripheral-blood reticulocytes was examined by measuring the incorporation of [H3]leucine, followed by column or high-performance liquid chromatography. Analyses of the karyotype and the pattern of globin synthesis were performed on peripheral blood and bone marrow on day 13 and were repeated at two-week intervals until day 60.
Survival distributions were estimated by the product-limit method8 and tested for equality by the Mantel-Cox9 and Breslow10 statistics. In estimating event-free survival, we identified rejection, the recurrence of thalassemia, and death as events. Rejection was defined as the development of complete marrow aplasia or the recurrence of thalassemia (a return to the pretransplantation pattern of globin-chain synthesis). Jeffreys minisatellite single-locus DNA probes (MS1, MS31, MS43, and g3) were used to distinguish donor and recipient cells in an analysis of restriction-fragment-length polymorphisms of variable-number tandem-repeat loci11,12.
Patients were advised to return for evaluation 6 and 12 months after transplantation and annually thereafter. The marrow and peripheral blood were examined by chromatography and, when appropriate, with cytogenetic techniques at the six-month, one-year, and two-year visits. Other characteristics evaluated included iron metabolism; liver, cardiac, and endocrine function; and growth and development. Height and weight were measured, and physical examinations were conducted yearly. The Z score was calculated with the formula (X - Xm)/SD, where X is the patient's height, Xm is the mean height for chronologic age, and SD is the standard deviation of the reference value obtained from Tanner growth charts13. The pubertal stage was evaluated according to the criteria of Tanner and Whitehouse14.
At the beginning of the study, liver biopsies were performed annually, but because the results of early post-transplantation examinations showed little change, patients enrolled later in the study had the first post-transplantation liver biopsy after three years. Serologic tests were performed on all patients to determine whether they had been exposed to hepatitis B virus. Tests for antibodies against hepatitis C virus (by second-generation enzyme-linked immunosorbent assay) were not available until later in the series, and they were performed in 14 patients.
Informed consent was obtained from the parents of all patients after the procedures and the risks involved in marrow transplantation had been explained in detail. The explanations given to the patients varied in complexity according to their age. Particular emphasis was placed on the option of continuing conventional management with transfusions and deferoxamine.
Results
Survival and Rejection
The number of patients available for follow-up at the end of each of the first nine years after transplantation is shown in Table 2. Figure 1 shows the probabilities of survival, event-free survival, death from causes other than rejection, and rejection for all patients. The last death occurred on day 101, and the last episode of rejection was on day 365. The curves for survival, event-free survival, mortality unrelated to rejection, and rejection leveled off at 0.92, 0.85, 0.06, and 0.08, respectively.
|
|
The probabilities of survival, event-free survival, death from causes unrelated to rejection, and rejection for the 64 patients treated according to protocol 6 leveled off at 0.97, 0.93, 0.03, and 0.04, respectively. The two deaths occurred 94 and 101 days after severe acute GVHD and were due to pulmonary infection and liver failure, respectively. Two patients had graft rejection, both of whom survived with recurrent thalassemia. Sixty of the patients were alive and free of thalassemia at the most recent follow-up. Table 3 summarizes the outcomes in the patients in the study.
|
The probability of acute GVHD of grade 2 or worse was 27 percent by day 100 (Figure 2). Acute GVHD of grade 3 developed in 14 patients, and acute GVHD of grade 4 developed in 2 patients.
|
Causes of Death
Seven patients died, all within 101 days of transplantation. One of these patients had graft rejection. Among the 73 patients who did not have acute GVHD of grade 3 or 4, 1 died of fungal disease on day 8. Five patients died of severe acute GVHD: two had acute GVHD of the liver, and three had pulmonary complications associated with severe, cutaneous acute GVHD.
Status of Patients Who Survived without Thalassemi
Hematopoiesis
Two survivors had transient episodes of mixed chimerism without anemia or the need for therapeutic intervention, and one of these patients had cytogenetic evidence of mixed chimerism in peripheral blood and marrow. As of September 1992, none of the survivors had mixed chimerism for patterns of globin-chain synthesis, and with the exception of the two patients with recurrent thalassemia, the patients' red-cell types were those of the donors.
Iron Metabolism
Figure 3 shows the mean serum ferritin levels each year after transplantation together with the number of patients in whom serum ferritin was measured and the number available for evaluation. The mean serum ferritin level of 1575 ng per milliliter before transplantation increased to 1685 ng per milliliter 12 months after transplantation, with a progressive decrease thereafter. Iron-binding capacity progressively increased over the same seven-year period.
|
Hepatic siderosis was categorized as absent, mild, moderate, or severe. All 58 patients who underwent liver biopsy at the time of transplantation had hemosiderosis. Follow-up data were available for 31 patients, 29 of whom had mild-to-moderate hemosiderosis at the time of transplantation. In 10 of these patients the severity of hepatic hemosiderosis improved after transplantation, with a return to normal in 7. Two patients with severe hepatic hemosiderosis at transplantation had subsequent improvement (to mild disease in one patient and to moderate in the other). Because of the decision to restrict liver biopsies as described in the Methods section, only limited information was available from subsequent samples.
Hepatitis
As of September 1992, the 2 patients with mild chronic aggressive hepatitis and 9 of the 15 patients with mild chronic persistent hepatitis before transplantation had not yet reached the time for another biopsy. Four of the six patients who presented with mild chronic aggressive hepatitis and who had had subsequent examinations had no signs of hepatitis; in one the disease had not progressed, and in another the disease progressed to moderate chronic aggressive hepatitis, first detected 12 months after transplantation and persisting for 48 months. This patient was one of two in whom nodular cirrhosis developed.
Portal Fibrosis
Additional biopsy samples were available for evaluation in 34 of the 58 patients with confirmed absence of portal fibrosis at the time of transplantation. In one of these patients, changes related to portal fibrosis developed that appeared mild 12 months after transplantation and moderate at 24, 36, and 48 months, and progressed to nodular cirrhosis at 60 months. Severe fibrosis developed in a second patient at 12 months and progressed to cirrhosis at 36 months. Both patients with cirrhosis underwent transplantation before tests for hepatitis C virus were available, and both had antibodies against hepatitis C virus when first tested after transplantation.
Growth and Development
Growth and development were evaluated in 70 patients followed for a minimum of one year. The mean Z score for height was -1.22 before transplantation and improved to +0.16 seven years after transplantation in patients in whom chronic GVHD did not develop. In two patients in whom chronic GVHD developed after transplantation, the Z scores were -1.77 and -1.86.
Two girls and one boy were pubertal at the time of transplantation; all other patients were prepubertal. As of September 1992, 13 girls were older than 11 years, 7 of whom had pubertal development, and 8 boys were older than 12 years, 4 of whom had pubertal development.
Discussion
In the first patient in this series, who was among the first to undergo transplantation for thalassemia in Pesaro, engraftment did not occur and thalassemia returned. We abandoned the regimen employed for this patient because it was insufficiently immunosuppressive, and we treated the next five patients with higher doses of busulfan along with cyclophosphamide, and in some instances with total-body irradiation. Three patients died of transplant-related causes and one of rejection with failure to reconstitute host marrow. These regimens were considered too toxic, and we replaced them with protocol 3, which included a lower dose of busulfan. In 2 of the 15 patients in class 1 who received this regimen, the donor marrow was rejected and autologous reconstitution occurred. The rejection rate with this protocol was particularly high in patients in class 2 or 3, and the GVHD prophylaxis was modified by adopting the regimen reported by Storb et al16. In a randomized study of patients in all classes, we demonstrated that cyclosporine alone was as effective as the combination of methotrexate and cyclosporine but was associated with less early mucosal toxicity. This regimen was adopted for protocol 6, which has been used for all patients in class 1 since January 1986.
Among the 64 patients in class 1 who underwent transplantation after the use of protocol 6 was instituted, there were 62 survivors, 60 of whom were free of hematologic thalassemia. These 60 patients have not required transfusions since they were discharged from the hospital after transplantation. Preliminary studies of iron transport show that iron overload does not progress after successful transplantation and that unloading of the tissue iron deposits is possible. This process appears to take several years, even in patients in class 1. Of six patients with histologic evidence of hepatitis at transplantation who could be evaluated, only two had such evidence at the most recent follow-up, and the condition progressed in one. The absence of evidence of hepatitis in the other four patients is an encouraging sign. It is not known whether the progress of similar lesions will be reversed in patients with more advanced disease who undergo transplantation.
In a recent report of adults undergoing transplantation for thalassemia,3 all patients were in class 2 or 3 with extramedullary organ damage due to the disease and its treatment. There was little hesitation in using bone marrow transplantation in such patients because of the poor prognosis with continuing conventional therapy. Transplantation eliminated the marrow defect in most (80 percent) of the patients. Removing the ongoing cause of the organ damage in these patients is clearly desirable, but the potential for healing and recovery is unknown in patients with advanced disease, and some damage, such as chronic persistent and chronic active hepatitis, may be progressive.
Patients in class 1 have less iron overload than patients in class 2 or 3. Other complications of intensive transfusion therapy, such as liver damage (often mediated by hepatitis B and C infection), are seen more frequently in patients in class 2 than in patients in class 1, and they are uniformly present in patients in class 3. Unfortunately, we do not know what the probability is that a patient receiving conventional therapy will at some point be reassigned to a higher-risk category, but transplantation centers frequently encounter patients in class 2 or 3 who have not responded to conventional treatment. Delaying transplantation until patients are in class 2 or 3 substantially reduces the probability of success of transplantation and may jeopardize the reversibility of liver and cardiac damage.
Although this setting does not permit the use of a randomized, controlled trial, our findings of a high cure rate with low rates of early and late morbidity and mortality suggest that patients with thalassemia in class 1 should be treated by bone marrow transplantation if HLA-identical donors are available.
Supported by the Italian Association against Cancer, Milan, the Italian Association against Leukemia, Pesaro, and the Regione Marche, Ancona.
Source Information
From the Divisione Ematologica e Centro Trapianto Midollo Osseo di Muraglia, Ospedale di Pesaro, Pesaro, Italy (G.L., M.G., P.P., E.A., D.B., C.G., M.A., F. Agostinelli, F. Albertini), and the Fred Hutchinson Cancer Research Center, Seattle (R.A.C.).
Address reprint requests to Mr. Clift at 2919 129th Ave. N.E., Seattle, WA 98104.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
-Thalassemia.