Purpose: Allogeneic bone marrow transplantation (BMT) has been shown to provide effective therapy for chronic myelogenous leukemia (CML), but previous reports have also demonstrated the persistence of bcr-abl-positive cells for months to years after BMT in the majority of patients. To evaluate the biologic significance of persistent bcr-abl-positive cells, we examined the relationship between clinical parameters known to affect the risk of relapse and the ability to detect bcr-abl-positive cells post-BMT.
Patients and methods: We analyzed 480 samples from 92 patients at two transplant centers for the presence of bcr-abl-positive cells by polymerase chain reaction (PCR). Two different BMT preparative regimens and protocols for prevention of graft-versus-host disease (GVHD) were used. One center used cyclophosphamide plus total-body irradiation (CY/TBI) and T-cell-depleted marrow; the second center used busulfan plus cyclophosphamide (Bu/CY) and untreated marrow with cyclosporine and methotrexate (Csp/MTX) as GVHD prophylaxis.
Results: We first determined the percent of patients at each center with > or = one PCR-positive (PCR+) result at defined intervals post-BMT. Between 0 and 6 months post-BMT, the majority of patients (80% to 83%) in both populations had PCR-detectable bcr-abl-positive cells. Between 6 and 24 months post-BMT, 80% to 88% of patients who received T-cell-depleted marrow remained PCR+, as compared with 26% to 30% of patients who received unmodified marrow. After 24 months post-BMT, the percentage of PCR+ patients was not significantly different in the two populations. This pattern of detection of bcr-abl-positive cells post-BMT followed the development of chronic GVHD in patients who received unmodified marrow. All patients were also divided into three groups based on post-BMT PCR results as follows: (1) persistent PCR+ (n = 29), (2) intermittent PCR-negative ([PCR-] n = 40), and (3) persistent PCR- (n = 23). These three groups were found to have a low, intermediate, and high probability of maintaining remission and disease-free survival, respectively (P = .0001). Intermittent or persistent PCR- results, which reflect levels of minimal residual disease < or = the limit of detection by PCR, were clearly associated with both acute (P = .004) and chronic (P = .000005) GVHD. Nevertheless, 44% of patients without GVHD also had intermittent or persistent PCR- assays.
Conclusion: The persistence of PCR-detectable bcr-abl-positive cells early post-BMT in more than 80% of patients suggests that neither BMT preparative regimen effectively eradicates CML cells in most patients. Subsequently, acute and/or chronic GVHD are associated with a decreased ability to detect residual bcr-abl-positive cells, which suggests that immunologic mechanisms mediated by donor cells are important for inducing long-term remissions after BMT. The demonstration that 44% of patients without GVHD had either low or undetectable levels of residual leukemia suggests the presence of mechanisms capable of suppression or eradication of CML independent of GVHD.