Polymerase chain reaction-based detection of minimal residual disease in acute lymphoblastic leukemia predicts relapse after allogeneic BMT.

We studied 20 patients who had received allogeneic bone marrow transplants for acute lymphoblastic leukemia. Using the polymerase chain reaction (PCR), we examined them for the presence of immunoglobulin heavy chain variable-diversity-junctional (IgH V-D-J) gene rearrangements as a marker of minimal residual disease (MRD). Seventeen of the patients were transplanted in relapse or for primary refractory disease, and three were transplanted in "high risk" remission. Thirteen patients had at least one positive PCR assay for MRD in the first 100 days after transplant; two died of nonleukemic causes, while all the remaining 11 relapsed. Seven patients had all negative PCR assays; two died of nonleukemic causes, four remain alive and well, and one patient with consistently negative PCR assays relapsed 47 days after his last negative test. Considering PCR status as a time-dependent covariate, the relative risk of relapse for PCR-positive patients compared to PCR-negative patients is 6.4 (p = 0.0067; 95% CI, 1.6-24.4), and the relative risk of relapse or death is 4.8 (p = 0.0052; 95% CI, 1.6-14.2). The Kaplan-Meier estimate of relapse at 1 year for PCR-positive patients is 100% compared to 33% for PCR-negative patients. The estimate for disease-free survival at 1 year for PCR-positive patients is 0% compared to 44% at 3 years for the PCR-negative group. The PCR assay for MRD appears to be a sensitive and specific test, identifying patients at high risk of relapse following allogeneic BMT who might benefit from further therapeutic interventions.