Chimerism Analysis of Free Circulating DNA in the Prediction of Relapse in Patients with Acute Leukemia Treated with Stem Cell Transplantation,

Predicting relapse after hematopoietic stem cell transplantation (HSCT) in hematologic malignancies remains a challenge, especially when there is no specific molecular marker for the leukemic cells. Early detection of relapse and intervention prior to florid relapse will, in general, improve outcome. Donor chimerism has been extensively explored in predicting relapse, but specificity and sensitivity of this approach remain low. Currently, post-transplant chimerism evaluation is performed on all circulating cells or on subpopulation of the circulating cells. However, decrease in donor DNA (cells) can be caused by factors other than relapsing leukemic cells. Free circulating DNA in plasma has been used for diagnosis and prediction of cancer and for early detection of relapse. Leukemic cells have high turnover pouring their DNA into circulation at a higher rate than normal cells. Because of this higher rate of turnover, plasma is enriched by leukemia-specific DNA. We hypothesized that plasma chimerism analysis may detect leukemia relapse earlier than cell chimerism. We studied DNA chimerism in the plasma from patients with hematologic malignancies and compared the pattern with that of cell chimerism. Most of our patients had acute myeloid leukemia (AML) or acute lymphoblastic leukemia (ALL) and were all treated with allogeneic hematopoietic stem cell transplantation (HSCT).

CD3+ cells were separated from the myeloid cells using beads. Chimerism analysis was performed on both populations to determine the relative ratio of donor DNA. In addition, DNA was extracted from the plasma from the same samples and chimerism evaluation was carried out in the same fashion as cells.

We first analyzed samples from patients who had aplastic anemia but treated with allogeneic HSCT (N=11). These samples demonstrated that the plasma DNA chimerism is comparable to that seen in the granulocytes and significantly different from the chimerism in the lymphocytes. Since these patients are physiologically normal, this suggests that the majority of the circulating plasma DNA is generated by the turnover of granulocytes. Then we analyzed patients transplanted as a treatment for leukemia (N=84) and had 100% donor DNA in their granulocytes. Of these, 16 (19%) patients had clinical evidence of relapse. All patients with relapse had >10% recipient DNA in the plasma reflecting the relapsing leukemic cells. One of these patients had a relapse only in testes without bone marrow involvement and his plasma chimerism was positive for recipient DNA while both lymphocytes and granulocytes showed 100% donor DNA. Only three of the 16 patients (19%) had more recipient DNA in the lymphocytes chimerism analysis. However, of the 84 samples, 16 samples (19%) showed more recipient DNA (>10%) in plasma than in granulocytes without evidence of relapse, but almost all these patients had neutropenia or thrombocytopenia. These patients are being followed up to determine if they will develop leukemia. Eight additional patients with mixed chimerism in granulocytes and lymphocytes were studied for DNA chimerism in plasma. Three of these patients had more recipient DNA in plasma than in granulocytes and the three patients had evidence of relapse, while the rest of the patients had no significant relative increase in recipient DNA (>10%) in plasma and had no evidence of relapse.

Chimerism studies of plasma DNA might be useful in predicting early relapse after HSCT in patients with acute leukemia. Although further studies are needed, our data suggests that any increase above 10% in percentage of recipient DNA in plasma as compared to that in granulocytes should alert to potential relapse. This approach in predicting early relapse has significant advantage because it can be applied to all leukemias and does not require a leukemia specific marker.

No relevant conflicts of interest to declare.