Donor CD3 Chimerism After Allogeneic Transplantation Predicts Probability Of Relapse For Patients Treated For Acute Myelogenous Leukemia (AML), Acute Lymphoblastic Leukemia (ALL), and Non-Hodkgin's Lymphoma (NHL)

The ability to predict relapse in patients after HSCT is crucial for tailoring therapy including pre transplant conditioning regimens and post transplant interventions. Donor bone marrow CD34+ cell chimerism analyses has previously been reported to be predictive for post transplant relapse for acute leukemia. Rapid immune reconstitution after HSCT may be important in the development of a robust graft versus leukemia response.

We performed for a sequential series of patients a retrospective analysis of donor cell chimerism (DCC) of peripheral blood CD3+ cells at 4 week intervals starting at day 28 after transplantation. Donor bone marrow CD34+ chimerism was evaluated at day 84 after transplantation. Percent chimerism was measured by PCR based amplification of short tandem repeats (STR). From 2009-2012, percent DCC data were collected for patients with ALL (n=39), NHL (n=26), and AML (n=68). Median age at diagnosis was 45 (range 25-69), 51 (range 23-66), and 57 (range 38-75), respectively. Patients received either PBSC or bone marrow cells from related (n=81) or unrelated donors (n=52). At transplantation, patients with ALL were in CR1 (n=27), CR2 (n=4), CR3 (n=1), PIF (n=3), relapse (n=2), PR (n=5); patients with NHL were in PR (n=9), CRU1 (n=4), CRU2 (n=3), Cr1 (n=4), CR2 (n=3), CR3 (n=4); patients with AML were in CR1 (n=38), CR2 (n=6), relapse (n=8), PIF (n=13). Conditioning regimens for patients with ALL were cyclophosphamide (cy) and fractionated total body irradiation (TBI) (n=27) or fludarabine (flu) and melphalan (mel) (n=11). Patients with NHL were conditioned with flu/cy and rituximab (n=15), or flu/mel (n=8). Patients with AML were conditioned with either a myeloablative regimen of flu/busulfan+TBI (n=51), or a reduced intensity regimens of flu/mel or flu/cy/TBI (n=16). Those who had NHL or AML and received unrelated donor cells were given rituximab. Post transplant immunosuppression was tacrolimus and methotrexate for most patients. Patients were censored for relapse or death. For analysis of DCC, patients who relapsed or did not relapse were divided into quartiles by diagnosis. The probability of relapse was tested for quartile groups using chi square analysis. Adjustments were not made for competing risks.

ALL patients in the upper three quartiles of CD3+ DCCs (1 of 34 patients, median=100%, range=84-100%, Table 1) compared to patients in the bottom quartile were less likely to relapse (3 of 7 patients, median=94%, range=82-95%, p<0.002). Similarly, the upper 2 quartiles of NHL patients who had at least 95% day 28 CD3+ DCC were less likely to relapse (2 of 16 patients, median 95%, range 63-100%, p<0.007, Table 1). 7 of 48 patients with AML in the upper 3 quartiles who had day 28 CD3+ DCC of at least 90% (median=100%, range 60-100%) relapsed, trending lower compared to those in the lower quartile (p<0.06). CD3+ DCCs for day 56 and day 84 were not statistically significant for predicting future relapse.

Day 84 CD34+ DCCs were also analyzed as described. The upper quartile of patients with ALL who did not relapse achieved a 100% DCC (1 of 23 patients relapsed, median 100%, range 82-100%, Table 2). Those that did not achieve 100% DCC for CD34+ cells by day 84 more likely to relapse (p<0.009). For patients with NHL, the upper 3 quartiles of those who did achieve 83% CD34+ DCC (1 of 7, median=90%, range 83-100%, p<0.008) by day 84 were less likely to relapse. The same method was used for patients with AML and those that did achieve 90% CD34+ DCC for cells by day 84 were not likely to relapse (5 of 35, median=96%, range=63-100%, p<0.09).

These findings suggest that failure to achieve early donor cell engraftment predicts for relapse. Relapse remains a major cause for treatment failure. Improvements in transplant conditioning may improve early lymphoid (CD3+) and myeloid (CD34+) engraftment and decrease the risk of disease relapse. Early monitoring of CD3+ DCC may identify patients who would benefit from post transplant intervention.

No relevant conflicts of interest to declare.