Reversal of Low Chimerism Following Nonmyeloablative (NM) Hematopoietic Cell Transplantation (HCT) Using Pentostatin and Donor Lymphocyte Infusions (DLI).

Our group has previously shown that DLI was ineffective for converting <40% donor CD3 chimerism to full donor chimerism. If successful, however, donor CD3 chimerism >90% elicited successful GVT effects (Blood2004;103:790). To facilitate effectiveness of DLI for patients (pts) with low or declining donor CD3 chimerism, a multicenter trial assessing the safety and efficacy of adding pentostatin was developed. Pts receiving HLA-matched NM HCTs and at risk for rejection, defined as low (<50%) but persistent (≥5%) donor CD3 chimerism and with stable/in remission disease, were included. Pts were excluded if they had evidence of relapse/progression, ongoing grades II–IV aGVHD, or chronic extensive (ce)GVHD. Since 2003, 15 pts have been enrolled on this study. Diagnoses included AML (n=5), NHL (n=4), CLL (n=3), CML (n=1), MDS/AML (n=1), and MM (n=1). Original HCT conditioning consisted of 2 Gy TBI with (n=13) or without (n=2) fludarabine (90 mg/m2); pts received PBSCs from HLA-matched related (n=11) or unrelated (n=4) donors. Postgrafting immunosuppression consisted of MMF with CSP (n=12) or tacrolimus (n=3). Median age at HCT was 54(44–66) years. DLI was given a median of 84(54–339) days after HCT. A single dose of pentostatin (4 mg/m2,adjusted for renal function) was given 2 days before infusing 107 CD3 cells/kg; no immunosuppression was given after DLI. Median donor CD3 chimerism before pentostatin/DLI was 27(5–38)%. Nine pts (all with increasing chimerism) developed aGVHD [grade I (n=2), II (n=4), III (n=2), IV (n=1)] a median of 13(0–49) days after DLI, of whom 6 developed ceGVHD a median of 85 (13–348) days after DLI. Of those who developed grades III–IV GVHD (n=3), none had evidence of GVHD post-NM HCT. The median number of significant infections post-DLI was 3(0–17) per pt. Efficacy, defined by absolute increases in CD3 chimerism ≥ 10% maintained to at least day 56 post-DLI, was seen in 10 of 15 pts with a median CD3 chimerism of 93(39–100)%. Three of 10 relapsed a median of 97(47–299) days after DLI. Four of 10 died from respiratory failure/infections (n=2), relapse (n=1), or grade IV GVHD (n=1) a median of 277(67–499) days post-DLI. The remaining 6 pts with increased chimerism are alive in CR (n=4), PR (n=1), or in relapse (n=1). In contrast, 5 of 15 had CD3 chimerism levels that were decreased/unchanged; 3 had eventual relapse a median of 61(26–530) days after DLI. None developed aGVHD. Two received 2nd NM HCTs for persistent low chimerism and both died. In total, 4 of 5 with decreased/unchanged chimerism died from either relapse (n=3) or sepsis (n=1) a median of 198(100–676) days after DLI. The remaining pt received a 2nd pentostatin/DLI (3 × 107 CD3/kg) and is under evaluation. In conclusion, pentostatin/DLI siginificantly increased donor CD3 chimerism in most pts without severe aGVHD. Lack of improving chimerism after DLI appears to correlate with disease relapse. Pts with aGVHD after initial HCT did not necessarily develop aGVHD after DLI. Infections likely related to DLI-induced neutropenia/lymphopenia need to be managed aggressively. Kinetics of CD3 Chimerism Post-DLI Kinetics of CD3 Chimerism Post-DLI