Donor Lymphocyte Infusions for Chronic Myeloid Leukemia Relapsing after Allogeneic Stem Cell Transplant: May We Predict Graft vs Leukemia without Graft vs Host Disease?.

Donor lymphocyte infusions (DLI) have radically changed the prognosis of patients relapsing after allogeneic hematopoietic stem cell transplant (SCT) for chronic myeloid leukemia (CML). Major obstacles to success with DLI are represented by leukemia resistance and by secondary GvHD (GvHD2). The best result of is when a patient treated with DLI achieve a durable molecular remission without experiencing GvHD2. It is unclear which factors may predict for such a favourable outcome when CML patients are treated with DLI.

We retrospectively identified 500 patients (59% males, median age 39 years, range 4–64), treated with DLI for CML relapse (81 molecular [16%], 150 cytogenetic [30%], 211 hematological chronic [42%], and 58 hematological accelerated [12%]) at 68 EBMT centers before 2004 with adequate information collected on disease response, GvHD2 and survival after DLI. Donor was an HLA-identical sibling in 73%, unrelated in 27%. DLI started with a cell dose <2×10⁷ CD3+ cells/Kg in 62% of the cases; 208 patients (42%) received 2 or more additional infusions of donor cells. Cumulative cell dose ranged from 1×10⁵ CD3+ cells/Kg to 1.4×10⁹ (median 7×10⁷). Molecular remission and/or cytogenetic complete remission was achieved in 340 patients (68%) in a median of 7.5 months (95% within 41 months). GvHD2 occurred in 60% of patients at a median of 3 months from 1^st transfusion of donor lymphocytes (95% within 24 months). Sixteen recurred at a median of 19 months (range 3–48). Actuarial probability of being alive and responsive to DLI without experiencing any GvHD2 was 29% (95% confidence interval [95CI]: 27–31%) and 27% (95CI: 24–30%) at 5 and 10 years after DLI, respectively.

We studied the prognostic effect of following factors: patient age at DLI, donor type, donor sex, sex mismatch with the donor, phase at SCT, stem cell source, T-depletion, total body irradiation in the conditioning regimen, GvHD prior to DLI, interval from SCT to DLI, type of relapse. Multivariate analysis with a Cox model adjusted for the period of DLI (≤1997 vs >1997), showed that chronic GvHD after transplant and prior to relapse (hazard ratio [HR]: 1.5, 95CI: 1.2–1.9, p<0.001), an interval from SCT to DLI <1 year (HR: 1.7, 95CI: 1.3–2.2, p<0.001), and hematological relapse (HR: 1.6, 95CI: 1.2–2.0, p<0.001), were adverse features. 94 patients (20%), 222 (48%), 133 (28%), and 17 (4%) had 0, 1, 2, and 3 adverse features, respectively. Survival in remission without experiencing GvHD2 at 5 years improved from 14%, 30%, to 56% in patients with 2–3, 1, and 0 adverse features, respectively.

We conclude that:

• “pure” GvL effect (ie. durable remission without GvHD2) was observed in more than 25% of patients treated with DLI for CML relapsing after allogeneic SCT;

• occurrence of chronic GvHD prior to relapse, the interval from SCT to DLI, and the type of relapse are the main factors associated with the chance of a “pure” GvL effect;

• patients treated with DLI beyond 1 year from SCT for a molecular/cytogenetic relapse that was not preceded by chronic GvHD have more than 50% chance of exploiting the “pure” GvL effect.