Donor Lymphocyte Infusion after Allogeneic Haematopoietic Stem Cell Transplant, a Single Centre Retrospective Study

Introduction

Allogeneic haematopoietic stem cell transplant (allo-HSCT) is the only curative treatment for many haematological malignancies, however, relapse remains the main cause of treatment failure. Relapse is associated with dismal prognosis and poor survival, therefore, post allo-HSCT strategies to monitor, recognise and prevent impending relapse are of major value. Donor lymphocyte infusions (DLI) are used pre-emptively in patients with mixed donor chimerism (MDC) to enhance the graft-versus-leukaemia effect, in high risk patients as a prophylactic measure to prevent relapse and in relapsed patients with or without salvage chemotherapy.

Patients and Methods

One hundred and twenty-five patients received DLI between January 2010 and June 2017 in our institution. Median age at transplant was 57 years (range 25-74). Sixty-three patients had acute myeloid leukaemia (AML), 16 had myelodysplastic syndrome (MDS), 17 had multiple myeloma (MM), 12 had lymphoproliferative disorders, 11 had myeloproliferative disorders (MPD) and 6 had acute lymphoblastic leukaemia (ALL). Sixty-five percent of patients were in complete remission at the time of allo-HSCT, 24% in partial remission and 11% underwent an allo-HSCT without any previous treatment. Fifty-eight patients had a matched sibling donor, 56 a matched unrelated donor and 11 a mismatched unrelated donor. Ninety percent of patients received a reduced intensity conditioning regimen and 83% with T-cell depletion. MDC was defined as donor CD3 percentage less than 95%. DLI was given pre-emptively in patients with MDC, prophylactically for high risk AML/MDS patients and to patients with haematological relapse. DLI was administered using escalating doses.

Results

Sixty-six patients received pre-emptive DLI, 9 patients prophylactic DLI and in 50 patients DLI was a part of their salvage treatment. Median doses were 0.1 x10⁷/kg for the first DLI, 0.5 x10⁷/kg for the second, 1.0 x10⁷/kg for the third, 5 x10⁷/kg for the fourth and >5 x10⁷/kg for the subsequent doses. Median follow up from day of first DLI was 12 months (range 1-88).

In the pre-emptive group median donor chimerism for CD3 at the time of DLI was 64.5% (range 5-94), median days from transplant to first DLI was 161 days (range 74-715) and median DLI administration was 1 (range 1-6). A total of 32 patients (48%) developed graft-versus host disease (GvHD) post-DLI; grade 1-2 in 21 patients (32%), grade 3 in 8 (12%) and grade 4 in 3 (4%). Two patients died due to GvHD. Median time between the last DLI and GvHD was 51 days (range 13-112). Seventy percent achieved full donor chimerism. Relapse was seen in 13 patients (20%), among them 4 are still alive.

In the prophylactic group all patients but 1 received FLAMSA-TBI as a conditioning regimen. All had full donor chimerisms when they received DLI. Median time from transplant to DLI was 173 days (range 43-263) with a median of 2 DLI (range 1-4). Only two patients developed GvHD, there was no grade 3-4 GvHD. Five patients (56%) relapsed and died, and one patient died due to a secondary malignancy being in complete remission.

In the treatment group, DLI was combined with Azacitidine in 14 patients, with Lenalidomide in 13 patients and with chemotherapy in 10 patients, 13 patients received DLI without any salvage treatment. Median time to DLI was 271 days (range 67-2476) with a median of 2 DLI (range 1-6). GvHD occurred in 13 patients (26%), one patient had grade 3 GvHD, all others had grade 1-2 GvHD. Sixty-eight percent died because of disease progression.

Overall survival (OS) for the whole cohort at 1, 2 and 5 years after the first DLI was 65%, 52% and 39% respectively. Patients receiving pre-emptive DLI had better survival (OS of 83% at 1 year, 74% at 2 years and 57% at 5 years) compared to the treatment group (OS of 75% at 1 year, 55% at 2 years and 24% at 5 years) and to the prophylactic group (OS of 44% at 1 year, 33% at 2 years and 16% at 54 months) (p < 0.005). GvHD was associated with better survival for the entire group (p=0.009).

Conclusions

In our cohort of patients pre-emptive DLI showed better results when compared to prophylactic and treatment DLI, although the numbers in the prophylactic group are scanty. GvHD post-DLI was associated with better outcomes.