Human-Leukocyte-Histocompatibility Antigens (HLA-A1+, A2-, B44-) and Serum Immunoglobulin (Ig)G with CD4 Levels Predict Response to Graft-Versus-Leukemia (GVL) and Overall Survival, Respectively, After Non-Myeloablative Allogeneic Stem Transplantation (NST) for Chronic Lymphocytic Leukemia (CLL).

The major focus of NST in CLL has been to reduce mortality in this elderly patient (pt) population and attempt to exploit GVL to achieve remission (CR). Donor T cells recognizing host-derived minor histocompatibility antigens presented on the cell surface by MHC class I and II molecules have been implicated in GVL reactions.

We aimed to analyze predictors of survival after NST in CLL and whether certain HLA alleles are associated with responses to immunomanipulation (IMM)(withdrawal of immunosuppression ± rituximab and step-wise donor lymphocyte infusions (DLI).

We included all 86 CLL pts who had undergone NST in sequential phase II protocols from February 1996 to August 2007. Eligibility criteria included pts aged 19-70 years with CLL who have failed purine analogs + rituximab and had a suitable donor. The preparative regimen used was FCR(fludarabine, 30 mg/m² , and cyclphosphamide 750 mg/m², each given intravenously on days –5 to –3 prior to transplant, 375 mg/m² of rituximab on day –13, and 1000 mg/m² on days -6, +1, and +8 ). Patients and donors were typed for alleles at HLA-A, B, C, DRB1, DRB3/4/5 and DQB1 by PCR amplification. Donors were matched siblings in 43 (50%) pts, matched unrelated in 30 (35%), and mismatched unrelated in 13 (15%) pts. Graft was from peripheral blood (PBSC) in 61 (71%) of pts, and marrow in 25 (19%) other pts. GVHD prophylaxis consisted of tacrolimus (administered for 6 months) and methotrexate. Pts who underwent non-sibling transplantation also received anti-thymocyte globulin or alemtuzumab (n=28, total dose 45mg). Post-transplant IMM was performed in pts with progressive or residual disease..

Median age (range) was 58 (36-70), and the median Hematopoietic Stem Cell Comorbidity Index (HSCT-CI) was 3 (range, 0-8). Prior to their transplantation, pts were exposed to a median of 3 prior chemotherapies, and 42 pts (48%) had failed alemtuzumab. IgVH was unmutated in 28 (72%) of 39 pts tested, and 15 of 65 (23%) had p53 mutation. The clinical factors used to assess outcome included: recipient and donor age, disease stage, % lymphocytes, %CD5-CD19, b2-microglobulin, LDH, PET-status, number of prior treatments, use of alemtuzumab, HSCT-CI, serum levels of IgG, IgA, IgM, CD4 and CD8, source of graft, # cells (CD34, CD3) infused, chimerism at day 90, acute and chronic GVHD. In addition, IgVH and P53 mutation, and HLA factors were also analyzed. In a multivariate model, the major determinant of survival was the level of immunosuppression as determined by CD4 <100/L, and IgG below normal levels at the time of transplantation (HR 7.1; p<0.0001). With a median follow-up of 37.2 months (range, 11.4-131.1 months), the 1- and 5-year overall survival rates were 93% and 68% vs. 25% and 17% (p<0.0001) for pts who had CD4>100/L and IgG level within the normal range vs those whose a CD4<100/L and IgG level below normal at time of study entry, respectively. Pts' age (< or > 60 years), IgVH and p53 mutation, and HSCT-CI were not found to be important factors in determining survival. Pts had refractory disease at transplantation, and 43 pts required IMM after NST. Of these 43 pts, 20 (47%) achieved CR. Pts characteristics at study entry (described above) for NST, and at the time of initiation of IMM, including number of DLI infused (median 2, range 1-6) and maximal dose (median 43.6, range 1-200, x 10⁶ CD3+/Kg), T-cell chimerism and GVHD, as well as HLA subtypes were assessed. The major determinants to achieve CR following IMM included receipt of a PBSC graft and achievement of a good (≥ 90%) donor T-cell chimerism at day 90 (p=0.035), and having a combination of HLA-A1+/A2-/B44- (p= 0.0009). The rate of CR to IMM was 9%, 36%, 50%, 91% respectively in patients who had none of the HLA factors described, I, 2, and all 3 respectively. There was no statistically significant difference in pts and disease characteristics, nor in the risks of GVHD between the respective HLA-subtypes described. The Current-progression-free survival rates at 5-year for patients with HLA-A1+/A2-/B44- vs those who had none of those HLA types were 68% vs 15%, respectively, (p<0.0001).

These data provide first evidence that IgG and CD4 levels may predict survival after NST in CLL. Our results also represent the first report showing certain HLA alleles might be predictive for response to GVL and achieving long-term remission in CLL. Verification of our results in a larger cohort of pts is highly warranted for better selecting pts for IMM for the treatment of recurrent malignancy in CLL.

No relevant conflicts of interest to declare.