Kiler Immunoglobulin Receptor(KIR) Ligands Are Associated with At Least Partial (HLA A*03, B*44) or Poorer (HLA B*51, C*04) Response to Thalidomide Among Patients with Multiple Myeloma Treated in Two Turkish Centers

We have previously presented HLA DRB1*13 and 11 tobe associated with response (IMMW, Paris, 2011). In this analysis we have expanded the sample size and have included Killer Immunoglobulin Receptor (KIR) ligands HLA A, Bw4 and C epitopes to further understand the mechanisms underlying this association.

115 patients diagnosed and treated in Ankara and Gazi University Hospitals and who have received a minimum of two months of Thalidomide (T), either alone or within a combination for the treatment of newly diagnosed or relapse myeloma were included. Treatment was given as either part of induction or consolidation: monotherapy (n= 54) or combination (n= 61), at a dose of median 100 mg (50–600 mg) for a median of 7 months (2–114 mo). Following consent HLA typing was done using Sequence-Specific Oligonucleotide primers. IMWG criteria was used for response assessment. Statistics were performed using the SPPS 15. 0 program. The most frequently observed three alleles at each loci and KIR ligands (C epitopes 1, 2 and Bw4) were evaluated to determine effects on response to T. Logistic regression analysis was used for univariateand multivariate analysis to evaluate the factors affecting response to thalidomide.

Fifty-two (45%) of the patients achieved at least PR. ≥PR and <PR responders were similar in respect to age, time from the diagnosis to initiation of thalidomide, gender, serum β₂-microglobulin, albumin, LDH, and creatinin levels, paraprotein characteristics, bone marrow plasma cell ratio, presence of ASCT or allo-HCT, response to induction treatments except for the median number of chemotherapy lines prior to T use which was higher in the less responding group compared to ≥PR responders (p=0. 016). Comparison of these KIR ligand frequencies revealed HLA A*03, B*51, C*04 and DRB1*13 to be associated with response significantly (Tables). The frequency of A*03∼B*44∼DRB1*13, A*03∼ DRB1*13 or A*03∼ B*44, B*44∼ DRB1*13 (contains ≥PR response associated specificities) patients responded frequently:1/1, 7/9, 4/5 or 3/3. Similarly A*11∼B*51∼C*04 or A*11∼B*51∼DRB1*11, A*11∼B*51 or B*51∼DRB1*11 (combining <PR response associated specificities) were not responding as well: 1/4, 0/2, 1/5 or 2/13. The OR for HLA A*03∼DRB1*13, B*51∼DRB1*11, B*51∼C*04, but not C*04 ∼DRB1*11 showed a synergistic effect of the combination (Tables).

Host related factors such as HLA DRB1 and KIR ligands are found to exert either inhibitory or activating effects on response to T. In addition to the number of previous lines of therapy, HLA, deserves to be included into the list of factors effecting response.

No relevant conflicts of interest to declare.