IntroductionAllogeneic Hematopoietic Stem Cell Transplantation (HSCT) is the best curative option for many patients with hematologic malignancies. Still, disease relapse remains a major clinical issue, and therapeutic options available to date are largely unsatisfactory. Thus interest has risen in how to anticipate the detection of residual malignant cells by more sensitive techniques, such as those based on quantitative PCR (qPCR), to prompt preemptive treatments. Adding further complexity to the issue of leukemia recurrence, our group described in the context of HSCT from partially incompatible donors novel variants of post-transplantation leukemia relapse, characterized by genomic loss of the mismatched HLA (Vago et al, N Engl J Med, 2009). These leukemia variants are not detected by routine diagnostic assays and are expected to be resistant to donor lymphocyte infusions, warranting the development of novel specific diagnostic assays and therapeutic algorithms. MethodsHere we prospectively validated a commercial hematopoietic chimerism assay based on qPCR detection of 34 insertion/deletion (indel) polymorphisms mapping on 19 different chromosomes, all outside the HLA complex (AlleleSEQR® Chimerism Assay, Celera Genomics). The assay was tested on 403 bone marrow samples harvested from 115 transplanted patients during 18 months of follow-up. Results obtained from qPCR-based chimersim were compared with standard Short Tandem Repeat (STR) chimerism analysis performed in parallel on the same samples. To discriminate between “classical” and “HLA loss” relapses, we developed chimerism assays targeted to specific HLA-A allele groups, based on the same qPCR technology of the commercial “outside HLA” assay and targeted to display its same sensitivity. ResultsqPCR chimerism displayed high informativity, providing at least a patient-specific indel marker in all the patients analyzed and at least two markers in 109/115 patients (94.7%). When two polymorphic indel markers were analyzed in parallel on the same sample concordance was extremely high (r2=0.97). As expected, qPCR chimerism demonstrated a higher sensitivity in detection of residual host cells as compared to STR: 131/403 samples exhibit a host chimerism above 0.5% (32.5%) and 91/403 above 1% (22.6%), whereas 79/395 (20%) were positive in STR. qPCR chimerism could predict impending relapse with a sensitivity of 37.5% and a specificity of 72.5% for a 0.5% host threshold, and with a sensitivity of 29.2% and a specificity of 83.5% for a 1% host threshold, comparing favourably in terms of sensitivity with STR (sensitivity 20.8%, specificity 84.8%). We designed and validated 5 qPCR reactions able to specifically detect 9 different HLA-A allele groups, displaying a combined frequency of 81% in Caucasians, and providing an informative patient-specific HLA-A allele as chimerism marker to more than 50% of the patients transplanted from mismatched related donors in our centre. Technical validation of the newly developed reactions with artificial mixes of HLA-typed DNA demonstrated for each reaction the sensitivity to detect at least 0.2% target cells and efficiency above 80%. Clinical utility was tested on 80 bone marrow samples harvested from transplanted patients, using the new “inside HLA” reactions in combination with the commercial “outside HLA” assay: combination of the results obtained from the two assays allowed us to unequivocally discriminate HLA loss relapses (n=5) from classical relapses (n=15). ConclusionHematopoietic chimerism detection by qPCR appears promising, displaying a higher sensitivity and a superimposable specificity as compared to techniques currently in use, and allowing by the novel assay we developed the simultaneous detection of leukemic variants characterized by genomic HLA loss. Disclosures:Bonini:MolMed SpA: Consultancy. Bost:Celera Genomics: Former Employee, Former Employee Other; KimerDx: Employment. Fleischhauer:Celera Genomics: Research Funding.