We previously demonstrated frequent genomic alterations measured by microsatellite instability (MSI) in non-neoplastic epithelial tissues of pts who underwent allogeneic Hematopoietic Cell Transplantation (HCT) but not after autologous transplantation (Blood 2006; 107:3389–3396). Confirmation in a larger and independent patient cohort and demonstration in an in vitro system was needed. In total 176 buccal samples obtained from 71 unselected patients who underwent allogeneic HCT were analyzed for MSI with standard PCR for microsatellite markers (THO-1, SEE33, D14S120 and D1S80) and analysis by capillary electrophoresis. MSI was observed in 37 (52%) allo-transplanted pts but never in the 31 controls (16 healthy and 15 pts after auto-HCT, 47 samples). MSI+ pts were significantly older than MSI-pts (median age 60y vs 48y, p<0.05). Other clinical features were not significantly different between MSI+ and MSI− pts: gender, myeloid vs lymphoid disease, sibling vs unrelated HCT, myeloablative vs reduced intensity conditioning, sex mismatched transplantation, bone marrow vs mobilized peripheral blood graft, T-cell depletion of the graft, use of methotrexate for GvHD prophylaxis, number of CD3+ cells/kg b.w. infused, occurrence and grade of mucositis during neutropenia, CMV reactivation, median ferritin levels post-transplant and occurrence of GvHD vs no GvHD. When GVHD was analyzed according to its severity we found that there was a trend for increasing risk of MSI for pts who experienced severe GvHD (RR=1.8) as compared to pts with no GvHD (RR=1). Secondary malignancy (5 skin-, 1 adeno-Ca) was diagnosed in 5 (14%) of the MSI+ pts and only in 1 (3%) MSI− pts. In an in vitro model of mutation analysis we tested the hypothesis that an alloantigenic stimulus is substantially involved in the mutation process. Briefly, keratinocyte DNA-repair proficient HaCaT cells were transfected with a reporter plasmid and genomic instability (GI) was detected and measured as hygromycin resistant clones. Treatment of HaCaT cells with H2O2 resulted in a time and dose dependent GI induction (1.9–5.7 fold). Cytokines abundant in alloreactions such as IFN-γ, TNF-α, TGF-β didn't result in any GI in various concentrations and incubation times. Also, treatment with supernatants from different MHC non-matched Mixed Lymphocyte Cultures (MLC-SN) or with MLC separated from HaCaT cells with a semipermeable membrane didn't induce any significant GI. In contrast, exposure of HaCaT cells to the whole MLC which includes activated cellular components (measured as BrdU+) resulted in significant induction of GI in 6 out of 8 different cases (mean 2.4-fold). GI was also evaluated in terms of DNA strand break formation by using the neutral comet assay. Treatment of HaCaT cells with whole MLC resulted in significant increase of DNA strand breaks as compared to HaCaT cells exposed to MLC-SN (p=0.039) or untreated controls (p=0.048). Western blot analysis revealed no significant alterations in the levels of MMR proteins upon treatment with MLC or MLC-SN. Exposure of HaCaT cells to MLC or MLC-SN resulted in increase of intracellular ROS levels as compared to untreated cells. The MLC treatment caused significantly higher increase of ROS levels than MLC-SN. The results of the present study confirm that GI of epithelial cells is a frequent phenomenon in patients after allogeneic HCT and give in vivo and in vitro support that alloantigeneic processes can be the cause of such a phenomenon. Progress in understanding DNA damage and repair after allogeneic HCT can potentially provide molecular biomarkers and therapeutic targets.