Diagnostik der minimalen Resterkrankung beim Multiplen Myelom anhand von Immunglobulin-Leichtketten-Rearrangements in zellfreier DNA

Along with the introduction of several new agents, the prognosis for patients with multiple myeloma (MM) has improved substantially in the past few decades. Nowadays, the treatment goal is not only to control the disease, but also to achieve deep responses and long survival times. A large fraction of patients achieves complete responses (CR) under currently available treatment protocols. Nonetheless, most patients eventually relapse. This might be due to per-sistence of minimal residual disease (MRD) which cannot be assessed adequately by routine workups, such as laboratory studies, imaging and evaluation of plasma cell infiltration in the bone marrow (BM). In order to allow for MRD assessment, more sensitive cytological (flow cytometry) and mo-lecular approaches, such as next generation sequencing (NGS) and ASO RQ-PCR, are needed. Recent MM-MRD studies mainly focused on evaluation of MRD in the BM. Available studies suggest that negativity for MRD in the BM is associated with significantly improved survival rates, as compared to MRD positivity. Fewer emphasis has been placed on noninvasive MM-MRD monitoring using NGS of periph-eral blood (PB) samples. Even though there are a few pilot studies available which provided evidence for the general applicability of this approach, the specific clinical utility has not yet been established and there are still a lot of unanswered questions. MM-MRD monitoring using NGS is generally carried out by targeting MM-specific immuno-globulin (Ig) gene rearrangements which are thought to be an unambiguous molecular mark-er. Recent studies often put their emphasis on analysis of Ig heavy chain (IGH) rearrange-ments. In the present work, it is evaluated whether a genomic light chain test, i.e. sequencing of Igκ (IGK) as well as Igλ (IGL) light chain rearrangements without consideration of IGH, can also be applied for disease monitoring and MRD assessment in MM. For this purpose, 130 PB samples were collected from 65 MM patients. Cell-free DNA (cfDNA) was isolated and variable regions of Ig light chains were amplified using well-validated BIOMED-2 primers. PCR products were further processed and eventually sequenced on an Illumina® MiSeq™ instrument. The obtained sequencing data were analyzed using MiXCR and VDJtools (bioinformatical software). Additionally, in order to allow for an unam-biguous identification of MM-related rearrangements, 37 genomic DNA (gDNA) tissue sam-ples (mainly FFPE BM) from 35 patients of the same study cohort were sequenced and ana-lyzed. PB samples obtained at the time of „high disease burden“ (samples taken at first diagnosis/Dx, progressive disease/PD, stable disease, minimal response or partial response before autologous stem cell transplantation) exhibited higher median and mean cfDNA concentrations than “good responders” (samples taken at CR, very good partial response/VGPR, partial response after autologous stem cell transplantation). Clonotype diversity was lower in PB samples obtained at the time of high disease burden compared to good responders. Overrepresented clonotypes, i.e. clonotypes accounting for >33.3% of the Ig light chain repertoire in their sample, were not only identified in 61.1% of Dx and 82.4% of PD samples, but also, inter alia, in 60.0% of CR samples. Based on the data from this work, criteria for identification of probable MM clones were estab-lished. In total, 35 patients of the study cohort met the criteria for potential identification of MM clones. A probable MM clone was found for 16/35 (45.7%) patients. Solely based on analysis of cfDNA samples, i.e. without consideration of FFPE/gDNA samples, probable MM clones could be identified for seven patients. All evaluable CR samples (n=5) were negative for the probable MM clones. Clonotype fre-quencies of probable MM clones in cfDNA were significantly higher during high disease bur-den and in samples with a detectable paraprotein in serum protein electrophoresis, as com-pared to good responders and samples with no detectable serum paraprotein, respectively. Fur-thermore, there was a significant correlation between (probable) MM clonotype frequencies in cfDNA and serum free light chain ratios. Tracking probable MM clonotypes over time, it was found that an increase of (probable) MM clonotype frequencies in cfDNA preceded clinical progression in some cases, as confirmed by established serological and clinical parameters. Comparing parallelly obtained cfDNA and FFPE/gDNA samples, a moderate positive correla-tion was observed regarding (probable) MM clonotype frequencies. Of note, nearly all overlap-ping >5% clonotypes in the matching samples corresponded to probable MM clones. In summary, in order to allow for reliable identification of probable MM clones, additional analysis of Dx/PD tissue samples (gDNA/FFPE) was very helpful. The results suggest that the sensitivity of the assay used in this work does not exceed the sensitivity of serological parame-ters for MRD detection. However, the method presented here might be particularly valuable for monitoring patients with non- and oligosecretory myeloma.