Genomic and Phenotypic Determinants of Resistance to Immunotherapies in Multiple Myeloma

Aim 1: Evaluation of cell-intrinsic mechanisms on BM.

1. Whole Genome Sequencing (WGS) and Whole Exome Sequencing (WES) will be performed on marker CD138+ purified cells to evaluate their genomic profile, and on buccal swab DNA to restrict the analysis to variants or structural abnormalities that have a clear somatic status, and are therefore specific to the tumor cells. In details: 1.1 WGS: libraries will be prepared with TruSeq™(Kit Illumina) DNA Polymerase Chain Reaction (PCR)-Free Library Preparation Kit (Illumina, San Diego, CA) from 500ng of genomic DNA, aiming for an average target insert of 300bp. Sequencing will be performed on a 150bp-paired end protocol, at a target depth of 40x for tumor samples and 30x for normal samples. 1.2 WES: libraries will be prepared with SureSelectXT Human All Exon V6 (Agilent technologies int., Santa Clara, CA) from 100ng of genomic DNA, aiming for an average target insert of 300bp. Sequencing will be performed on a 150bp-paired end protocol, aiming for a target depth of 200x for tumor samples and 100x for normal samples. 1.3 Data analysis: next generation sequencing scoring system output format files (*.FASTQ files) will be aligned to the reference genome using Burrows-Wheeler Alignment Tool (BWAmem), and deduplicated aligned Binary Alignment Map (BAM) files will be analyzed using the following published tools available at the Wellcome Trust Sanger Institute (WTSI):
2. accurate genome-wide allele-specific copy number (ASCAT) and Battenberg for clonal and subclonal copy number changes.
3. BRASS for structural variations (large inversions and deletions, translocations, internal tandem duplication).
4. Caveman and Pindel for Single Nucleotide Variants (SNVs) and small insertion-deletions (indels). The clonal composition of the sample and the genomic evolution of myeloma over time will be inferred from the adjusted cancer cell fraction of the variants identified, clustered and analyzed using a hierarchical bayesian Dirichlet process. The mutational processes operative at various phases of MM will be analyzed using a Non-Negative Matrix Factorization (NNMF) approach to extract mutational signatures from the array of substitutions in their 5' and 3' context. The possible driver mutation role of all extracted missense mutation will be evaluated by the recently published dN/dS algorithm.
5. RNA-seq on marker CD138+ purified cells to evaluate transcriptomic profile will be performed using TruSeq Stranded Messanger RiboNucleic Acid (mRNA) Library Prep Kit (Illumina, San Diego, CA) on 500 ng total RNA, followed by sequencing, aiming for 100x106 total reads per sample. DNA excision repair protein (ERCC) spike-in mix will be added to facilitate normalization of the expression levels between samples. Reads will be aligned with Tophat2 to call SNVs, indels, and detect gene fusions. Cufflinks2 will be used to profile gene expression and detect novel transcript isoforms. Overall gene transcript expression levels will be quantified using the Reads Per Kilobase Million (RPKM) metric based on uniquely mapping reads.
6. Flow cytometry analysis will be performed on BM samples to examine potential determinants of immunotherapy sensitivity/resistance and the expression of specific targets including marker Cluster of Differentiation 38 (CD38), B-cell maturation antigen (BCMA), marker Cluster of Differentiation 33 (CD33), Programmed death-ligand 1 (PDL1), and marker Cluster of Differentiation 19 (CD19) prior to treatment, at response and at relapse. We will evaluate MM percent positive cells and Mean Fluorescence Intensity (MFI) in order to monitor the antigen expression during the evolution of the disease. Receptor density will also be performed. Moreover the European cytoflow consortium of International Myeloma Foundation (EuroFlow-IMF) MM minimal residual disease (MRD) panel will be applied to monitor MRD in particular by using a multiepitope (ME) antiCD38 to detect possible determinants of resistance. Moreover this panel will allow us to monitor the phenotype evolution of the clonal population, looking in particular at the shift towards more immature cells, which has been suggested as a mechanism of resistance to bortezomib.
7. Storage of viable marker CD138-: we will evaluate distribution of marker CD38 also on marker CD138- cells and we will determine if genomic or immunophenotypic lesions responsible for resistance could be present also in the marker CD138- fraction.

Aim 2: Evaluation of cell-extrinsic mechanism on BM and PB.

1. Flow Cytometry analysis of lymphocyte subpopulations will be performed in the same BM and PB to evaluate T-cells population (marker CD38+, marker Cluster of Differentiation 4 (CD4+), marker Cluster of Differentiation 8 (CD8+), Tregs cells) and other regulatory and suppressive immune populations like Myeloid-Derived Suppressor Cells (MDSCs). T cells will be measured at baseline, response and at relapse, and they could help in evaluating the interaction between B and T cell compartments in patients receiving immunotherapies. Moreover, evaluation of some immune checkpoints on T cells at baseline and post-treatment will be performed.
2. RNA-seq of lymphocyte subpopulations will be performed to identify the frequencies of the various helper and effector lymphocyte populations, and correlate those with response to treatment or lack thereof. We will also perform RNA-seq on different marker CD4+ T-lymphocytic subpopulations in responsive and non-responsive MM to identify potential suppressive signatures in the latter group. The same lymphocyte subpopulations will be analyzed in BM samples from 10 healthy subjects (BM biopsy in lymphoma negative staging).
3. Measurement of a broad spectrum of cytokines produced and secreted by MM and other cells within the BM microenvironment: cytokines and chemokines related to MM bone and microenvironment will be measured in BM and PB plasma from patients at each time-point. The laboratory assays will be performed by using an enzyme-linked immunosorbent assay (ELISA) kit.
4. BM biopsy: bone biopsies from MM patients at baseline and after Monoclonal Antibody Therapy (mAbs therapy) will be evaluated for the expression of suppressive molecules such as, marker Cluster of Differentiation 80/86 (CD80/86), marker Cluster of Differentiation 40 (CD40) in the tumor cells and in the BM lymphoid population by immunohistochemistry.

Aim 3: After comprehensively characterizing the genomic, transcriptomic and immunophenotypic features of CD138+ cells, and having a clear picture of the effector/suppressive immune population in MM, we will then correlate these features with clinical data. In detail, we will create a database including the following columns:

• Baseline clinical characteristics
• Prognostic factors: International Staging System (ISS), Revised International Staging System (R-ISS), Lactate Dehydrogenase (LDH), cytogenetic analysis by Fluorescence In Situ Hybridization test (FISH)
• Prior therapies and relevant clinical results
• Best response: responses will be defined according to the International Uniform Response Criteria. Responders are defined as subjects with at least a VGPR.
• Duration of Response (DOR), Progression Free Survival (PFS), Overall survival (OS) and Time to Progression (TTP) data In this aim, we will look for correlation between biological features and disease response or lack thereof, to understand which cell-intrinsic and cell-extrinsic features are better predictors of response. Because of the time needed before disease response can be assessed, this analysis will be performed after at least 1 year of treatment or at earlier progression. In a first analysis, biological features will be associated with best response to treatment, PFS and other baseline clinical, prognostic and treatment variables using linear models. Subsequently, starting from year 3, when enough follow-up will allow a meaningful analysis of PFS and OS, Kaplan-Meier and Coxregression models will be fitted to identify possible independent prognostic factors.

Although the relatively small size of the cohort will limit statistical power and the possibility to perform subgroup analysis, this attempt to identify biomarkers could improve the clinical management of the patient, by prioritizing the vast array of salvage treatments in MM and thus decreasing costs.