Identification of novel antigenic determinants of pediatric cancer for the development of targeted therapies

Background: In recent years, immunotherapy has drawn increasing attention and has revolutionized the course of cancer treatment. In the development of novel immunotherapy strategies, the step of target identification is a challenging process, because it aims at identifying robust tumor-associated antigens (TAAs) specific for the target cells and causing no off-target effects. We here propose: a) CD72 as a novel and robust TAA for acute leukemias, in particular B-cell precursor acute lymphoblastic leukemias (BCP-ALL) and acute myeloid leukemias (AML), with the purpose of using it in advanced diagnostics and as a target for novel immunotherapies; b) the identification of novel druggable targets and characterization of bone marrow (BM) niche for neuroblastoma (NB), the most frequent extracranial pediatric solid tumor, for which BM infiltration represents a negative prognostic factors. Methods: a) We profiled CD72 expression by flow cytometry (FCM) on a variety of pathological and healthy cell lines and primary samples. We included analysis on the major populations of BM and studied CD72 expression on hematopoietic stem and progenitor cells (HSPCs). CD72 expression was assessed by FCM on 538 pathological pediatric BM samples. b) We used a combination of whole-transcriptome analysis (WTA) by single-cell RNA sequencing (scRNA-seq), single-nuclear RNA sequencing (sNucRNA-seq), and FCM analysis to possibly identify novel therapeutic targets and study BM metastatic from HR-NB patients at diagnosis. Results: a) CD72 is expressed by B-lineage cells, for which we defined a precise expression pattern in B cell regeneration. Other normal BM subpopulations, skin fibroblasts, mesenchymal stromal cells, skeletal myoblasts and endothelial cells did not express the antigen. CD72 was expressed by BCP-ALL and AML cell lines and absent on T-ALL and other solid tumor cell lines. CD72 is not expressed by HSPCs. CD72 was expressed in 211/215 (98.1%) BCP-ALL cases; in all B-NHL blasts; in 79/156 (50.6%) AML cases; in 5/83 (6.0%) T-ALL; in none of the T-NHL and non-hematopoietic solid tumors. In our cohort, CD72 does not correlate with specific BCP-ALL subtypes as classified according to EGIL. CD72 was expressed by all the 16 BCP-ALL patients that had been previously treated with CD19-directed immunotherapies, including 7 patients that underwent CD19 loss. CD72 was expressed by three rare cases of CD19-negative BCP-ALL at diagnosis. Within our AML cohort, we did not find any significant difference in CD72 expression among AML genetic subgroups. b) Bioinformatics tools allowed to distinguish among NB infiltrating cells and BM different subpopulations in four samples analyzed by scRNA-seq. NB cells were further study to identify genes overexpressed on the cell membrane and obtain a final list of 32 potential TAA. We developed a bioinformatic pipeline for the analysis of one sample by sNucRNA-seq, thanks to which we identified NB infiltrating population and 29 additional markers. We performed FCM on 13 patients, either metastatic or nonmetastatic for the BM. We studied NB cells immunophenotype and annotated cell abundance of different BM compartments, identifying regulatory T and PD-1+ cells as the most significant populations differentially present between BM-NB samples and healthy controls. Conclusions: The here reported data highlight: a) the usefulness of CD72 in advanced diagnostics, being able to recognize blasts in case of CD19 loss, and in the development of novel immunotherapy approaches which may target both BCP-ALL and AML; b) preliminary results of a characterization of BM-NB metastatic niche. In the present study, we managed to identify a number of TAA candidates which will be validated to check for feasibility of new therapeutic strategies. We identified an immunosuppressive TME, for which clinical significance it will be necessary to implement more cases.