Combined use of multiparametric flow cytometry and cytomorphology to enhance detection of neuroblastoma metastatic cells in bone marrow

A Multi-Color Flow Cytometric Assay for Quantifying Dinutuximab Binding to Neuroblastoma Cells in Tumor, Bone Marrow, and Blood

GD2, a disialoganglioside, is present on the surface of most neuroblastomas, as well as on some other cancers, such as melanoma and osteogenic sarcoma. The anti-GD2 antibody ch14.18 (dinutuximab) has an FDA-registered indication for use as maintenance therapy for high-risk neuroblastoma with cytokines and 13-cis-retinoic acid after myeloablative therapy. Recent studies using immunohistochemistry of tumor or tumor cells in marrow have shown that some neuroblastomas are negative for GD2. Dinutuximab and other anti-GD2 antibodies are increasingly used in combination with cytotoxic chemotherapy for treating relapsed neuroblastoma, so it is important to be able to identify patients with tumor cells with low GD2 expression, as such patients may experience toxicity but not benefit from the antibody therapy. As the most common clinical samples available for relapsed neuroblastoma are bone marrow aspirates, we developed a method to quantify dinutuximab binding density and the frequency of neuroblastoma cells positive for the antibody in bone marrow aspirates. Here, we describe a multi-color flow cytometry assay that employs non-GD2 antibodies to identify neuroblastoma cells in a mixed population (tumor, bone marrow, or blood) and an anti-GD2 antibody to quantify both the frequency and density of GD2 expression on neuroblastoma cells.

Pro-metastatic and mesenchymal gene expression signatures characterize circulating tumor cells of neuroblastoma patients with bone marrow metastases and relapse

Existing marker-based methods of minimal residual disease (MRD) determination in neuroblastoma do not effectively enrich for the circulating disease cell population. Given the relative size differential of neuroblastoma tumor cells over normal hematogenous cells, we hypothesized that cell size-based separation could enrich circulating tumor cells (CTCs) from blood samples and disseminated tumor cells (DTCs) from bone marrow aspirates (BMA) of neuroblastoma patients, and that their gene expression profiles could vary dynamically with various disease states over the course of treatment. Using a spiral microfluidic chip, peripheral blood of 17 neuroblastoma patients at 3 serial treatment timepoints (diagnosis, n=17; post-chemotherapy, n=11; and relapse, n=3), and bone marrow samples at diagnosis were enriched for large intact circulating cells. Profiling the resulting enriched samples with immunohistochemistry and mRNA expression of 1490 cancer-related genes via NanoString, 13 of 17 samples contained CTCs displaying cytologic atypia, TH and PHOX2B expression and/or upregulation of cancer-associated genes. Gene signatures reflecting pro-metastatic processes and the neuroblastoma mesenchymal super-enhancer state were consistently upregulated in 7 of 13 samples, 6 of which also had metastatic high-risk disease. Expression of 8 genes associated with PI3K and GCPR signaling were significantly upregulated in CTCs of patients with bone marrow metastases versus patients without. Correspondingly, in patients with marrow metastases, differentially-expressed gene signatures reflected upregulation of immune regulation in bone marrow DTCs versus paired CTCs samples. In patients who later developed disease relapse, 5 genes involved in immune cell regulation, JAK/STAT signaling and the neuroblastoma mesenchymal super-enhancer state (OLFML2B, STAT1, ARHGDIB, STAB1, TLR2) were upregulated in serial CTC samples over their disease course, despite urinary catecholamines and bone marrow aspirates not indicating the disease recurrences. In summary, using a label-free cell size-based separation method, we enriched and characterized intact circulating cells in peripheral blood indicative of neuroblastoma CTCs, as well as their DTC counterparts in the bone marrow. Expression profiles of pro-metastatic genes in CTCs correlated with the presence of bone marrow metastases at diagnosis, while longitudinal profiling identified persistently elevated expression of genes in CTCs that may serve as novel predictive markers of hematogenous MRD in neuroblastoma patients that subsequently relapse.

Multiparametric flow cytometry highlights B7-H3 as a novel diagnostic/therapeutic target in GD2neg/low neuroblastoma variants

Background

High-risk neuroblastomas (HR-NBs) are rare, aggressive pediatric cancers characterized by resistance to therapy and relapse in more than 30% of cases, despite using an aggressive therapeutic protocol including targeting of GD2. The mechanisms responsible for therapy resistance are unclear and might include the presence of GD2neg/low NB variants and/or the expression of immune checkpoint ligands such as B7-H3.

Method

Here, we describe a multiparametric flow cytometry (MFC) combining the acquisition of 10⁶ nucleated singlets, Syto16pos CD45neg CD56pos cells, and the analysis of GD2 and B7-H3 surface expression. 41 bone marrow (BM) aspirates from 25 patients with NB, at the onset or relapse, are analyzed, comparing results with cytomorphological analysis (CA) and/or immunohistochemistry (IHC). Spike in experiments assesses the sensitivity of MFC. Kaplan-Meier analysis on 498 primary NBs selects novel prognostic markers possibly integrating the MFC panel.

Results

No false positive are detected, and MFC shows high sensitivity (0.0005%). Optimized MFC identifies CD45negCD56pos NB cells in 11 out of 12 (91.6%) of BM indicated as infiltrated by CA, 7 of which coexpress high levels of GD2 and B7-H3. MFC detects CD45negCD56posGD2neg/low NB variants expressing high surface levels of B7-H3 in two patients with HR-NB (stage M) diagnosed at 53 and 139 months of age. One of them has a non-MYCN amplified tumor with unusual THpos PHOX2Bneg phenotype, which relapsed 141 months post-diagnosis with BM infiltration and a humerus lesion. All GD2neg/low NB variants are detected in patients at relapse. Kaplan-Meier analysis highlights an interesting dichotomous prognostic value of MML5, ULBPs, PVR, B7-H6, and CD47, ligands involved in NB recognition by the immune system.

Conclusions

Our study validates a sensitive MFC analysis providing information on GD2 and B7-H3 surface expression and allowing fast, specific and sensitive evaluation of BM tumor burden. With other routinely used diagnostic and prognostic tools, MFC can improve diagnosis, prognosis, orienting novel personalized treatments in patients with GD2low/neg NB, who might benefit from innovative therapies combining B7-H3 targeting.

Disialoganglioside GD2 Expression in Solid Tumors and Role as a Target for Cancer Therapy

Gangliosides are carbohydrate-containing sphingolipids that are widely expressed in normal tissues, making most subtypes unsuitable as targets for cancer therapy. However, the disialoganglioside GD2 subtype has limited expression in normal tissues but is overexpressed across a wide range of tumors. Disialoganglioside GD2 can be considered a tumor-associated antigen and well-suited as a target for cancer therapy. Disialoganglioside GD2 is implicated in tumor development and malignant phenotypes through enhanced cell proliferation, motility, migration, adhesion, and invasion, depending on the tumor type. This provides a rationale for targeting disialoganglioside GD2 in cancer therapy with the development of anti-GD2 monoclonal antibodies and other therapeutic approaches. Anti-GD2 monoclonal antibodies target GD2-expressing tumor cells, leading to phagocytosis and destruction by means of antibody-dependent cell-mediated cytotoxicity, lysis by complement-dependent cytotoxicity, and apoptosis and necrosis through direct induction of cell death. Anti-GD2 monoclonal antibodies may also prevent homing and adhesion of circulating malignant cells to the extracellular matrix. Disialoganglioside GD2 is highly expressed by almost all neuroblastomas, by most melanomas and retinoblastomas, and by many Ewing sarcomas and, to a more variable degree, by small cell lung cancer, gliomas, osteosarcomas, and soft tissue sarcomas. Successful treatment of disialoganglioside GD2-expressing tumors with anti-GD2 monoclonal antibodies is hindered by pharmacologic factors such as insufficient antibody affinity to mediate antibody-dependent cell-mediated cytotoxicity, inadequate penetration of antibody into the tumor microenvironment, and toxicity related to disialoganglioside GD2 expression by normal tissues such as peripheral sensory nerve fibers. Nonetheless, anti-GD2 monoclonal antibody dinutuximab (ch14.18) has been approved by the U.S. Food and Drug Administration and dinutuximab beta (ch14.18/CHO) has been approved by the European Medicines Agency for the treatment of high-risk neuroblastoma in pediatric patients. Clinical trials of anti-GD2 therapy are currently ongoing in patients with other types of disialoganglioside GD2-expressing tumors as well as neuroblastoma. In addition to anti-GD2 monoclonal antibodies, anti-GD2 therapeutic approaches include chimeric antigen receptor T-cell therapy, disialoganglioside GD2 vaccines, immunocytokines, immunotoxins, antibody-drug conjugates, radiolabeled antibodies, targeted nanoparticles, and T-cell engaging bispecific antibodies. Clinical trials should clarify further the potential of anti-GD2 therapy for disialoganglioside GD2-expressing malignant tumors.