Evidence of Graft-versus-Tumor Effect in Refractory Metastatic Neuroblastoma

Abstract

Listen

Translate article

There are several reports describing graft-versus-tumor (GVT) effects in patients with solid tumors (1). Recently, GVT effects against advanced pancreatic cancer were added in this journal (2). To our knowledge, GVT effects in patients with high-risk neuroblastoma have never been reported and it has been believed that the effects are questionable in neuroblastoma (3). Here, we report eradication of tumor cells following the induction of acute graft-versus-host disease (GVHD) in a child with refractory metastatic neuroblastoma. The clearance and early detection of minimal residual disease (MRD) were confirmed by flow cytometric analysis of CD9+CD56+CD45− bone marrow cells, a population we previously showed that is consistent with neuroblastoma cells (4). A 7-year-old girl had a history of disseminated stage IV neuroblastoma with N-myc amplification. She had a recurrence of the tumor with the elevation of urinary vanillylmandelic acid two years after tandem highdose chemotherapy with autologous peripheral blood stem cell rescue. Magnetic resonance imaging showed bilateral multiple massive paraaortic metastases. I-123-meta-iodobenzylguanidine (MIBG) scan revealed to metastases of generalized lymph nodes and bone marrow, and abdominal mass. Bone marrow aspiration revealed presence of CD9+CD56+CD45− neuroblastoma cells (Fig. 1, B–i). Clustering of neuroblastoma cells in bone marrow was confirmed by immunohistochemistry using anti-tyrosine hydroxylase monoclonal antibody (Fig. 1, C–i) (5). Three courses of conventional chemotherapy and topotecan-based chemotherapy were relatively effective, reducing the tumor mass size by 30 percent (Fig. 1A). She received allogeneic peripheral blood stem cell transplant from an HLA-compatible sister after total surgical resection of the mass and intraoperative 20 Gy irradiation of tumor bed to achieve complete remission (Fig. 1, B-ii and C-ii). On day +40 the minimal residual tumor in bone marrow reappeared (Fig. 1, B-iii and C-iii), suggesting a recurrence of neuroblastoma, although urinary vanillylmandelic acid, magnetic resonance imaging, and MIBG scan were not positive. At least two slides were examined. It was reported that the sensitivity of morphological and histologic analysis is approximately 1 per 103–4 cells (6). Flow cytometric analysis could detect occult neuroblastoma cells at a level of 1 per 104–5 cells (4).FIGURE 1.: Induction of a graft-versus-neuroblastoma effect. (A) A patient with recurrent refractory neuroblastoma received allogeneic peripheral blood stem cell transplantation. A graft-vs.-neuroblastoma effect was induced by discontinuing cyclosporine A. (B) The expression of CD9 antigen on gated CD56+/CD45− is shown in histograms. The percentage of CD9+CD56+CD45− cells detected in bone marrow pooled samples at the recurrence (i), before transplantation (ii), day 40 after transplantation (iii) and two weeks after withdrawing cyclosporine A (iv). Total number of cells analyzed was 4x104 in (i) and 2×106 in (ii)-(iv). (C) Immunohistochemistry of bone marrow with anti-tyrosine hydroxylase (TH) monoclonal antibody provided the detection of neuroblastoma cells. Typical neuroblastoma cell clustering is shown in (i) and a TH-positive cell with a round nucleus is shown in (iii; arrow). Negative staining is confirmed in (ii) and (iv) taken from at least two sites of bone marrow. CsA, cyclosporine A; VMA, vanillylmandelic acid; HVA, homovanillic acid.Resolution of bone marrow metastasis was observed in association with the development of grade III acute GVHD two weeks after withdrawing cyclosporine A (Fig. 1, B-iv and C-iv). Acute GVHD was successfully treated and she remained free of MRD for 30 months after transplantation, when she relapsed with left tibia bone involvement. Bone marrow involvement was not observed. At this time, she was placed on tacrolimus and prednisolone for extensive type of chronic GVHD. Our observation here is a first reported definite evidence of graft-versus-neuroblastoma effect, which was confirmed by disappearance of tumor cells after the induction of acute GVHD. Target structure in GVT effects is unknown. However, minor histocompatibility antigen may be a possible target. Clinical experience on a large scale would be required to determine the clinical efficacy of GVT effects in patients with neuroblastoma. ACKNOWLEDGMENTS This work was supported by a Research Grant on Human Genome, Tissue Engineering (H17-014) from the Japanese Ministry of Health, Labor & Welfare. Masahiro Hirayama Eiichi Azuma Mariko Araki Yoshihiro Komada Department of Pediatrics and Cell Transplantation Mie University School of Medicine Mie, Japan Atsuko Nakagawa Division of Pathology Department of Clinical Laboratory Medicine National Center for Child Health and Development Tokyo, Japan