Medulloblastoma (MB), tumor of the cerebellum, is a leading cause of cancer related mortality in childhood. In a mouse model of spontaneous MB, correlation between a defect of the migration of cerebellar precursor cells (GCPs) and an increased frequency of MB has been demonstrated. The chemokine Cxcl3, responsible for the inward migration of GCPs, can suppress the development of the MB lesions.
In the United States the annual incidence for medulloblastoma was reported at 6 per million children, in other words, approximately 450 new cases per year [Smoll et al., J Clin Neurosci 2012; 19: 1541]. Incidence of medulloblastoma decreases with age: from 2000 to 2013 in the United States the MB incidence was 0.55, or 0.57, or 0.32, and 0.16 per 100,000 population in children aged 0–4, 5–9, and 10–14 years, and adolescents aged 15–19 years, respectively [Khanna et al., J Neurooncol 2017; 135: 433]. In Italy, the age-standardized incidence rate of MB was 6.9 cases per million/year among males and 4.4 among females [AIRTUM Working Group et al., Epidemiol Prev 2013; 37: 1]. Males show higher incidence rate relative to females (1.5 : 1 for males), with a significant difference between the sexes also in the survival rate of patients older than 3 years [Curran et al., Pediatr Blood Cancer 2009; 52: 60]. Although MB arises mainly in childhood, about 30% of medulloblastomas occur in adults (20 years of age and older), but they only account for 1.0% of adult brain tumors [Crawford et al., Lancet Neurol 2007; 6: 1073].
After leukemias, central nervous system tumors are the prevalent pediatric cancers and represent the leading cause of cancer‐related mortality in children; among these, medulloblastoma (MB) is the most common, representing about 20% of all childhood brain tumors [Rossi et al., Clin Cancer Res 2008; 14: 971]. MB is frequently associated with an increase in intracranial pressure and its predominant symptoms are morning vomiting, headache, ataxia and nausea. An insidious feature of MB is its propensity to metastasize: at diagnosis, 11-43% of patients show dissemination in the central nervous system, particularly in the cerebrospinal fluid and in the meninges, while extraneural spread is an uncommon event, with bone, bone marrow, lymph nodes, liver, and lung involvement occurring in decreasing order of frequency [MacDonald et al., Oncologist 2003; 8: 174]. The current multimodality treatment of MB, including surgery, radiotherapy and chemotherapy, allows acceptable survival rates, but patients suffer severe side effects such as permanent neurocognitive dysfunctions and secondary malignancies. Therefore, new, more effective and less toxic therapeutic options are urgently needed.
Current Technology Limitations
Currently, the MB treatment, which includes surgical resection followed by cranio-spinal irradiation and chemotherapy, allows a 5‐year disease‐free survival rate ranging from 50% to 80%, depending on the extent of residual tumor mass after surgery, the presence of metastases and the age of the subject (less than or older than 3 years) at the time of diagnosis, prognostic parameters that distinguish high risk patients from standard risk patients. The multimodality treatment, however, causes devastating side effects such as retarded skeletal growth, endocrine dysfunction, progressive cognitive impairment, psychiatric disorders and social difficulties [Crawford et al., Lancet Neurol 2007; 6: 1073], as well as relapses and secondary malignancies [Goldstein et al., Cancer Causes Control 1997; 8: 865]. Thus, discovering new treatment options, aimed to increase the specificity for cancer cells and minimize the damage to the developing brain, is urgently needed.
Until recently, the therapeutic stratification of MB patients was carried out on the basis of the extent of resection, the presence of metastases and the histological subtype. A new, more accurate patient stratification method based on tumor gene expression profiles has lately been proposed: this method identifies 4 main subgroups (Wnt, Shh, Group 3 and Group 4), that differ in cellular origin, molecular etiology and prognosis [Hatten et al., Trends Neurosci 2011; 34: 134; Taylor et al., Acta Neuropathol 2012; 123: 465]. Our team have previously demonstrated the effectiveness of Cxcl3 treatment in murine Shh-type medulloblastoma [Ceccarelli et al., Front Pharmacol 2016; 7: 484], originating from GCPs that proliferate on the cerebellar surface and express the Cxcl3 chemokine receptor, Cxcr2, on their plasma membrane [Farioli-Vecchioli et al., J Neurosci 2012; 32: 15547]. Therefore, we expect the therapeutic treatment with Cxcl3 to be effective also in human Shh-type tumors, that have the same cellular origin. An initial possible application of Cxcl3 therapy could be in Gorlin’s syndrome, because the affected individuals develop Shh-type medulloblastoma with hereditary predisposition (frequency 1: 50,000), and therefore MB onset is more predictable and monitorable from the early stages of development. If successful, the application of the Cxcl3 therapy could be extended to all patients with Shh-type MBs.
Our Technology and solutions
We have observed that in the mouse model of spontaneous MB, Ptch1+/-/Tis21-/-, a dramatic increase of MB frequency is caused by an impairment of migration of granule precursor cells (GCPs) from the surface of cerebellum to the internal layers, during cerebellar development. We proved that this increase of MB occurs because GCPs remain longer in the external proliferative area, rather than differentiating and migrating internally, becoming potential targets of transforming insults. We identified the chemokine Cxcl3 as responsible for the inward migration of GCPs [Farioli-Vecchioli et al., J Neurosci 2012; 32: 15547] and we demonstrated that the chronic administration (for 4 weeks) of the chemokine Cxcl3 into the cerebella of the 1-month-old Ptch1+/-/Tis21-/- mice suppresses the development of MB lesions, rescuing the defect of migration of GCPs and inducing their differentiation [Ceccarelli et al., Front Pharmacol 2016; 7: 484].
The therapeutic approach against MB that we propose here – the Cxcl3 administration – would bypass the current cancer therapies targeting proliferation and survival of tumor cells, which mainly use toxic chemicals. In fact, the chemokine Cxcl3 forces the neoplastic cells at the cerebellar surface to migrate out of the proliferative area towards the internal layers and to differentiate, withdrawing from the tumor program. It is possible that Cxcl3 leads only to a reduction, and not to the disappearance, of the tumor mass: even in this case, Cxcl3 could still be used in the MB treatment as adjuvant therapy, aimed at reducing the tumor extent before surgical resection or increasing the tumor differentiation, making it more sensitive to the chemotherapeutic agents. Moreover, the control of migration of neural precursors, i.e., the underlying therapeutic mechanism of Cxcl3, is operative also in other neural tumors, thus this potential therapy may have wider application.
Key advantage of Cxcl3 is that, being a chemokine, has receptors on the surface of cerebellar cells and therefore does not need to be vehiculated inside the neoplastic cells by a virus. Furthermore, Cxcl3, as a recombinant molecule, shows no toxicity effects and is widely available from pharmacological suppliers, which facilitates its use.
Once we have confirmed the effectiveness of the Cxcl3 chemokine treatment to inhibit the development of murine MB at more advanced stages of development, the next step will be to test the possibility to use the Cxcl3 chemokine in human MB therapy. This objective initially requires the analysis of human medulloblastomas to determine if they are able to respond to treatment with Cxcl3: the tumors will be collected during surgery and, therefore, the project will be carried out in collaboration with doctors and researchers of the Bambino Gesù Pediatric Hospital in Rome. Subsequently, the explanted tumor cells will be used for the production of PDX (patient derived xenograft) and 3D cell cultures, in order to verify both in vivo and in vitro the ability of the chemokine Cxcl3 to induce migration and differentiation into neurons of the human cancer cells, causing medulloblastoma regression.
We have previously shown that intracerebellar administration of Cxcl3 is effective in the MB treatment at early tumor stages [Ceccarelli et al., Front Pharmacol 2016; 7: 484]; currently, our team is studying the effectiveness of Cxcl3 treatment at the more advanced stages of MB development. For this purpose, 3-month-old Ptch1+/-/Tis21-/- mice were subjected to chronic administration of recombinant Cxcl3 for 28 days through implantation of an Alzet osmotic minipump in the cerebellum. Preliminary results indicate that treatment with Cxcl3 compared to treatment with vehicle alone (CSF, cerebrospinal fluid) is able to promote the migration towards the internal cerebellar layers and the differentiation of neoplastic GCPs, with consequent exit from the neoplastic program and reduction of the frequency and extension of hyperplastic cerebellar lesions. These preliminary results support the hypothesis that the chemokine Cxcl3 is able to prevent the development of MB even in a more advanced stage of tumorigenesis, when all the preneoplastic lesions on the cerebellar surface are irremediably committed to a full-blown tumor [Farioli-Vecchioli et al ., J Neurosci 2012; 32: 15547].