PURPOSE Before anaplastic lymphoma kinase (ALK) inhibitors, treatment options for ALK-positive inflammatory myofibroblastic tumors (AP-IMTs) were unsatisfactory. We retrospectively analyzed the outcome of patients with AP-IMT treated with crizotinib to document response, toxicity, survival, and features associated with relapse. METHODS The cohort comprised eight patients with AP-IMT treated with crizotinib and surgery. Outcome measures were progression-free and overall survival after commencing crizotinib, treatment-related toxicities, features associated with relapse, outcome after relapse, and outcome after ceasing crizotinib. RESULTS The median follow-up after commencing crizotinib was 3 years (range, 0.9 to 5.5 years). The major toxicity was neutropenia. All patients responded to crizotinib. Five were able to discontinue therapy without recurrence (median treatment duration, 1 year; range, 0.2 to 3.0 years); one continues on crizotinib. Two critically ill patients with initial complete response experienced relapse while on therapy. Both harbored RANBP2-ALK fusions and responded to alternative ALK inhibitors; one ultimately died as a result of progressive disease, whereas the other remains alive on treatment. Progression-free and overall survival since commencement of crizotinib is 0.75 ± 0.15% and 0.83 ± 0.15%, respectively. CONCLUSION We confirm acceptable toxicity and excellent disease control in patients with AP-IMT treated with crizotinib, which may be ceased without recurrence in most. Relapses occurred in two of three patients with RANBP2-ALK translocated IMT, which suggests that such patients require additional therapy.

Radiation-induced glioma (RIG) is a highly aggressive brain cancer arising as a consequence of radiation therapy. We report a case of RIG that arose in the brain stem following treatment for paediatric medulloblastoma, and the development and characterisation of a matched orthotopic patient-derived xenograft (PDX) model (TK-RIG915). Patient and PDX tumours were analysed using DNA methylation profiling, whole genome sequencing (WGS) and RNA sequencing. While initially thought to be a diffuse intrinsic pontine glioma (DIPG) based on disease location, results from methylation profiling and WGS were not consistent with this diagnosis. Furthermore, clustering analyses based on RNA expression suggested the tumours were distinct from primary DIPG. Additional gene expression analysis demonstrated concordance with a published RIG expression profile. Multiple genetic alterations that enhance PI3K/AKT and Ras/Raf/MEK/ERK signalling were discovered in TK-RIG915 including an activating mutation in PIK3CA, upregulation of PDGFRA and AKT2, inactivating mutations in NF1, and a gain-of-function mutation in PTPN11. Additionally, deletion of CDKN2A/B, increased IDH1 expression, and decreased ARID1A expression were observed. Detection of phosphorylated S6, 4EBP1 and ERK via immunohistochemistry confirmed PI3K pathway and ERK activation. Here, we report one of the first PDX models for RIG, which recapitulates the patient disease and is molecularly distinct from primary brain stem glioma. Genetic interrogation of this model has enabled the identification of potential therapeutic vulnerabilities in this currently incurable disease.

We describe a case of medulloblastoma maturating into ganglioglioma during therapy. A 10-month-old boy was diagnosed with a desmoplastic medulloblastoma and was treated with gross total resection followed by induction chemotherapy. A recurrence in the tumor bed during therapy was managed with focal radiation therapy and consolidation chemotherapy. After further progression, the recurrent tumor was resected completely. The histopathology revealed a benign ganglioglioma with no residual medulloblastoma. This case raises the possibility that a malignant medulloblastoma can differentiate into a benign tumor and suggests that differentiation therapy may have value in the treatment of medulloblastoma.