Subependymal giant cell astrocytoma harboring a PRRC2B‐ALK fusion: A case report

Abstract

To the Editor: Subependymal giant cell astrocytoma (SEGA) is a WHO grade 1 tumor with a reported 5-year overall survival rate of >90%.1 However, unresectable recurrences, metastasis, and SEGA-associated death have also been reported.1–4 Typically, SEGA develops in children with tuberous sclerosis complex (TSC), which is characterized by manifestations involving the skin, brain, kidney, and retina.1,5 The pathophysiology of TSC is attributed to the altered function of hamartin and tuberin encoded by TSC1 and TSC2, respectively, and subsequent activation of the mTOR pathway.5 The mTOR inhibitor is a highly effective targeted therapy for SEGA.6,7 Here, we report a case of ventricular SEGA with no TSC1/TSC2mutations, but a PRRC2B-ALK fusion was identified. A previously healthy 3-year-old male with an unremarkable family medical history was referred to our hospital because of an intraventricular mass incidentally detected by computed tomography, which was ordered after a minor head injury. Magnetic resonance imaging showed a heterogeneously enhanced, well-demarcated mass in the left lateral ventricle (1A-E). The patient underwent gross total resection of the lesion. Microscopically, the tumor comprised large polygonal cells resembling gemistocytic astrocytes (Figure 1A). Spindle cells in a fibrillary background and an ependymoma-like lesion with vague perivascular pseudorosettes were also observed (Figures 1B and 1C). Immunohistochemically, the tumor cells were positive for both glial markers (GFAP, S100 protein, and nestin) (Figure 1D) and neuronal markers (synaptophysin and TUJ-1) (Figure 1E), as well as SOX2. The MIB-1 labeling index was 3% (Figure 1F). The case was reviewed by independent central pathologists, and the patient was diagnosed with SEGA. The patient showed no dermatological or neurological signs associated with TSC. The results of an abdominal ultrasound and of examinations by an ophthalmologist and a cardiologist were also unremarkable. The patient has been in complete remission for 3 years post surgery. Targeted sequencing of all coding exons for 93 brain tumor-related genes, shown in Table S1, revealed no significant pathogenic variant in TSC1 and TSC2s. However, RNA sequencing revealed a PRRC2BALK fusion (Figure 1G). Existence of the ALK fusion was validated by immunohistochemistry (clone 5A4, Abcam, Cambridge, UK) and by fluorescence in situ hybridization (Vysis ALK Break Apart FISH Probe, AbbottMolecular, Abbott Park, IL) (Figures 1H and 1I). Methylation analysis was carried out by using the Heidelberg classifier available at https://www.molecularneuropathology.org/mnp. The sample was classified as “glioblastoma, IDH wild type, subclass mesenchymal,” for which typically adult hemispheric glioblastomas are classified.8 However, the calibrated score was 0.52 (only a score ≥0.9 is considered a match) and the result was interpreted as nomatch. Approximately 60-80% of SEGA occur in association with germline TSC1/TSC2 pathogenic variants, including single-nucleotide variants, insertions, and deletions.9,10 However, TSC1/TSC2 germline alterations may not be detected in the remaining patients.9,10 This may be due in part to the limited sensitivity of conventional detection methods. With the advent of next-generation sequencing, the detection rate of TSC1/TSC2 variants, including mosaicism, has improved; however, cases remain in which no variants are found.10–12 Martin et al reported that somatic or germline TSC1/TSC2 variants were not detected in two out of 66 patients with TSC when using whole-exome sequencing, single-nucleotide polymorphism arrays, and targeted deep sequencing.12 Furthermore, some patients developed only one TSC-associated manifestation, that is, sporadic SEGA or sporadic lymphangiomyomatosis.4,13,14 Somatic TSC1/TSC2 variants were detected in some but not all of these cases. ALK fusions are well-known therapeutic targets detected in various types of cancer.15–21 In the present case, as observed in other recurrent ALK fusions, the kinase domain of ALK and the coiled-coil domain of its fusion partner, PRRC2B,were retained.15 Oligomerization via the coiled-coil domain leads to constitutive kinase activation. Therefore, the PRRC2B-ALK fusion may play a key oncogenic role in this particular SEGA tumor. Thus, our case suggested that ALK alterations may be present on rare occasions of TSC1/TSC2-wildtype SEGA, for which an ALK inhibitor may be an effective targeted therapy. In fact, five brain tumors of various types with ALK fusions have been reported, four of whichwere hemispheric gliomas diagnosed in children less than 1 year old (age data were unavailable for the remaining case), whereas the present case of intraventricular SEGAwas diagnosedwhen the patient was 3 years old.18–22 This case emphasizes the importance of investigating targetable alterations in pediatric brain tumors with particular histological subtypes, especially when typical genetic signature is absent. In summary, we reported a case of SEGA with PRRC2B-ALK fusion. Although the current treatment of choice is generally an mTOR inhibitor, potentially targetable gene fusions may be detectable by RNA sequencing in a few cases.