Recurrent somatic alterations of FGFR1 and NTRK2 in pilocytic astrocytoma

PMID: 23817572 · DOI: 10.1038/ng.2682 · Journal: Nature Genetics (2013)

TL;DR

The ICGC PedBrain Tumor Project performed whole-genome sequencing (with matched RNA-seq in 73 cases and long-insert mate-pair sequencing in 68 cases) of 96 pediatric pilocytic astrocytomas. They found an extremely low mutation rate (<0.1/Mb) and confirmed that pilocytic astrocytoma is a single-pathway disease: 100% of tumors harbored a MAPK pathway alteration. Beyond known KIAA1549:BRAF fusions, they identified four novel BRAF fusion partners (RNF130, CLCN6, MKRN1, GNAI1), a novel BRAF p.R506_insVLR insertion, two new NTRK2 fusions (QKI:NTRK2 and NACC2:NTRK2) in non-cerebellar tumors, recurrent hotspot mutations in FGFR1 (N546 and K656; 5 tumors in the WGS cohort, 9 more in a screening cohort of 45 non-cerebellar tumors), and recurrent PTPN11 mutations (E69K, E76A) in FGFR1-mutant tumors. The same FGFR1/H3-3A K27M/NF1 triple-mutation constellation was identified in 3/48 pediatric glioblastomas, suggesting FGFR1 signaling is a therapeutic target in both diseases.

Cohort & data

• 96 pilocytic astrocytomas (PAST) with matched blood DNA subjected to whole-genome sequencing on Illumina HiSeq2000.
• Matched RNA sequencing (n=73) and long-insert mate-pair sequencing (n=68) on the same samples for structural variant detection.
• Tumor locations spanned the CNS (~half cerebellar, ~half non-cerebellar/extra-cerebellar/midline); cohort identifier: past_dkfz_heidelberg_2013.
• Additional independent screening cohort of 45 non-cerebellar pilocytic astrocytomas (negative for KIAA1549:BRAF) screened for FGFR1 exons 12 & 14 and PTPN11 exon 3.
• Comparison cohort: re-examination of previous pediatric GB exome sequencing data (48 samples).
• Gene expression array data from 118 pilocytic astrocytomas (Affymetrix U133 Plus2.0, including 66 from the present series) vs. 158 other astrocytic tumors.
• Methods: whole-genome-seq, rna-seq, mate-pair sequencing, BWA + samtools + Picard alignment to hg19, pindel, CREST, DELLY for SV detection, TopHat-Fusion and deFuse for fusion calling, genome-music for significantly mutated genes, annovar + oncotator for annotation, PCR + sanger-sequencing for verification (>98% SNV verification, >70% InDel verification).
• Accession: European Genome-phenome Archive EGAS00001000381.

Key findings

• 100% of 96 pilocytic astrocytomas harbored a MAPK pathway alteration; pilocytic astrocytoma is confirmed as a prototypic single-pathway disease.
• Mean somatic mutation rate <0.1/Mb; mean 1.6 non-synonymous SNVs per tumor (range 0–9); <1 coding InDel per case.
• Genome-wide mutation rate positively correlated with patient age (Pearson r = 0.42, P = 2.3 × 10⁻⁵), driven by C-to-T transitions at CpG sites.
• 70/96 cases harbored a KIAA1549:BRAF fusion; one FAM131B:BRAF fusion; four BRAF V600E mutations; one BRAF p.ins599T.
• Four novel BRAF fusions identified: RNF130:BRAF (n=2, from reciprocal t(5;7)(q35;q34)), CLCN6:BRAF, MKRN1:BRAF, and GNAI1:BRAF (each n=1); all variants delete the BRAF amino-terminal regulatory region.
• Novel BRAF p.R506_insVLR (3-aa insertion in the interdomain cleft) in ICGC_PA65 — stabilizes a dimeric form of BRAF, confirmed by co-immunoprecipitation, and increased ERK phosphorylation as effectively as BRAF V600E.
• Two cases (ICGC_PA117, ICGC_PA142) harbored compound KRAS mutations on the same allele (E63K/R73M and L19F/Q22K respectively) — non-hotspot residues, distinct from classical codon 12/13/61 hotspots.
• All but one cerebellar tumor in the series harbored a BRAF fusion (the exception had a KRAS alteration); 9/48 (19%) non-cerebellar tumors lacked any classical MAPK alteration on initial screening.
• Two novel NTRK2 fusions identified in 3 non-cerebellar samples: QKI:NTRK2 and NACC2:NTRK2; both 5’ partners contain dimerization domains, predicted to drive ligand-independent NTRK2 (TrkB) dimerization. Verified by PCR.
• Recurrent FGFR1 kinase-domain hotspot mutations (N546, K656) in 5/96 WGS tumors plus 9/45 in the screening cohort (14 total).
• One additional FGFR1 alteration (ICGC_PA89): a ~4.5 kb internal tandem duplication (ITD) of the portion encoding the kinase domain, reminiscent of FLT3 ITDs in AML.
• Strong, diffuse phospho-FGFR1 immunohistochemistry in 7/7 FGFR1-mutant tumors vs. 0/11 FGFR1-WT tumors; all samples were strongly phospho-ERK positive.
• Recurrent PTPN11 (SHP-2) mutations E69K and E76A — both in FGFR1-mutant tumors (ICGC_PA84 and ICGC_PA166); one additional E69K in the screening cohort, again in an FGFR1-mutant case.
• In vitro: PTPN11 mutant alone did not elevate pERK; FGFR1 mutants alone or with mutant PTPN11 upregulated pERK — supporting PTPN11 as a modifier of FGFR1-driven activation.
• FGF2 and PTPN11 (SHP-2) expression were significantly elevated in pilocytic astrocytoma vs. 158 other astrocytomas or normal tissues — present in both FGFR1-mutant and WT tumors, suggesting a broader ligand-mediated activation role.
• BRAF, FGFR1, KRAS, and NF1 were the only genes significantly mutated by Genome MuSiC (max-FDR 0.05); except for FGFR1+PTPN11 co-occurrence, alterations were mutually exclusive (permutation test P < 0.0001).
• All FGFR1-mutant pilocytic astrocytomas were extra-cerebellar, mostly midline — overlapping the anatomic distribution of H3-3A K27M-mutant gliomas.
• Cross-disease finding: 3/48 (6%) pediatric GB samples from a prior exome cohort harbored the same triple constellation of FGFR1 activation + H3-3A K27M + somatic NF1 alteration; all three were TP53-WT. One previously misclassified medulloblastoma case carried H3-3A K27M + NF1 + FGFR2 K659E (homologous to FGFR1 K656E), totaling 5 triple-mutant cases.

Genes & alterations

• BRAF — activating fusions (KIAA1549:BRAF in 70/96; novel partners RNF130 n=2, CLCN6, MKRN1, GNAI1; plus FAM131B:BRAF), V600E in 4 tumors, p.ins599T in 1, and novel p.R506_insVLR (interdomain-cleft insertion that stabilizes dimerization and matches BRAF V600E for ERK phosphorylation).
• FGFR1 — recurrent kinase-domain hotspot mutations at N546 and K656 (5/96 WGS + 9/45 screening cohort); plus a ~4.5 kb kinase-domain internal tandem duplication (ICGC_PA89). Mutants drive ERK phosphorylation in NIH3T3. Phospho-FGFR1 IHC positive in all FGFR1-mutant tumors examined. All FGFR1-mutant tumors were extra-cerebellar, mostly midline. Same FGFR1 mutations recurred in 3/48 pediatric glioblastomas in a published exome cohort.
• NTRK2 — two novel kinase-domain-retaining fusions in 3 non-cerebellar samples: QKI:NTRK2 and NACC2:NTRK2; 5’ partners contribute dimerization domains, predicted to confer ligand-independent activation.
• KRAS — non-hotspot compound mutations on the same allele in 2 cases (E63K/R73M; L19F/Q22K) — distinct from classical codon 12/13/61 hotspots; one cerebellar tumor lacking a BRAF fusion harbored a KRAS alteration.
• PTPN11 — E69K and E76A activating mutations (previously described in juvenile myelomonocytic leukemia), exclusively co-occurring with FGFR1 mutations; PTPN11/SHP-2 expression also globally elevated in pilocytic astrocytoma.
• NF1 — germline + somatic second hit in 3 NF1-associated cases (classical tumor-suppressor model); somatic NF1 alterations in pediatric glioblastomas with FGFR1+H3-3A K27M.
• H3-3A (formerly H3F3A) — K27M mutation co-occurring with FGFR1 activation and NF1 alteration in one ambiguous pilocytic astrocytoma/GB case (ICGC_PA69) and in 3/48 pediatric GB samples.
• FGF2 — significantly increased expression in pilocytic astrocytoma vs. other astrocytomas/normal tissues; not restricted to FGFR1 status, suggesting ligand-mediated MAPK activation across the entity.
• KIAA1549 — most frequent 5’ fusion partner for BRAF in pilocytic astrocytoma (70/96 cases).
• FAM131B — alternative 5’ fusion partner for BRAF (1 case).
• RNF130, CLCN6, MKRN1, GNAI1 — newly identified 5’ fusion partners for BRAF.
• QKI, NACC2 — newly identified 5’ fusion partners for NTRK2; both contribute dimerization domains.
• TP53 — wild-type in all triple-mutant FGFR1/H3-3A K27M/NF1 pediatric GB cases (distinct from most K27M-mutant GBM/DIPG, which typically carry TP53 mutations).
• FGFR2 — K659E mutation (homologous to FGFR1 K656E) observed in a triple-mutant case originally classified as medulloblastoma but reclassified by expression as glioblastoma.
• FGFR3, TACC1, TACC3, RAF1, NTRK1, NTRK3, FLT3 — discussed as background context (homologous FGFR3:TACC3/FGFR1:TACC1 fusions in adult GB; analogous NTRK1/NTRK3 fusions in other cancers; RAF1 fusions in pilocytic astrocytoma; FLT3 ITDs in AML serve as the conceptual parallel for the FGFR1 kinase-domain ITD).

Clinical implications

• Identifies FGFR1, NTRK2, and MEK inhibition as rational targets for the ~5–20% of non-cerebellar pilocytic astrocytomas lacking KIAA1549:BRAF, where surgical resection is often impossible and tumors progress chronically.
• BRAF V600E-specific agents are a logical option for ~5% of pilocytic astrocytomas with V600E. The authors caution against non-V600E-selective BRAF inhibitors in KIAA1549:BRAF-fusion tumors (the majority) because of preclinical evidence of paradoxical MAPK activation in that context.
• Combination strategies (FGFR + NTRK2 + MEK inhibitors) may be warranted given that downstream NTRK2 signaling converges on the MAPK pathway.
• The discovery of the same FGFR1 hotspot mutations in 3/48 pediatric GB with H3-3A K27M and NF1 loss extends a potential FGFR-targeted therapeutic axis to a clinically devastating tumor subset.
• Diagnostic implication: phospho-FGFR1 IHC reliably stratifies FGFR1-mutant vs. wild-type tumors and may be deployable as a clinical assay.

Limitations & open questions

• The downstream signaling and transforming potential of the two non-hotspot compound KRAS mutations (E63K/R73M; L19F/Q22K) on the same allele remain to be functionally characterized.
• ICGC_PA69 carried the triple-mutation constellation (FGFR1 + H3-3A K27M + NF1) and was histologically read as pilocytic astrocytoma by three neuropathologists, but a GB diagnosis could not be conclusively excluded due to limited material — the biological identity of triple-mutant tumors is unresolved.
• Why the FGFR1 + H3-3A K27M + NF1 mutations occur in concert, and whether they together specify a fate of oncogene-induced senescence (pilocytic astrocytoma) vs. aggressive malignancy (GB), is unknown.
• The NF1-associated cohort (3/96) is smaller than the prospective 5–10% rate, reflecting limited tissue availability from optic-pathway tumors.
• The authors do not report functional dissection of every novel BRAF fusion partner (CLCN6, MKRN1, GNAI1).
• Mate-pair coverage limited to 68/96 cases, so some structural variants may still be undetected.
• The model that NF1 loss may mimic elevated PTPN11 activity (both blocking RasGAP recruitment) is proposed but not directly tested.

Citations from this paper used in the wiki

• “We identified recurrent activating mutations in FGFR1 and PTPN11 and novel NTRK2 fusion genes in non-cerebellar tumors. New BRAF activating changes were also observed. MAPK pathway alterations affected 100% of tumors analyzed, with no other significant mutations, indicating pilocytic astrocytoma as predominantly a single-pathway disease.” (Abstract)
• “The average somatic mutation rate was extremely low (<0.1/Mb), with a mean of 1.6 non-synonymous single nucleotide variants (SNVs) per tumor (range 0–9).” (Results, p.2)
• “Most of the known MAPK pathway activating events were also found in this series, including KIAA1549:BRAF fusion variants (70 cases), a FAM131B:BRAF fusion, four BRAFV600E mutations and one BRAFins599T.” (Results, p.2)
• “An RNF130:BRAF fusion derived from a reciprocal t(5;7)(q35;q34) translocation was seen in two cases, with single examples of CLCN6:BRAF, MKRN1:BRAF and GNAI1:BRAF fusions.” (Results, p.2)
• “9/48 (19%) of the non-cerebellar tumors, however, lacked any of the above alterations. Further assessment of structural rearrangements revealed two novel gene fusions in a total of 3 samples, involving the kinase domain of NTRK2.” (Results, p.3)
• “A second new recurrent alteration, namely mutation of two hotspot residues (N546 & K656) in the kinase domain of FGFR1, was seen in five tumors.” (Results, p.3)
• “Nine cases harbored FGFR1 mutations at N546 or K656, and one additionally carried a PTPN11/SHP-2 E69K change.” (Results, p.4 — screening cohort)
• “By examining previous exome sequencing data for pediatric GBM, we identified 3/48 samples (6%) with an FGFR1 mutation. Strikingly, all three harbored the same constellation of H3F3A K27M, somatic NF1 alteration, and FGFR1 activation.” (Results, p.4)
• “Single-drug or combination therapy with FGFR, NTRK2 and/or MEK inhibitors … may therefore represent rational treatment options. BRAFV600E-specific agents may also be a logical choice for ~5% of patients.” (Discussion, p.5)
• “Emerging preclinical data suggest that BRAF inhibitors may trigger paradoxical activation in tumors harboring KIAA1549:BRAF, i.e. the majority of pilocytic astrocytomas.” (Discussion, p.5)

This page was processed by crosslinker on 2026-05-09.