Dual-Genotype Diffuse Low-Grade Glioma: Is It Really Time to Abandon Oligoastrocytoma As a Distinct Entity?

Journal of Neuropatholgy & Experimental Neurology Vol. 0, No. 0, 2017, pp. 1 5 doi: 10.1093/jnen/nlx024 BRIEF REPORT Dual-Genotype Diffuse Low-Grade Glioma: Is It Really Time to Abandon Oligoastrocytoma As a Distinct Entity? Valeria Barresi, MD, PhD, Simona Lionti, MD, Laura Valori, BSc, Giovanna Gallina, PhD, Maria Caffo, MD, PhD, and Sabrina Rossi, MD, PhD Abstract We report a unique case of dual-genotype oligoastrocytoma characterized by IDH2 gene mutation. The tumor was resected from the temporal lobe of a 25-year-old man. At histological examination with hematoxylin and eosin stain, it showed distinct oligodendroglial and astrocytic areas. The former retained alpha-thalassaemia/ mental retardation X-linked (ATRX) immuno-expression and had absent staining for p53, while the latter had ATRX loss and p53 over-expression. Molecular analyses were separately assessed in the 2 tumor components. Gene sequencing disclosed IDH2 mutation in both, whereas oligodendroglial, but not astrocytic areas, had 1p/19q codeletion and telomerase reverse transcriptase promoter mutation. Distinction of dual-genotype oligoastrocytoma from oligodendroglioma and astrocytoma might be clinically relevant for prognosis and therapy. Because most studies that investigated the molecular phenotype of oligoastrocytomas have focused on IDH1 R132H mutated cases, we suggest further analyses on diffuse gliomas with heterogeneous (astrocytic and oligodendroglial) morphology before oligoastrocytoma is dismissed as a distinct nosological entity. Key Words: 1p/19q Codeletion, ATRX, IDH, Oligoastrocytoma, TERT. INTRODUCTION In the fourth edition of the World Health Organization (WHO) Classification of Tumours of the Central Nervous System, oligoastrocytoma was listed as a separate entity among the diffuse gliomas (1). In particular, it was defined as a diffusely infiltrating glioma composed of a conspicuous mixture of 2 distinct neoplastic cell types morphologically resembling the tumor cells of oligodendroglioma or diffuse astrocytoma (1). Thereafter, several studies showed that nearly all oligoastrocytomas are monoclonal tumors and that they can be classified as astrocytoma or oligodendroglioma on their biomolecular From the Department of Human Pathology, University of Messina, Messina, Italy (VB and SL); Department of Pathology, Treviso General Hospital, Treviso, Italy (LV, GG and SR); and Department of Neurosciences, University of Messina, Messina, Italy (MC) Send correspondence to: Valeria Barresi, MD, PhD, Department of Human Pathology, Polyclinic G. Martino, Pad D, Via Consolare Valeria, 98125 Messina, Italy; E-mail: vbarresi@unime.it The authors have no duality or conflicts of interest to declare. profile (2 5). In detail, oligoastrocytomas with mutations in isocitrate dehydrogenase 1 and 2 (IDH1 and IDH2) genes and combined whole-arm losses of 1p and 19q (1p/19q codeletion) can be classified as oligodendrogliomas, while those with mutations in IDH1/IDH2, alpha-thalassaemia/mental retardation X-linked (ATRX) andp53 genes can be categorized as astrocytomas (2 5). For this reason, the most recent update of WHO classification states that the histopathological diagnosis of oligoastrocytoma should be reserved to tumors with ambiguous morphology when diagnostic molecular testing cannot be performed (so-called oligoastrocytoma not otherwise specified) or in the very rare instance of a dual-genotype oligoastrocytoma (6). Herein we describe a case of low-grade diffuse glioma composed of distinct areas with astrocytic and oligodendroglial morphology and that had different molecular signature. This case demonstrates that oligoastrocytoma is not just a tumor with ambiguous histology, but a rare neoplasia that represents a distinct nosological entity. MATERIALS AND METHODS Clinical history was obtained from retrospective review of the medical records. The whole surgical specimen was formalin-fixed and paraffin-embedded for histological examination. Four micrometer consecutive sections were cut from the paraffin blocks for microscopic examination with hematoxylin and eosin (H&E) stain, immunohistochemistry and fluorescent in situ hybridization (FISH) analysis. Through histological examination and immunohistochemistry, we identified tumor fragments with different morphology and immuno-phenotype (Fig. 1). Those fragments were dissected and collected separately for IDH1 and IDH2 sequencing and for telomerase reverse transcriptase (TERT) promoter mutational analysis. Similarly, 1p/19q FISH analysis was performed separately in areas with different morphology and immuno-phenotype. Immunohistochemistry Immunohistochemistry was performed using an automated Immunostainer (Dako Autostainer Link 48 Instruments; Dako, Glostrup, Denmark) and the following antibodies against olig-2 (clone 211F1.1, Cell Marque, Rocklin, CA; 1:100), glial fibrillary acidic protein (GFAP) (clone 6F2; VC 2017 American Association of Neuropathologists, Inc. All rights reserved. 1

Barresi et al J Neuropathol Exp Neurol Volume 0, Number 0, Month 2017 IDH1 and IDH2 were as follows: (1) initial denaturation step at 95 C for 5 minutes; (2) 40cyclesat95 C/30 seconds, 58 C/30 seconds, and 72 C/30 seconds; and (3) afinalstep at 72 C/5 minutes. We used the following primers: IDH1-F CCATCACTGCAGTTGTAGGTT; IDH1-R GCAAAAT CACATTATTGCCAAC; IDH2-F TGCAGTGGGACCACT ATTATC; IDH2-R GTGCCCAGGTCAGTGGAT. PCR conditions for TERT were as follows: (1) initial denaturation step at 95 C for 10 minutes; (2) 35cyclesat95 C/30 seconds, 60 C/30 seconds, and 72 C/1 minute; and (3) afinal step at 72 C/7 minutes. We used the following primers: TERT-F GTCCTGCCCCTTCACCTT and TERT-R GCACCT CGCGGTAGTGG. FIGURE 1. Pre-operative axial T2-weighted magnetic resonance imaging showing a hyperintense mass in the temporal lobe. Dako; 1:500), ATRX (Polyclonal; Life Science Sigma, St Louis, MO; 1:750), IDH1 R132H (clone H09, Dianova, Germany; 1:200), p53 (clone DO-7, Dako; 1:100) and Ki-67 (clone MIB-1, Dako; 1:100). FISH Formalin-fixed, paraffin-embedded tissue specimens were processed with LSI 1p36/19q13 Dual-Color Probe Sets (Vysis/Abbott, Molecular Europe, Wiesbaden, Germany) assays, following manufacturer-recommended protocols. Slides were examined by Olympus BX61 fluorescence microscope equipped with a 100 oil immersion objective and a triple band pass filter for simultaneous detection of Spectrum Orange, Spectrum Green, and DAPI signals; 200 nonoverlapping nuclei containing a minimum of 2 reference probe signals were counted. The ratio of 1p/1q and 19q/19p was calculated by dividing the number of orange and green signals. Counting was performed separately in tissue fragments showing different morphology (astrocytic vs oligodendroglial) and nuclear expression of ATRX (ATRX loss vs ATRX maintenance). Tumor fragments were considered to be deleted when the final ratio was <0.8; a target-to-reference probe signal ratio >0.8 indicated no allelic loss of target locus. Molecular Analyses Fragments with differential expression of nuclear ATRX were preliminarily identified and marked in the ATRX immunostained slide (Fig. 2). Thereafter, 5-mm-thick serial sections of those fragments were cut from the corresponding paraffin block, put on uncoated slides and separately dissected and collected in different tubes. Genomic DNA from each area was isolated for molecular analyses using the QIAamp DNA Mini Kit (Qiagen, Venlo, The Netherlands). All areas had at least 80% cellularity. IDH1, IDH2, and TERT were amplified by PCR and both strands were sequenced using the ABI PRISM 3500 Genetic Analyzer (Applied Biosystems). PCR conditions for 2 RESULTS A 25-year-old man referred to our hospital because of vomiting and headache persisting for 1 week. Past clinical history was unremarkable and it was negative for brain irradiation. There was no significant family history of brain tumors or other cancers, and there were no stigmata of neurofibromatosis or enchondromatosis present. Magnetic resonance imaging showed an infiltrating lesion in the right temporal lobe. The lesion was hypointense in T1 and hyperintense in T2 weighted images and it was not contrast-enhanced (Fig. 1). The radiological findings suggested low-grade glioma. The patient was submitted to craniotomy for surgical removal of the lesion and post-surgical RMI demonstrated complete surgical resection of the tumor. Surgical specimen consisted of several tissue fragments that were entirely paraffin embedded for histological examination. Microscopic evaluation with H&E stain revealed a glial tumor with low cellularity and mainly composed of cells with ovoid nuclei and mild nuclear atypia (astrocytic morphology) (Fig. 2A). However, some fragments had a different morphology and showed monotonous neoplastic cells with round, hyperchormatic, nuclei, and clear cytoplasmic halo and thin vessels with chicken-wire pattern (oligodendroglial morphology) (Fig. 2D). GFAP was diffusely expressed in the areas with astrocytic morphology, whereas it was overall negative in the fragments with oligodendroglial morphology. No immunohistochemical staining for IDH1 R132H was found in both the astrocytic and oligodendroglial components. However, ATRX and p53 immunostaining was different in the areas with astrocytic and oligodendroglial morphology. Indeed, the former were positive for p53 and negative for ATRX (Fig. 2B, C), while the latter were negative for p53 and positive for ATRX (Fig. 2E, F). FISH analyses correlated with ATRX expression. Specifically, in the areas which retained ATRX expression, 1p/1q and 19q/19p ratios were 0.54, and 0.55, which corresponds to 1p/19q codeletion (Fig. 2); on the other hand, in the areas with ATRX loss, 1p/1q was 0.89 and 19q/19p was 0.9, which indicates absence of 1p/19 codeletion (Fig. 3). Molecular analysis of IDH1/IDH2 genes revealed IDH2 gene mutation c.515g>t (p.arg172met) at codon 172 in both oligodendroglial and astrocytic areas. Molecular analysis of TERT promoter showed 124C>T (C228T) mutation in the oligondendroglial areas, and absence of mutation in the

J Neuropathol Exp Neurol Volume 0, Number 0, Month 2017 Dual-Genotype Oligoastrocytoma FIGURE 2. Areas with astrocytic morphology were composed of neoplastic cells with oval nuclei and mild nuclear atypia (A) (H&E stain; original magnification, 200); they showed ATRX nuclear loss (B) (ATRX stain; original magnification, 200) and p53 over-expression (C) (p53 stain; original magnification, 200). Areas with oligodendroglial morphology were composed of cells with uniform round nuclei (D) (H&E stain; original magnification, 200) that retained ATRX nuclear expression (E) (ATRX stain; original magnification, 200) and were negative for p53 (F) (p53 stain; original magnification, 200). astrocytic component. Proliferative index assessed through Ki- 67 staining was 2% in both astrocytic and oligodendroglial areas. No tumor recurrence was documented after 4 months follow-up. DISCUSSION In this study, we report on a case of so-called oligoastrocytoma dual-genotype, that is a diffuse glioma with distinct areas (mostly confined in different fragments) showing morphology and molecular signature of oligodendroglioma and astrocytoma. Interestingly, this case was resected from the temporal lobe of a male patient, similarly to 3 previous cases of dual-genotype oligoastrocytoma (7 9) (Table); however, further reports on this entity are needed to clarify whether the association with male sex and temporal location is coincidental or not. Differently from previously reported dual-genotype oligoastrocytomas (7, 8), ours was characterized by IDH2 and not by IDH1 gene mutation (Table). In gliomas, IDH2 mutations are rare as compared with IDH1 mutations, and they are mainly found in tumors with oligodendroglial morphology (10 12). In particular, IDH2 gene mutation c.515g>t, which was detected in this case, is found in only 0.8% 1.8% of gliomas (12). In our case, demonstration of the same IDH2 mutation in astrocytic and oligodendroglial areas suggests common clonal origin and successive differentiation into 2 distinct subclones, one characterized by 1p/19q codeletion and the other by ATRX loss and p53 mutation. This supports the hypothesis that IDH mutation is an early event in gliomagenesis (2). Then, 3 possible scenarios may happen: (1) acquisition of ATRX and p53 mutations and development of astrocytoma; (2) 1p/19q codeletion and origin of oligodendroglioma; (3) splitting into 2 subclones, one acquiring ATRX and p53 mutations and one 1p/19q codeletion, and creation of oligodendroglial-astrocytic tumor (oligoastrocytoma), as in the present case. Interestingly, IDH2 somatic mutations were found in patients with enchondormatosis (Oliier-Maffucci syndrome) (13) and anaplastic astrocytoma was reported in a patient with enchondromatosis and IDH2 R172S mutation (14). Our patient had no clinical signs of enchondromatosis. However, the possibility of enchondromatosis due to constitutional mosaicism for IDH2 p.r172m mutation cannot be excluded, and that this tumor actually represents a collision of two separate gliomas due to this possible underlying glioma predisposition syndrome remains an alternative explanation for the findings. Wang et al (15) recently described low incidence (2.2%) of IDH2 gene mutations a in large series of diffuse gliomas (15). Interestingly, none of their IDH2 mutated cases had P53 or ATRX mutations (15); for this reason, they suggested that microenvironment in IDH2-mutated gliomas may not favor the development of other mutations (15). However, we do not agree with this assumption for several reasons. First, P53 mutation was previously reported in IDH2 mutated anaplastic astrocytoma (14). In addition, ATRX mutational analysis was not available in the majority of IDH2 mutant cases (13/18) (15). Finally, 10 out of the 18 IDH2-mutated gliomas in that series had ambiguous morphology and were classified as oligoastrocytomas 3

Barresi et al J Neuropathol Exp Neurol Volume 0, Number 0, Month 2017 FIGURE 3. Fragments with ATRX expression and those with loss of ATRX were identified and marked in green and red, respectively, so that molecular analyses were carried out separately in the 2 tumor components. Both tumor areas bore IDH2 mutation. Areas with retained ATRX expression, but not those with ATRX loss, had 1p/19q codeletion and TERT promoter mutation. TABLE. Clinicopathological and Molecular Features of Dual-Genotype Oligoastrocytoma in the Literature Dual-genotype oligoastrocytoma Age Gender Site IDH mutation 1p/19q status TERT promoter mutation Grade Recurrence Ou et al (10) 44 Male Temporal lobe NA 1p/19q codel (oligo)/ NA II Unknown del 1p (astro) Huse et al (8) 53 Female Frontal lobe IDH1 R132H 1p/19q codel (oligo)/ NA II (oligo)/iii (astro) Yes (2 months) GBM Wilcox et al (9) 30 Male Temporal lobe IDH1 R132H 1p/19q codel (oligo)/ C250T (oligo) III Unknown Wilcox et al (9) 57 Male Temporal lobe IDH1 R132H 1p/19q codel (oligo)/ NA II Yes (2 years) astrocytic Present case 25 Male Temporal lobe IDH2 R172M 1p/19q codel (oligo)/ C228T (oligo) II None (4 months) NA: not assessed. (15). However, whether P53 and ATRX analyses had been carried out separately in the 2 morphological components it is not specified (15). Therefore, we cannot exclude that accurate and specific investigation of the astrocytoma-like and oligodendroglioma-like areas in those tumors may lead to results equivalent to those found in the present case. Since virtually all 1p/19q codeleted tumors (oligodendrogliomas) exhibit TERT promoter mutation, while only IDH wild type, usually high-grade, astrocytic tumors have such 4 genetic alteration (16), we analyzed also TERT promoter mutation in our case. As expected for dual-genotype diffuse glioma, we observed TERT promoter mutation in oligodendroglial, but not in astrocytic areas. Noteworthy, TERT promoter mutation had been previously analyzed in only 1 case of dual-genotype oligoastrocytoma (Table); in that case, TERT promoter mutation was at highest levels in the oligodendroglial region, absent in the astrocytic region, and at intermediate levels in the mixed region (7).

J Neuropathol Exp Neurol Volume 0, Number 0, Month 2017 The most recent update of WHO classification (6) strongly discourages the histopathological diagnosis of oligoastrocytoma, due to the evidence that nearly all oligoastrocytomas can be categorized as astrocytomas or oligodendrogliomas on their biomolecular profile (3, 17). However, the present case and other previous rare cases of dual-genotype oligoastrocytoma (7 9) suggest that further studies are needed before the concept of oligoastrocytoma is definitively abandoned. In addition, we believe that the incidence of dual-genotype oligoastrocytoma might be currently underestimated. Indeed most studies on oligoastrocytoma were focused on morphologically ambiguous tumors with positive immunohistochemical staining for IDH1 R132H mutant protein (3, 5, 8). However, diffuse gliomas may also have other IDH1 mutations not detectable through immunohistochemistry and IDH2 mutations (12), as in the present case. In addition, a recent study demonstrated that more than a half of IDH2 mutations are found in gliomas with morphological features of oligoastrocytoma (15). Underrating of dual-genotype oligoastrocytoma might also depend on sampling biases. Indeed, in our case we did not observe tumor regions with admixed oligodendroglial and astrocytic neoplastic cells. Hence, it might happen that one of the components of a dual-genotype oligoastrocytoma is missed because of incomplete surgical removal or due to partial histological examination of the tumor. Finally, immunohistochemistry and molecular analyses are usually performed on only one tumor block in routine practice. Therefore, if ATRX immunohistochemistry that predicts 1p/19q codeletion with higher accuracy than morphology is performed on a slide containing only tumor fragments with homogeneous histological aspect, dual-genotype oligoastrocytoma might not be recognized. In conclusion, in this study we reported a unique IDH2 mutant dual-genotype oligoastrocytoma. Although further studies are warranted to clarify the biological behavior of this entity, we believe that the identification of an oligodendroglial, 1p/19q codeleted, component in an otherwise diffuse astrocytoma may be clinically relevant. Indeed, it is well known that 1p/19q codeletion is associated with longer overall survival (15) and with response to chemotherapy in patients with diffuse gliomas. On the other hand, demonstration of an astrocytic component in an otherwise oligodendroglial neoplasia may be predictive of higher recurrence risk (8). In the present case, clear distinction of astrocytic and oligodendroglial areas observed on H&E slides was possible by using ATRX staining. Indeed ATRX loss was found in the former, but not in the latter. Therefore, we suggest evaluating ATRX expression in all the histological slides of tumors with heterogeneous astrocytic and oligodendroglial morphology in order to identify tumor fragments to submit to following molecular analyses. Dual-Genotype Oligoastrocytoma Finally, in our opinion, more extensive studies including IDH1/IDH2 and TERT promoter sequencing are warranted in morphologically heterogeneous tumors before definitely dismissing oligoastrocytoma as a distinct entity. References 1. von Deimling A, Refeinberger G, Kros JM, et al. Oligoastrocytoma. In: Luois DN, Oghaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumors of the Central Nervous System. Lyon: IARC Press; 2007: 63 5 2. Jiao Y, Killela PJ, Reitman ZJ, et al. Frequent ATRX, CIC, FUBP1 and IDH1 mutations refine the classification of malignant gliomas. Oncotarget 2012;3:709 22 3. Sahm F, Reuss D, Koelsche C, et al. Farewell to oligoastrocytoma: in situ molecular genetics favor classification as either oligodendroglioma or astrocytoma. Acta Neuropathol 2014;128:551 9 4. Wiestler B, Capper D, Sill M, et al. Integrated DNA methylation and copy-number profiling identify three clinically and biologically relevant groups of anaplastic glioma. Acta Neuropathol 2014;128:561 71 5. Hewer E, Vajtai I, Dettmer MS, et al. Combined ATRX/IDH1 immunohistochemistry predicts genotype of oligoastrocytomas. Histopathology 2016;68:272 8 6. Louis DN, Perry A, Reifenberger G, et al. The 2016 World Health Organization Classification of tumors of the central nervous system: a summary. Acta Neuropathol 2016;131:803 20 7. Huse JT, Diamond EL, Wang L, et al. Mixed glioma with molecular features of composite oligodendroglioma and astrocytoma: a true oligoastrocytoma?. Acta Neuropathol 2015;129:151 3 8. Wilcox P, Li CC, Lee M, et al. Oligoastrocytomas: throwing the baby out with the bathwater? Acta Neuropathol 2015;129:147 9 9. Qu M, Olofsson T, Sigurdardottir S, et al. Genetically distinct astrocytic and oligodendroglial components in oligoastrocytomas. Acta Neuropathol 2007;113:129 36 10. Hartmann C, Meyer J, Balss J, et al. Type and frequency of IDH1 and IDH2 mutations are related to astrocytic and oligodendroglial differentiation and age: a study of 1,010 diffuse gliomas. Acta Neuropathol 2009; 118:469 74 11. Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in gliomas. N Engl J Med 2009;360:765 73 12. Waitkus MS, Diplas BH, Yan H. Isocitrate dehydrogenase mutations in gliomas. Neuro Oncol 2016;18:16 26 13. Pansuriya TC, van Eijk R, d Adamo P, et al. Somatic mosaic IDH1 and IDH2 mutations are associated with enchondroma and spindle cell hemangioma in Ollier disease and Maffucci syndrome. Nat Genet 2011;43: 1256 61 14. Moriya K, Kaneko MK, Liu X, et al. IDH2 and TP53 mutations are correlated with gliomagenesis in a patient with Maffucci syndrome. Cancer Sci 2014;105:359 62 15. Wang HY, Tang K, Liang TY, et al. The comparison of clinical and biological characteristics between IDH1 and IDH2 mutations in gliomas. J Exp Clin Cancer Res 2016;35:86 16. Arita H, Narita Y, Fukushima S, et al. Upregulating mutations in the TERT promoter commonly occur in adult malignant gliomas and are strongly associated with total 1p19q loss. Acta Neuropathol 2013;126: 267 76 17. Brat DJ, Verhaak RG, Aldpe KD, et al. Comprehensive, integrative genomic analysis of diffuse lower-grade gliomas. N Engl J Med 2015;372: 2481 98 5