A 47‐year‐old man with a recurrent glioma

Clinical and Radiological Findings

A 39‐year‐old man was admitted in 2009 following epileptic seizure and visual impairment. MRI revealed an expansive and infiltrative lesion in the right parieto‐occipital lobe with a cortico‐subcortical localization. This lesion was hypointense in T1‐weighted images and showed no contrast enhancement. Methionine (MET‐PET) revealed high methionine uptake in this lesion. Partial surgical resection was performed. There was no adjuvant therapy. MRI showed progression in 2014 but the patient refused any treatment. In 2017, MRI showed areas of enhancement (Figure 1A) and MET‐PET revealed heterogeneous increased metabolic uptake, findings consistent with progression. A second surgery was done.

Figure 1.

Pathology

The 2009 surgical resection was an infiltrating glioma of moderate cellularity, consisting of monomorphic cells with round nuclei and perinuclear haloes (Figure 1B). Few mitoses were observed and there was no necrosis. GFAP and p53 were strongly positive (Figure 1C) and the Ki67 was 1%. IDH1 R132H was negative. There was no 1p/19q co‐deletion by FISH or NGS. The 2017 specimen had a block with areas of round glial cells similar to the 2009 sample (Figure 1D). But the areas of infiltrating tumor had hyperchromatic, pleomorphic nuclei (Figure 1E). Both areas had increased cellularity, mitoses, nuclear atypia and were strongly GFAP positive but negative for IDH1‐R132H. The infiltrating tumor was strongly positive for p53 (Figure 1F), showed loss of ATRX (Figure 1G) and again did not have 1p/19q co‐deletion by FISH and NGS. Unexpectedly, the round glial cells were negative for p53 (Figure 1H), retained normal ATRX staining (Figure 1I) and had 1p/19q co‐deletion by FISH and NGS. In addition, the NGS studies showed that an IDH2‐R172L mutation was present in the 2009 tumor as well as in both areas of the 2017 recurrence. What is your diagnosis?

Diagnosis

2009 sample: Diffuse, Grade II astrocytoma, IDH‐mutant, 1p/19q‐intact.

2017 sample: Anaplastic, Grade III oligoastrocytoma (OA) with a dual genotype.

Discussion

According to the 2016 WHO Central Nervous System tumor classification, which integrated molecular biology and histology into diagnosis, nearly all of the IDH mutant gliomas are classified as either astrocytomas—IDH mutant, or oligodendrogliomas—IDH mutant and 1p/19q co‐deletion. The diagnosis of oligoastrocytoma is strongly discouraged. Despite this, true oligoastrocytomas consisting of histologically and molecularly distinct astrocytic and oligodendroglial tumor populations have been reported, albeit rarely, with only five cases of oligoastrocytomas supported by the molecular findings 1, 2, 3, 4. The present case, to the best of our knowledge, is the first to be extensively analyzed by a large targeted NGS panel (Supplementary Data). This molecular profiling allowed us to identify different tumor cell subpopulations in the different morphological areas. The presence of the same IDH2 mutation in the 2009 and in both areas of the 2017 samples at similar allelic frequencies indicates that this is a driver event in the tumor. As proposed by many authors and the 2016 WHO, we can assume that OAs come from an IDH mutant cell of origin, but later during tumorigenesis develop cytologically distinct subpopulations, as observed in the present case. Indeed, the variation of allelic frequencies for TP53 mutations and the difference in ATRX and CIC mutations is consistent with the presence of molecularly distinct tumor cell subpopulations. The presence of CIC mutation and 1p19q co‐deletion in the tumor block of the 2017 sample is consistent with the diagnosis of an oligodendroglial tumor. In contrast, the absence of CIC mutation, 1p19q co‐deletion and ATRX IHC expression in the 2009 sample and in the infiltrating cells of 2017 sample is consistent with the diagnosis of an astrocytic tumor. This is also supported by the detection of ATRX deletion in the 2009 sample. Different hypotheses on the evolution of the tumor can be formulated. We cannot elucidate whether the two morphological components were already present in the 2009 tumor, or whether these two components evolved due to tumoral progression in 2017. It should be noted that this is the first description of a diagnosis of OA at recurrence. In conclusion, this case revealed the difficulties of diagnosing tumors with mixed morphologies and dual genotypes, when applying the 2016 WHO classification.

EDITOR’S NOTE: Using the standardized terminology recommended by the International Collaboration on Cancer reporting (ICCR) this complex case would be designated:

2016 CNS WHO diagnosis: Oligoastrocytoma, NOS

Histological appearance: Diffuse glioma

WHO (histological) grade: II

Molecular parameters:IDH2 R172L mutation, ATRX mutation, TP53 mutation, variable CIC mutation, variable 1p/19q co‐deletion.