Anatomical predilections and pathologic-genomics discrepancies in radiation induced meningiomas

Introduction: Radiation induced meningiomas (RIM) arise from exposure to radiation in survivors of childhood cancer. RIMs are thought of being more aggressive than sporadic meningiomas. While the understanding of genomic characteristics of sporadic meningiomas has advanced greatly in the past few years, the molecular and spatial characteristics of RIMs remains largely unknown.

Methods: We retrospectively analyzed 68 radiation-induced meningiomas from 60 patients for which genomic profiling was performed. We obtained histopathology reads and medical records since first radiation therapy. For the meningiomas resected at our institution, pre and post-operative MRIs were analyzed with volumetric contouring to assess anatomical tumor burden and location.

Results: Amongst patients with WBRT without focal boost, 36% developed first meningiomas in frontal regions, with 14.4% growing in the falx, suggesting potential selective supratentorial radiation vulnerability to meningioma genesis across intracranial locations. Surprisingly, the patients that received focal radiation either as part of WBRT or targeted radiation to tumor bed only, developed meningiomas also within the frontal lobes in a similar percentage of cases as those receiving whole brain radiation (40%, p=1).

As expected, primary RIMs had more copy-number-alterations per sample than sporadic samples (p < 0.0001). Chromosomal arm losses in 1p, 1q, and 9q, along with 4q gain (FDR < 0.05) were enriched in RIMs. Interestingly, there was a large subset of 37%) of primary RIM samples that had a high genomic burden, but with 2 or fewer mitoses, a marker of less aggressive tumors.

Conclusion : We observed a predilection of RIMs occurring above the tentorium even though a large fraction of our samples received focal radiation to the posterior fossa as part of medulloblastoma treatment. This might reflect a biological vulnerability of meningeal cells between these regions, warranting further research.
A subset of primary RIM samples with aggressive genomic features had low number of mitoses in pathology analysis, which paradoxically would suggest a more benign tumor progression. Further studies on the interplay of genomic aberrations with pathological characteristics are warranted to better predict the recurrence of RIM.