Proposal for a new risk stratification classification for meningioma based on patient age, WHO tumor grade, size, localization, and karyotype

Abstract

Background: Tumor recurrence remains the major clinical complication of meningiomas, the majority of recurrences occurring among WHO grade I/benign tumors. In the present study, we propose a new scoring system for the prognostic stratification of meningioma patients based on analysis of a large series of meningiomas followed for a median of >5 years.

Methods: Tumor cytogenetics were systematically investigated by interphase fluorescence in situ hybridization in 302 meningioma samples, and the proposed classification was further validated in an independent series of cases (n = 132) analyzed by high-density (500K) single-nucleotide polymorphism (SNP) arrays.

Results: Overall, we found an adverse impact on patient relapse-free survival (RFS) for males, presence of brain edema, younger patients (<55 years), tumor size >50 mm, tumor localization at intraventricular and anterior cranial base areas, WHO grade II/III meningiomas, and complex karyotypes; the latter 5 variables showed an independent predictive value in multivariate analysis. Based on these parameters, a prognostic score was established for each individual case, and patients were stratified into 4 risk categories with significantly different (P < .001) outcomes. These included a good prognosis group, consisting of approximately 20% of cases, that showed a RFS of 100% ± 0% at 10 years and a very poor-prognosis group with a RFS rate of 0% ± 0% at 10 years. The prognostic impact of the scoring system proposed here was also retained when WHO grade I cases were considered separately (P < .001).

Conclusions: Based on this risk-stratification classification, different strategies may be adopted for follow-up, and eventually also for treatment, of meningioma patients at different risks for relapse.

Keywords: SNP arrays; iFISH; meningioma; recurrence; risk stratification.

Fig. 1.

Clinical, biological, and genetic features of meningiomas that showed a significant impact on patient relapse-free survival (RFS; n = 261). Relapse-free survival curves of meningioma cases according to patient age (A), sex (B), tumor localization (C), and size (D), presence versus absence of edema (E), tumor histology (F), WHO grade (G), and interphase fluorescence in situ hybridization (iFISH) cytogenetic profile (H). The last 2 panels (I and J) show patients' RFS curves according to the new prognostic scoring system here proposed (I) and that previously reported by Maillo et al (J).

Fig. 1.

Clinical, biological, and genetic features of meningiomas that showed a significant impact on patient relapse-free survival (RFS; n = 261). Relapse-free survival curves of meningioma cases according to patient age (A), sex (B), tumor localization (C), and size (D), presence versus absence of edema (E), tumor histology (F), WHO grade (G), and interphase fluorescence in situ hybridization (iFISH) cytogenetic profile (H). The last 2 panels (I and J) show patients' RFS curves according to the new prognostic scoring system here proposed (I) and that previously reported by Maillo et al (J).

Fig. 2.

Clinical, biological, and genetic features of WHO grade I meningiomas with a significant impact on patient relapse-free survival (RFS; n = 227). RFS curves of WHO grade I meningioma cases grouped according to patient age (A), tumor localization (B), size (C), and interphase fluorescence in situ hybridization (iFISH) cytogenetic profile (D). RFS curves are shown for the same cases grouped according to the new prognostic scoring system proposed here (E) and that previously reported by Maillo et al (F).

Fig. 3.

Impact of tumor karyotype, defined by the chromosomal copy number profile, as analyzed by SNP arrays on relapse-free survival (RFS) of meningioma cases. (A) RFS curves of meningioma cases classified according to the copy number patterns (diploid, isolated alteration of a single chromosome, and complex karyotypes) for the 2 series of cases analyzed by single-nucleotide polymorphism (SNP) arrays (n = 132) published so far in the literature. Only cases with follow-up data (n = 108) were considered for RFS analysis. (B) RFS curves of the same cases grouped according to the new prognostic scoring system proposed here. (C) RFS curves of the same cases grouped according to the prognostic scoring system previously reported by Maillo et al. In both panels B and C the cases from the Lee series were adjusted for the information available (data on tumor localization and size were not available).