Prognostication for meningiomas: H3K27me3 to the rescue?

See the articles by Behling et al. pp. 1273–1281 and Nassiri et al. pp. 1282–1291.

Predicting the future is hard. Despite the substantial gains made in understanding the biology, genetics, and epigenetics of meningiomas, in daily clinical practice it remains difficult to identify which of these tumors will recur. Recent efforts, however, have started to chip away at the uncertainty. Two recent articles in Neuro-Oncology now support the value of evaluating the levels of trimethylation of histone H3 (ie, H3K27me3) in meningiomas,^1,2 adding to a growing list of markers that may be used to assess the prognosis for patients with this very common brain tumor.

Histopathology has long provided the first line of stratification for identifying meningiomas at high risk of recurrence. The World Health Organization (WHO) classification system denotes three grades of meningiomas. Grade 1 meningiomas are least likely to recur, whereas grade 3 anaplastic meningiomas frequently recur and are lethal. Nonetheless, up to 25% of grade 1 meningiomas recur and the time to recurrence for grade 3 meningiomas can be highly variable. Even more challenging are grade 2 atypical meningiomas, which have a recurrence rate of 30%-50%. Better markers are needed to supplement traditional histopathology so that radiotherapy and other treatments can be used judiciously.

Immunohistochemistry (IHC) for H3K27me3 has become a useful marker in the diagnosis and prognosis of a number of tumors. In the peripheral nervous system, reduced H3K27me3 is seen in a subset of malignant peripheral nerve sheath tumors with aggressive clinical behavior.³ In the brain, reduction in H3K27me3 is a sensitive marker for H3F3A K27M-mutant gliomas⁴ and for Group-A posterior fossa ependymomas,⁵ tumor types with poor prognosis that require intensive therapy.

Recently, H3K27me3 loss has also been observed in a subset of meningiomas. The first study by Katz and colleagues showed that H3K27me3 loss was associated with significantly shorter time to recurrence in WHO grade 1 and 2 meningiomas.⁶ A subsequent study by Gauchotte and colleagues using a larger set of WHO grade 3 anaplastic meningiomas (n = 47) linked H3K27me3 loss with reduced overall survival but not progression-free survival (using a cutoff of 50% loss of staining).⁷ Although interobserver reliability is often reported as high, challenges in scoring the IHC have been noted due to heterogeneous and indeterminate staining patterns in which areas of loss of H3K27me3 are interspersed within areas of retained trimethylation.

Behling and colleagues now report an analysis of H3K27me3 expression in over 1200 meningiomas resected at the University Hospital Tübingen and arrayed in tissue microarrays.² They found that H3K27me3 is lost in ~5% of tumors overall, with the frequency of loss increasing with grade (grade 1: 3%, grade 2: 10%, grade 3: 18%) and in proliferative tumors (those with Ki-67+ >6.9%). Similar to the study by Katz and colleagues, H3K27me3 provided the most value in discriminating the rate of recurrence in patients with WHO grade 2 atypical meningiomas but may also be useful in characterizing recurrence risk in patients with grade 1 meningiomas. Only 14 grade 3 tumors were studied so more work with larger cohorts will be required to adequately evaluate a role for H3K27me3 IHC in discriminating between anaplastic meningiomas.

The publication by Nassin et al. is the most recent and reports the use of H3K27me3 IHC in a multi-institutional study enriched for grade 2 meningiomas and tumors with challenging prognoses.¹ The authors evaluated whole slides in this report and identified loss of H3K27me3 in ~14% of the samples; in agreement with Behling et al., ~4% of grade 1 meningiomas were H3K27me3-negative. Similar to prior studies, the authors found that loss of H3K27me3 was associated with increased recurrence in grade 2 tumors while a trend was observed in grade 1 tumors. Heterogeneous staining was also noted, and similar to Katz et al., the authors labeled cases as H3K27me3-negative only if complete loss of signal was observed in tumor cells.

These two recent studies combined with prior work provide further support for neuropathologists who are interested in using H3K27me3 IHC as a prognostic marker in the evaluation of meningiomas, particularly for identifying grade 2 tumors with an increased risk of recurrence. Developing and implementing prognostic markers for clinical practice requires careful validation to assess reproducibility in multiple institutions; even markers as widely used as Ki-67 can be challenging to implement and standardize. As neuropathologists gain experience using H3K27me3 IHC in routine practice, they will be able to further assess whether the heterogeneous staining described in research studies is manageable and whether true negatives can be identified reproducibly. The successful implementation of H3K27me3 IHC for meningioma will require standardization of staining protocols and of approaches for scoring and reporting. Sharing of information and data between centers is therefore critical. Ultimately, prospective trials evaluating H3K27me3 IHC performed on whole slides will be needed, but such efforts will take years to complete.

Despite the important contributions provided by these studies on H3K27me3 IHC as a prognostic marker for meningioma, efforts are needed that integrate genomic, proteomic, and epigenomic analyses with assessments of clinical outcomes. As for other cancers, comprehensive phenogenomic TCGA-type studies that are underway are likely to provide important insights into the pathophysiology of these tumors and supply additional biomarkers to more accurately determine patient outcomes and treatment. A brighter future lies ahead.