To the Editor: In the 2016 WHO Classification of Central Nervous System Tumors, microscopic evidence of brain invasion has been added as a new grading criterion for meningiomas.1 In several studies, increased frequency of brain invasion in males2-4 and sex-related invasion patterns were shown.4 Sex hormone receptor expression in meningiomas, especially for estrogen and progesterone (ER/PR), has been widely investigated5-9 and, of note, distinct alterations of extracellular matrix genes were shown comparing tumors with and without PR expression.10 However, it remains unclear if sex-related differences of brain-invasive growth are related to distinct sex hormone receptor expression in brain-invasive vs noninvasive meningiomas. We therefore compared PR and ER expression in brain-invasive and noninvasive meningiomas by immunohistochemical staining. Thirty-two brain-invasive and 32 age-, sex-, tumor location-, and WHO grade-matched noninvasive high-grade (Table) as well as 25 randomly selected grade I meningiomas were chosen from a previously published series. Clinical data were obtained from medical and operative reports.4 No recurrent tumors were included. All specimens had been neuropathologically reviewed according to the 2016 WHO Classification of Central Nervous System Tumors. Accordingly, brain invasion was a stand-alone grading criterion for atypia and diagnosed in case of microscopic evidence of tumor tissue within the adjacent brain without a separating connective tissue layer.1 Pattern of invasive growth was classified into (a) diffuse (single cells spreading into the brain parenchyma), (b) nests/clusters of tumor cells, and (c) finger-like tumor expansion into the adjacent brain (see figures in Spille et al.4). TABLE. - Results of the Matched-Pair Analyses of Brain-Invasive and Noninvasive High-Grade Meningiomas. Statistical Analyses Show no Differences Concerning Age, Sex, Tumor Location, and WHO Grade Comparing Both Groups Brain-invasive Noninvasive P-value Number 32 (50%) 32 (50%) Median age (years, range) 61 (32-84) 61 (27-83) .865 Female:male ratio 0.88 0.88 1.00 Location Convexity 12 (38%) 16 (50%) .198 Falx/parasagittal 8 (24%) 5 (16%) Skull base 12 (38%) 8 (25%) Othersa 0 3 (9%) WHO grade II 29 (91%) 31 (97%) .613 III 3 (9%) 1 (3%) aincluding 2 posterior fossa and 1 intraventricular meningioma. Immunohistochemical staining for ER (anti-ER, dilution 1:25, M7047, Dako, Glostrup, Denmark) and PR (anti-PR, dilution 1:100, M3569, Dako) was performed on representative formalin-fixed paraffin-embedded tumor samples according to the manufacturer's instructions using an automatic staining system (Dako). ER/PR expressing human breast cancer samples were used as positive controls (Figures A and B). Staining was analyzed by 2 independent observers (DCS and KH) according to the immunoreactive score (IRS) by Remmele et al.11 Hence, nuclear but not unspecific cytoplasmic staining was evaluated (Figure C). IRS was classified as absent (0-2 points), weak (3-4 points), moderate (6-8 points), or strong (9-12 points).FIGURE.: Positive controls and representative samples of ER/PR immunohistochemical staining. ER (A, anti-ER, M7047, Dako, Glostrup, Denmark) and PR (B, anti-PR, M3569, Dako) positive breast cancer samples were used as positive controls. Occasionally observed cytoplasmic staining was not considered positive (C, anti-ER, M7047, Dako). In comparative analyses of the invasion zone of the tumor D and distant inner tumor regions (E, magnification ×400), PR expression was found to be similar. Correspondingly, PR expression was also found in invasive tumor cell clusters distant from the main meningioma tissue (F, magnification ×200).Data were analyzed in univariate and multivariate analyses using commercial statistic software (IBM SPSS Statistics, Version 23, IBM, Armonk, New York). All P-values reported are 2-sided and considered significant if <.05. Data collection and scientific use was approved by the local ethics committee (Münster 2007-420-f-S) and permitted by the patients in each single case. In total, 42 males (47%) and 47 females (53%) with a median age of 59 years (range: 27-84) harboring 25 grade I (28%), 60 atypical (67%), and 4 anaplastic (5%) meningiomas were included. Tumors were located at the convexity (n = 38, 43%), the falx/parasagittal (n = 16, 18%), the skull base (n = 31, 35%), or in other locations (n = 4, 4%). In atypical meningiomas, grading was exclusively based on the microscopic evidence of brain invasion in 20 (33%, “otherwise benign”) or on other histopathological criteria in 31 specimens (52%), while a combination of both was present in 9 samples (15%). Brain invasion was diffuse in 3 (3%), cluster-like in 11 (12%), and finger-like in 18 (20%) patients. Invasion was found to be finger-like, diffuse, or cluster-like in 13 (76%), 1 (6%), and 3 (18%) of 17 males but in 5 (33%), 2 (13%), and 8 (53%) of 15 females, respectively (P = .037). Immunohistochemical staining for ER and PR was successful in all specimens. Despite clearly detectable staining on positive controls, ER expression was absent from all meningioma specimens. In contrast, PR expression was found in 68 samples (76%) and was weak, moderate, or strong in 16 (18%), 19 (21%), and 33 (37%) individuals, respectively. PR expression was detected in 18 benign (72%), 47 atypical (78%), and 3 anaplastic (75%) meningiomas (P = .820) and IRS did not significantly differ between the 3 groups (P = .363). Similarly, PR expression was detected in 25 of 32 brain-invasive and 43 of 57 noninvasive meningiomas (78% vs 75%; P = 1.00) and IRS was similar comparing both groups (P = .711). No differences of IRS and dichotomous PR expression were found when exclusively comparing 32 brain-invasive and 32 noninvasive high-grade meningiomas (P = .601 and P = .924). IRS was similar in tumors, in which grading was exclusively based on brain invasion or on other histopathological criteria (such as geographic necrosis, etc.) or on a combination of both (P = .593). Age-, sex-, tumor location-, and WHO grade-adjusted multivariate analyses confirmed a lack of any analyzed variable including brain invasion (odds ratio: 0.97, 95% confidence interval 0.27-3.52; P = .962) and PR expression. In 32 brain-invasive tumors, 14 of 18 (78%) patients with finger-like, 9 of 11 with cluster-like (82%), and 2 of 3 individuals (67%) with diffuse invasion showed PR expression (P = 1.00). Correspondingly, subtype of invasion was not correlated with IRS (P = .971). Finally, PR expression in the brain-invasive tumor margins and representative, distant regions of the same slides were compared. However, no difference of distribution of PR expression was found when comparing both regions dichotomously (P = .101) and according to the IRS (P = .339, Figures D-F). Frequency of PR expression in our series matches observations from previous studies.7-9 Similarly, although ER expression has been commonly reported,8 weak to absent expression in benign and especially high-grade meningiomas have been previously shown.7,12 Throughout the whole analysis, ER and PR expression did not differ between brain-invasive and noninvasive tumors in both cumulative and subgroup analyses of high-grade meningiomas. Additionally, expression was similar among the varying morphological invasion subtypes and did not differ between the invasion zone and distant inner tumor regions. These findings are remarkable regarding the higher frequency of brain invasion in male meningioma patients and the sex-related differences of invasion patterns.2-4 Moreover, Claus et al.10 reported distinct alterations of genes in extracellular matrix pathways comparing meningiomas with and without PR expression. Although degradation of extracellular matrix proteins is considered a key feature of invasive growth,13 we could not further support their results in terms of different PR expression in brain-invasive and noninvasive tumors. In conclusion, our analyses provide no evidence that the sex-related differences of brain-invasive growth in meningiomas are correlated with ER or PR expression. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
No takes yet. Share an insight, caveat, or question.
Heß et al. (2017) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: