Molecular characteristics of meningiomas in a cohort of Indian patients: Loss of heterozygosity analysis of chromosomes 22, 17, 14 and 10
Correspondence Address: Source of Support: Funded by the Department of Neurosciences, Conflict of Interest: None DOI: 10.4103/0028-3886.111119
Source of Support: Funded by the Department of Neurosciences, Conflict of Interest: None
Background: Though, loss of heterozygosity (LOH) at chromosome 22q is considered to be the most likely initiating event in the formation of meningiomas, LOH at other chromosomes (1, 3, 6, 9, 10, 11, 14.17, and 18) have been implicated in its progression. The aim of this study was to analyze microsatellite markers on a select set of chromosomes including, 22q, 10q, 14q, and 17p for LOH in patients with meningiomas. Materials and Methods: Tumor tissue and its corresponding blood sample were collected from 27 patients with meningioma. Four polymorphic microsatellite markers (D10S520, D17S1289, D14S555, and D22S417) were characterized for LOH analysis. Results: There were 14 World Health Organization (WHO) grade I, 12 WHO grade II and 1 WHO grade III meningiomas. LOH was seen most often at D22S417 with an equal distribution between the grades (33% of informative samples in each grade). Though, LOH at D14S555 was seen in 50% of informative WHO grade II tumors, compared to 11.1% of informative WHO grade I tumors it did not reach statistical significance. However, allelic imbalance (AI) at D14S555 was significantly associated with atypia (P = 0.05). LOH at D17S1289 was seen only in one tumor sample, and none of the informative samples displayed LOH at D10S520. Conclusion: The frequency and equal distribution of LOH at chromosome 22 supports the hypothesis that it is an early event in the tumorigenesis of meningiomas. The association of AI at D14S555 in WHO grade II meningiomas needs to be investigated on a larger set of samples.
Keywords: Loss of heterozygosity, meningioma, microsatellite markers
Meningiomas are common tumors, with an approximate annual incidence of 1 in 16,000  accounting for roughly 25% of primary central nervous system neoplasms. Although, nearly 80% are classified as histologically benign (World Health Organization [WHO] grade I), more than 25% cannot be cured due to inaccessible location or invasiveness.  While the recurrence rate for atypical meningiomas is approximately 40%, this figure rises to 78% for anaplastic meningiomas.  The revised WHO classification system with more stringent criteria, yields nearly 20% atypical (WHO grade II) and 1-2% anaplastic (WHO grade III) meningiomas. 
The histological grade also shows a correlation with survival, with a mortality rate of 21% at 5 years for atypical meningiomas and a median survival time of less than 2 years for anaplastic meningiomas. , However, even histologically benign meningiomas recur after seemingly complete resection, with long-term follow-up studies suggesting recurrence rates as high as 19% at 20 years.  Though, the extent of tumor resection and histological grade are the two strongest predictors of tumor recurrence there is significant variability in the biologic behavior of these tumors that cannot be accounted for by just these two parameters. The only accepted therapy for meningiomas failing surgical intervention is radiation, which is by no means curative. , Investigation of the genes involved in the molecular pathogenesis of meningiomas has highlighted newer targets that have the potential to improve diagnostic and therapeutic strategies, particularly for those patients with aggressive meningiomas that are currently resistant to conventional forms of therapy.
Though, most molecular alterations are poorly characterized, and the genetic classification of meningiomas is still in its infancy, over the past decade, there have been significant advances in our understanding of the biology of meningiomas such as mutations in chromosome 22 initiating tumorigenesis.  In a significant number of meningiomas, loss of chromosome 22 is followed by the partial or complete loss of other chromosomes. These additional losses correlated with a shift toward an atypical or anaplastic histological appearance. Investigators have attempted to delineate the specific molecular sequence of genetic damage that causes progression to high grade states and have identified regions in chromosomal arms 1p, 3q, 6q, 9p, 10q, 14q, 17p, 18p, 18q, and 22q that are often lost in atypical and anaplastic meningiomas but are rarely altered in benign meningiomas and are suspected to contain tumor suppressor genes (TSG). ,,,,,,,, Though, the loss of function of a TSG on chromosome 22 may initiate the development of a meningioma it is not fully clear what additional genetic aberrations are required for its progression.
Based on the hypothesis that meningiomas progress to anaplastic forms due to an accumulation of genetic deletions, Lee et al.,  created a panel of 24 microsatellite markers (that were available in their laboratory) for chromosomal regions likely to be deleted in various grades of meningiomas. Forty-three meningiomas (34 benign, 6 atypical and 3 anaplastic) were analyzed, and D1S407 (P = 0.006), L-myc (P < 0.001), D10S520 (P = 0.003), D10S1173 (P = 0.042), D11S1920 (P = 0.001), D14S555 (P = 0.041), D17S1289 (P = 0.001), D22S417 (P = 0.0021), D22S431 (P = 0.019), and D22S532 (P = 0.028) were found to be important loci predictive of malignancy. Though, several such studies have been carried at centers world-wide data from India has been completely lacking. We therefore, decided to investigate a subset of the markers to establish baseline data in the Indian population on chromosomes involved with meningioma tumorigenesis including markers that found to be highly informative, as mutations in these sites were reported to be statistically significant predictors of atypia in meningiomas. 
The study was approved by the Institutional Review Board and is a cross-sectional observational study.
Residual tumor tissue samples from 27 patients with a clinical diagnosis of meningioma, treated at the Department of Neurosurgery from January 2005 to November 2006, were obtained at surgery after submitting adequate samples for biopsy. The tissue was snap frozen in liquid nitrogen and transported to the laboratory, where they were stored at −70°C. Residual peripheral venous blood samples were also collected in Ethylenediaminetetraacetic acid (EDTA) at the time of surgery. Genomic DNA was extracted from the blood and tumor of all the 27 patients included in this study.
Loss of heterozygosity analysis
DNA from tumor tissue and blood samples was extracted manually using the phenol chloroform method.  The extracted DNA samples were assessed for quantity and quality of DNA using the Nanodrop (Nanodrop technologies, USA) and stored at −70°C. Four polymorphic microsatellite markers-D10S520, D14S555, D17S1289, and D22S417, located on 10q, 14q, 17p, and 22q adjacent to known/possible TSG were selected for characterization of LOH. Primer sequences for all the four loci were obtained from UniSTS ( http://www.ncbi.nlm.nih.gov/unists ) and only the forward primers were 6-FAM-labelled (Sigma Aldrich, India) for fragment analysis. Polymerase chain reaction (PCR) was performed in 12.5 μL reaction volumes that contained Taq polymerase (Amplitaq Gold, Applied Biosystems, USA), 10X PCR Buffer II without MgCl2 (Applied Biosystems, USA), magnesium chloride 25 mM solution (Applied Biosystems, USA) and 200 μM of deoxynucleotide triphosphates (MB Fermentas, Germany). PCR reactions were carried out in a Veriti 96 well Thermal cycler (Applied Biosystems, USA) for 35 cycles, after an initial 8 min at 95°C. The amplified products were electrophoresed using 2% agarose gel (Agarose Low EEO, Medox) and one μl of each amplified product from blood samples for each of the four loci were individually added to 14 μl of formamide (Plus One formamide, Amersham Biosciences, Uppsala, Sweden) and 0.5 μl of the standard LIZ @ (Gene Scan 500 LIZ size standard, Applied Biosystems, Warrington, UK) and denatured for 10 min at 95°C and subsequently cooled on ice for 5 min. The samples were then subjected to capillary gel electrophoresis using a fluorescence based DNA fragment size analysis system (ABI 3130 Genetic Analyzer, Applied Biosystems, USA). Electropherograms were plotted with the fragment sizes and peak heights of alleles in relative fluorescence units. A blood sample was considered informative if it had two alleles.  Blood samples showing only one allele were considered non-informative [Figure 1]. Only informative blood samples [Figure 2] had their corresponding tumor samples analyzed for LOH. LOH and allelic imbalance (AI) were calculated based on the suggestion of Skotheim et al.  as follows:
A value ≤0.84 or ≥1.19 was chosen as the cut-off for interpretation as AI [Figure 3], versus retained heterozygosity.  A value ≤0.5 or ≥2.0 was interpreted as LOH [Figure 4], which requires at least 50% of the tumor cells in a sample to have lost the allele. ,
Statistical analysis was performed using the SPSS software (version 15). One meningioma (WHO grade III) was excluded from the analysis for LOH and AI at locus 17 as the tumor had a persistent non-specific band which was not seen in the blood and hence, could not be interpreted with certainty. As there was only one WHO grade III tumor, WHO grades II, and III were merged for the analysis of all the other variables in the study. Fisher exact test was used to compare variables between WHO grades I, II, and III.
There were 14 WHO grade I, 12 WHO grade II [Figure 1] and 1 WHO grade III meningioma among the 27 samples included in the study. Demographic information obtained showed that the age of patients ranged from 10 years to 67 years with a mean of 45.93 (±14.76) years and included 15 females and 12 males. DNA from all 27 meningioma samples and corresponding blood samples were amplified and analyzed. The informative rate of each locus in terms of LOH is shown in [Table 1] and ranged from 62.96% to 81.48%. The most informative locus was D17S1289. The frequency of AI and LOH of each locus is shown in [Figure 2]. LOH was most frequently seen at D22S417 with equal distribution across grades (33%). 50% of WHO grades II and III tumors displayed LOH at Chromosome 14, compared to 11% of WHO grade I tumors. However, the difference was not significant. Nevertheless, AI at DS14S555 between the benign and atypical groups was significantly different (P = 0.05). [Figure 3] is an illustrative example of an informative blood sample and tumor demonstrating AI, for D22S417 marker, whilst [Figure 4] is an example of LOH for the D22S417 marker. Whilst both the cases illustrated in [Figure 3] and in [Figure 4] show heterozygous alleles with a visibly lower peak for the tumor sample compared to blood, on applying Skotheim's formula,  the former case had a ratio of 1.3 whilst the latter of 3.1. Thus, the case illustrated in [Figure 3] had an AI and the case in [Figure 4] had a LOH. Analysis of the other markers for LOH demonstrated that there was variability in the frequency of LOH ranging from 0% at D10S520 to 33.33% at D22S417 for the corresponding tumor samples analyzed.
Recurrence/regrowth versus LOH
Follow-up data was available for 18 of the 27 samples characterized with the period of follow-up ranging from 9 months to 72 months. Recurrence was observed among 4 of these 18 cases where 2 tumors with recurrence were WHO grade II tumors and the remaining were WHO grade I. Three of these tumors with recurrence were informative at locus D14S555. While 1 of these showed AI, 2 were found to demonstrate an LOH (66.6%). Only one of the recurrent tumors showed LOH at D22S417. LOH at chromosome 17 was not seen in any of the tumors with recurrence/regrowth.
Though, significant progress has been made in delineating the molecular mechanisms of meningioma initiation and progression, the grading of meningiomas and the identification of tumors with higher morbidity still remains challenging with many unanswered questions.
In the present study, LOH was assessed at the following microsatellite targets-D10S520, D14S555, D17S1289, and D22S417. LOH was most often seen in chromosome 22, with an equal frequency (33% each) in WHO grade I and WHO grades II/III. The distribution of AI at chromosome 22 between the atypical group and benign group was not statistically significant. This is in contrast to the results obtained by Lee et al.,  where 3 of 24 benign tumors and 5 of 6 atypical/anaplastic tumors showed LOH (P = 0.0021). In previous studies, deletions in chromosome 22 involving the NF2 gene were seen across the entire histological spectrum of meningiomas  and additional mutations were rarely seen in the absence of monosomy 22  suggesting that LOH at this locus is likely to be an initiation event, rather than a progression event leading to higher states of atypia. The high frequency and equal distribution seen in benign and atypical meningiomas in the present study raises the possibility that LOH at the locus D22S417, is also an initiating event, and is unlikely to assist in determining if the tumor is atypical.
Though 50% of atypical tumors showed LOH 14 compared to 11% of benign meningiomas, this was not statistically significant (P = 0.22), contrary to other studies. ,, Although LOH 14 per se was not significantly associated with atypia we found AI at this locus to be significant between the benign and atypical groups (P = 0.050). Allelic imbalance can be caused by LOH in a smaller clone of at least 20% of the tumor cells or a total loss in tumor cells that are masked by normal cells.  One of the limitations of the study was the non-availability of laser-microdissection at our institution that precluded the ability to select the areas of highest atypia for the DNA extraction and LOH studies. Tumor DNA was extracted from the entire sample that was submitted and therefore it was not possible to ascertain if these areas were fully representative of the atypia or had a significant contribution from non-neoplastic cells (stromal cells, lymphocytes). Another factor might be the stringent cutoffs used to determine LOH in this study. LOH 14 in benign meningiomas has been associated with recurrence. ,,, Follow-up data for the only benign meningioma with LOH at chromosome 14 was not available in the present study.
As previously stated, only one benign tumor showed LOH at chromosome 17 and on reviewing the histopathological records for the case, this tumor had a mindbomb E3 ubiquitin protein ligase 1 (MIB) of 8%, with immunopositivity for bcl2 and mitosis of 3/10 hpf. It is possible that LOH at chromosome 17 in this tumor is suggestive of an evolving atypical clone, which could develop morphological features of atypia in due course. However, this patient died due to complications of urosepsis in the post-operative period and hence further observations were not possible. Interestingly in the study by Lee et al.  none of the benign tumors showed LOH at this site.
Other limitations of the present study was that only one locus was analyzed for each of the chromosomes selected as this was a pilot study and baseline data of the genetic profile of meningiomas among Indian patients was lacking. The small sample size did not permit stronger statistical associations. Moreover, point mutations and small deletions do not result in loss of alleles and hence cannot be detected by LOH analysis. LOH analysis therefore underestimates the extent to which a chromosome arm contains a TSG. Finally, as the samples were collected in 2005 with a possible follow-up period of only 6 years, recurrence could not be studied. Nevertheless, this study adds to the existing scientific body of knowledge. The validity of using D22S417 to differentiate between benign and atypical states is questionable as it was equally distributed in both groups in the present study. The AI seen at locus D14S555 in atypical meningomas compared to benign meningiomas is intriguing: Whether it represents a smaller atypical clone and/or masking of LOH by normal cells can only be determined by further studies using a larger sample size and laser microdissection to isolate regions of atypia. Lastly, the complete absence of LOH at D10S520 in any tumor, benign or atypical, could represent intrinsic biologic differences in the study population.
The development of a neoplasm is multifaceted. Since cells have several safeguards to protect themselves, mutations in multiple genes are required for cancer to develop. Identifying the genes/loci involved in the progression to atypical states would facilitate better understanding of the facets of neoplastic progression and possible incorporation of immunohistochemical markers reflecting relevant and synergistic mechanisms of tumor biology into the WHO grading. Development of diagnostic panels using microsatellite markers may lead to increased accuracy in the prognostication of meningiomas.
The study details the molecular characteristics of an Indian cohort of meningiomas highlighting the few intrinsic differences that may exist in this population.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]