Evaluation of 1p and 14q status, MIB-1 labeling index and progesterone receptor immunoexpression in meningiomas: Adjuncts to histopathological grading and predictors of aggressive behavior
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.141248
Source of Support: None, Conflict of Interest: None
Background: Meningiomas are benign central nervous system tumors; however, significant fraction recurs, irrespective of histological grade. Materials and Methods: We performed fluorescence in-situ hybridization for 1p36 and 14q32, and immunohistochemistry for progesterone receptor (PR), p53 and MIB-1 on 84 meningiomas. Results and Conclusion: Sixty-four were convexity tumors (30 grade I, 21 grade II, 13 grade III) and 20 petroclival (grade I; 10 with gross total resection (GTR), 10 with subtotal resection (STR)). Isolated 1p36 deletion was seen in 20% grade I, 28.6% grade II and 30.8% grade III convexity meningiomas, and isolated 14q deletion in one grade III convexity tumor. 1p/14q co-deletion was seen in none of grade I, 28.5% grade II and 30% grade III convexity meningiomas. PR immunoreactivity was less frequent in grade III tumors. Petroclival tumors did not show isolated deletion. However, 1p/14q co-deletion was seen in 20% of petroclival tumors with STR and in none with GTR. Frequency of chromosomal alterations and MIB-1 labeling index both increase with tumor grade. 1p/14q co-deletion is characteristic of grade II/III meningiomas, while PR immunoreactivity inversely correlates with grade, suggesting their use as surrogate markers for grading. Identification of 1p/14q co-deletion in grade I petroclival tumors with STR suggests that unresectable petroclival meningiomas are biologically more aggressive than their grade I convexity counterparts.
Keywords: 1p, 14q, co-deletion, meningioma, progesterone receptor, petroclival
Meningiomas are the most common benign brain tumors, accounting for 35.5% of primary central nervous system (CNS) tumors with an incidence of around 7.22/100,000.  They occur most commonly in middle-aged and elderly patients, with a peak during the 6 th and 7 th decades, and mean age of 56.3 years. , Based on their histomorphological features, these tumors have been classified into grade I, II and III.  Most slow growing benign tumors correspond to grade I. However, some are more aggressive, corresponding to WHO grade II (atypical) and III (anaplastic), and have a tendency to recur. , This grading system, based on histomorphological features, has proven useful in predicting outcomes and in defining treatment regimens for meningioma patients. However, there remains considerable variability in biological behavior and clinical outcome within each grade.  Even histologically benign meningiomas have been seen to recur in 7-20% of cases.  Although there are several documented predictors of meningioma recurrence, such as age above 65 years, female sex, tumor location, tumor size, extent of resection, tumor histology (especially higher grade), presence of calcification, proliferation rate i.e. MIB-1 labeling index (LI), loss of immunopositivity for progesterone receptor (PR), and p53 immunoreactivity, , these do not prove useful for prognostication in all instances.
Genomic characterization of meningiomas could be an efficient route to improved understanding of meningioma tumorigenesis, which is a necessary step in the design of newer rational therapeutic agents.  Studies on the genetic alterations in meningiomas have, however, lagged behind those of other CNS neoplasms. Most molecular alterations are poorly characterized and the genetic classification of meningiomas is still in its infancy. In recent years, cytogenetic fluorescence in-situ hybridization (FISH), comparative genomic hybridization (CGH) and array CGH methods have allowed various investigators to report deletions of a number of loci in different grades of meningiomas. The most common chromosomal aberration noted is the monosomic loss of chromosome 22, which occurs in a majority of these tumors.  However, losses have also been reported on chromosomes 1p, 3q, 6q, 9p, 10q, 14q, 15q and 18q, and have been found to correlate with higher grade of meningiomas.  The loss of 1p in particular has been suggested as a decisive step for meningioma progression. ,,, Loss of 14q is another frequently detected abnormality, which has been found to be associated with higher grade and worse prognosis.  Although meningiomas at different sites display similar histopathological features, there is considerable variability in the clinical outcome of tumors arising in different locations. The location of these tumors influences the extent of resection, which in turn has an impact on the development of recurrence.  While convexity tumors are easily accessible and well-circumscribed, skull base meningiomas, particularly petroclival tumors, are more invasive and are often intimately associated with critical neurovascular structures, thus presenting a unique therapeutic challenge for complete resection.  Understanding the biology and genetics of these aggressive and unresectable meningiomas might significantly deepen the understanding of meningioma progression and provide accurate classification, diagnosis and prognostic evaluation of these tumors. 
Although a few studies are available in Western literature on the molecular alterations in meningiomas, not much data is available on the genetic alterations in these tumors from India. A single study by Asirvatham et al. analyzed four chromosomes for loss of heterozygosity (LOH) or allelic imbalance in meningiomas from a cohort of Indian patients, using microsatellite markers.  The authors found that LOH was most frequently seen on chromosome 22q, across all grades. They also noted that LOH 14q was more frequent in high grade as compared to benign meningiomas. However, this finding did not achieve statistical significance, and they suggested that the role of chromosome 14q be further investigated in meningiomas in the Indian population. Also, this and the few other studies on molecular alterations in meningioma did not correlate chromosomal deletion with other alterations like p53 immunopositivity, hormone receptor status and MIB-1 LI. We therefore undertook this study to assess the frequency of 1p 36 and 14q 32 deletion in various grades and phenotypes of convexity meningiomas, and in grade I petroclival meningiomas that had undergone either gross total resection (GTR) or subtotal resection (STR), based on their invasiveness. The results obtained were correlated with MIB-1 LI, p53 immunoreactivity and PR status to assess the utility of these markers in predicting the biological behavior of these tumors.
All cases diagnosed as meningiomas between January 2002 and December 2011 were identified from a detailed review of the records of the Neuropathology Department at our Institute. Patient records were reviewed to obtain demographic data, including age, sex, tumor location and treatment received. Patients who received gamma knife therapy prior to surgery were excluded. All patients had a complete radiological workup in the form of NCCT/CECT or MRI with contrast. Patients were followed-up for recurrence of tumor or any complications.
The original H and E slides were re-evaluated independently by three neuropathologists (CS, MCS and VS) and the diagnoses reconfirmed as per the WHO 2007 classification.  Cases with adequate material in the paraffin-embedded tissue blocks were taken up for the study.
FISH analysis for 1p and 14q
Dual-probe FISH assay was performed on paraffin-embedded sections, with locus-specific probes for 1p 36 and 14q 32 paired, respectively, with the reference probes for 1q 25 and 18q 21 (Vysis, Downers Grove, IL, USA). Signals were scored in at least 200 non-overlapping, intact nuclei. Sections from non-neoplastic cortical tissue obtained from epilepsy surgery specimens were used as a control for each probe pair. A test: Control ratio of 1:2 in a minimum of 40% nuclei was scored as deletion. The cut-off value was determined by calculating the mean +3 SD of deletion seen in non-neoplastic brain tissue.
Immunohistochemistry for p53, progesterone receptor and MIB-1
Monoclonal antibodies for progesterone receptor (Neomarkers, Fremont, CA, USA; dilution, 1:50), p53- DO1 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA; dilution, 1:200) and MIB-1 (Dako, Glostrup, Denmark; dilution, 1:200) were used. Universal labeled streptavidin biotin kit was used as detection system (Dako, Glostrup, Denmark). Tumor cell staining for progesterone receptor was graded as 0 if no cells stained; 1+ if <10% cells stained; 2+ if 10-50% stained; 3+ if 51-80% stained and 4+ if >80% cells stained.  Staining for p53 was graded as 0 if no cells stained; 1 + if 0-10% stained; 2+ if 10-50% stained and 3+ if >50% stained.  Grade 2+ and above was considered positive for both markers. MIB-1 LI was calculated in the highest proliferating areas as the percentage of labeled nuclei per 1000 tumor cells.
A total of 8796 samples of CNS tumors were received by the Neuropathology laboratory over a period of 10 years. Of these, 1502 (17.1%) were diagnosed as meningiomas. Majority of meningiomas (90.7%) corresponded to WHO grade I, while grade II and grade III meningiomas comprised 7.8% and 1.5% of all cases, respectively. Age of the patients ranged from 10 to 65 years. Majority (95.3%) were adults while 71 were in pediatric age group (4.7%). There was a slight female preponderance (M:F = 1:1.3). Petroclival tumors constituted 5.4% of all meningiomas.
For further analysis, 30 consecutive grade I, 21 consecutive grade II and 13 consecutive grade III convexity meningiomas were taken up. In addition, 20 grade I petroclival meningiomas collected consecutively over the 10-year period were also studied. Ten of these had undergone GTR, and 10 had undergone STR as they showed invasion into surrounding structures.
Isolated 1p deletion
In convexity meningiomas, isolated 1p deletion was identified in 6/30 (20%) grade I tumors, 6/21 (28.6%) grade II tumors and 4/13 (30.8%) of grade III meningiomas. Among the petroclival meningiomas, none of the cases with GTR or STR showed isolated 1p deletion.
Isolated 14q deletion
Only a single case of grade III convexity meningioma (7.7%) revealed isolated 14q deletion, while none of the grade I and grade II convexity meningiomas showed this alteration. None of the petroclival tumors with GTR or STR revealed isolated 14q deletion.
Combined 1p/14q deletion
Deletion of both 1p [Figure 1]c and 14q [Figure 1]d were seen in grade II and grade III convexity meningiomas, in 28.5% and 30% of cases, respectively, while it was absent in grade I convexity tumors. The difference was statistically significant between grade I and grade II (P = 0.008) and grade I and grade III (P = 0.005) tumors. Two of 10 (20%) petroclival tumors with STR showed 1p/14q co-deletion; however, none of the cases with GTR showed co-deletion.
p53, PR and MIB -1 labeling index
Majority of convexity meningiomas analyzed were immunonegative for p53 (57/64; 89.1%). None of the grade I convexity meningiomas showed p53 immunoreactivity [Figure 2]a. Four out of 21 cases (19%) of grade II [Figure 2]b and 3/13 cases (23.1%) of grade III [Figure 2]c convexity tumors displayed p53 immunopositivity. All the p53 immunopositive cases showed only focal staining (1+). Total 20% cases of grade II convexity meningiomas with combined 1p and 14q deletion showed p53 immunopositivity. Although the frequency of p53 immunopositivity was relatively lower (10%) in tumors with normal 1p/14q status, the difference was not statistically significant (P = 1.000). Among grade III convexity tumors, p53 immunopositivity was observed in 1 out of 4 cases with normal 1p/14q status, while none of the cases with 1p/14q co-deletion showed p53 immunoreactivity. Most of the petroclival meningiomas with GTR (80%) and with STR (90%) were immunonegative for p53.
PR immunoreactivity was observed in 64% of convexity tumors included in the study. Majority of grade I (86.7%) [Figure 2]d and grade II [Figure 2]e convexity meningiomas (66.7%) revealed PR immunoreactivity while only 30.7% of grade III [Figure 2]f convexity meningiomas were immunopositive for this marker. Among the immunopositive tumors across all grades, strong PR immunoreactivity (3+ or 4+) was seen in 70.7% of cases. There was no significant difference in PR immunopositivity between male and female patients (P = 1.000). Among grade II convexity meningiomas, similar frequency of PR immunoreactivity was observed in cases with 1p/14q co-deletion (80%) and cases with normal 1p/14q status (70%). Interestingly, none of the grade III convexity meningiomas with co-deletion (0/4) showed PR immunoreactivity, whereas 50% of cases (2/4) with normal 1p/14q status were immunopositive for PR. However, the difference was not statistically significant (P = 0.42). PR positivity was observed in 80% and 90% of petroclival tumors with GTR and STR, respectively, with most cases showing 3+ or 4+ positivity.
Mean MIB-1 LI was 2.76%, 10.5% and 18.46% for grade I [Figure 2]g, grade II [Figure 2]h and grade III [Figure 2]i convexity meningiomas, respectively. The difference in MIB-1 LI was statistically significant between grade I and II (P = 0.016) and grade I and III (P = 0.002) convexity tumors. Mean MIB-1 LI in grade II (10.4%) and grade III (20.5%) convexity meningiomas with 1p/14q co-deletion was higher as compared to cases with normal 1p/14q status (8% in grade II and 12.5% in grade III tumors). This difference, however, was not statistically significant (P value = 0.64 and 0.47, respectively). There was no significant difference in mean MIB-1 LI in petroclival tumors with GTR (2%) and STR (3%).
Early studies on chromosomal alterations in meningiomas showed that one of the commonest cytogenetic changes detected in meningiomas was monosomy of chromosome 22, seen in 70% of cases, and the neurofibromatosis type 2 (NF2) gene localized to chromosome 22q12.2 is the most commonly involved tumor suppressor gene in the development of these tumors. This alteration is probably involved early on in tumorigenesis but not in tumor progression. ,
Recent studies have described the genetic alterations involved in progression of meningiomas.  Cytogenetic studies have indicated that atypical and anaplastic meningiomas frequently show complex numerical and structural aberrations. ,, LOH analyses have identified three chromosomal regions that are rarely altered in grade I, but are often lost in grade II and III meningiomas. These are located on 1p, 10q and 14q, and have been suspected to contain tumor suppressor genes associated with meningioma progression. ,, Thus, malignant progression of meningiomas appears to be associated with the sequential accumulation of various genetic abnormalities, most often deletions at 1p 36 and 14q 32.9.
In our study, the frequency of isolated 1p deletion was 20% in grade I convexity meningiomas, while 28.6% of grade II and 30.8% of grade III convexity meningiomas showed this alteration. Previous studies have shown similar results, with increase in frequency of isolated 1p deletion with increase in tumor grade, ,, recurrence  and tumor progression.  Our study and previous documented series thus support the existence of a tumor suppressor gene located on 1p 36.2, which is involved in meningioma tumorigenesis.  A candidate gene recently identified at this locus is TP73, which encodes a p53-related protein. Studies have shown that TP73 induces apoptosis, cell cycle arrest and differentiation through its interaction with p53, and thus acts as a tumor suppressor gene. 
Isolated 14q 32 deletion was extremely rare in our study, as also reported in previous studies. , In all but one case, loss of 14q was accompanied by 1p deletion. This suggests that while 1p deletion is acquired early on, is seen across all grades and its frequency increases with tumor grade, loss of 14q is an additional event that occurs later on in the process of tumorigenesis, and plays a role in tumor progression. It has long been suggested that chromosome 14q 32 harbors a tumor suppressor gene involved in meningioma pathogenesis and progression, and recent studies have identified the maternally expressed gene 3 (MEG 3) gene at this locus. The MEG 3 gene is normally expressed in arachnoidal cells, in which it functionally suppresses DNA synthesis. It suppresses cancer cell growth, stimulates p53-mediated transcriptional activation and selectively activates p53 target genes. , Thus, loss of MEG3 leads to uncontrolled proliferation of meningeal cells.
Further, we noted that the frequency of chromosomal alterations increases with increase in tumor grade, as in the present study 20% of grade I meningiomas, 52.3% of grade II and 69% of grade III meningiomas carried either one or both chromosomal aberrations, indicating that accumulation of chromosomal alterations is a requisite for malignant transformation of these tumors. Combined 1p/14q deletion showed increasing frequency with tumor grade, in that, it was not observed in any of the grade I meningiomas while it was seen in 28.5% of grade II and 30% of grade III meningiomas. Pfister et al. reported similar results based on analysis of 77 meningiomas, with 1p/14q co-deletion being present in 27% of grade I, 93% of grade II and 100% of grade III tumors.  In a subsequent study on 111 meningiomas (grade I and grade II), the same authors documented 1p/14q co-deletion in 25% of grade I and 93% of grade II tumors. 
High MIB-1 LI is known to be associated with increased mitotic activity  and with recurrence.  We observed a significant correlation between mean MIB-1 LI and tumor grade, similar to previously reported series. , The frequency of PR immunopositivity was 83%, 67%, and 30% in grade I, grade II and grade III convexity meningiomas, respectively. A significant correlation of PR immunoreactivity with histological grade (between grade I and grade III tumors) was observed in this study (P = 0.007), as in previously reported series. , Similar to other studies, , we observed that p53 immunoexpression was rarely identified, and when present, was seen only in higher grades (P = 0.02) and was restricted to weak, focal positivity, suggesting that TP53 mutation does not play a significant role in evolution of these tumors. However, mutation analysis may be required for confirmation of the same. Therefore, absence of PR immunoreactivity and high MIB-1 LI are indicative of higher tumor grade and may serve as markers of poor prognosis in patients with convexity meningiomas.
Meningiomas of the petroclival region usually involve the petrous apex and upper two-thirds of the clivus, and are among the most difficult lesions to manage surgically, given the critical neurovascular structures that they are intimately associated with. , The probability of recurrence and progression for petroclival meningiomas remains uncertain for any given tumor, and recurrence rates for published series have ranged from 0-42%.  The risk of recurrence seems to increase with partial resection, malignant histopathology and involvement of cavernous sinus. ,, In a study by Jung et al. on the long-term outcome of patients with residual petroclival meningiomas, tumor progression occurred in 42% of patients in whom tumor removal was subtotal.  None of the predictors of progression investigated, including age, duration of symptoms, initial symptoms, tumor size, extent of removal, pathological subtypes, presence of brainstem edema or growth pattern of the tumor were associated with a significant increase in rate of recurrence.  There has been no previous study on molecular biology of petroclival meningiomas except one case report in which a subtotally resected petroclival meningothelial meningioma transformed into an atypical meningioma within a year. Chromosomal DNA copy number losses were observed on 1p, 6q, 10, 14q and -22q detected in both the primary and the recurrent tumors. These results suggest that this tumor was potentially malignant at the initial stage, although it had the histological features of a benign meningioma.  Understanding the molecular biology of these invasive meningiomas enveloping sensitive neurovascular structures in which GTR is not safely feasible may provide an insight into factors influencing recurrence in petroclival meningiomas and provide targets for development of novel therapeutic agents. In our analysis, we observed that histopathological features, proliferation index, progesterone receptor or p53 expression in petroclival meningiomas with GTR or STR were similar to corresponding grade I convexity meningiomas. However, in contrast to grade I convexity meningiomas, 20% of grade I petroclival meningiomas with STR showed 1p/14q co-deletion, while those with GTR did not show this alteration in any of the cases. Thus, the presence of 1p/14q co-deletion in petroclival tumors with STR suggests that these have an inherently aggressive biological nature, which manifests as increased invasiveness and locally destructive behavior, making complete resection difficult. Hence, grade I petroclival tumors that show 1p/14q co-deletion should be treated with caution, and these patients should be kept on regular follow-up so as to detect recurrence at the earliest.
To conclude, chromosome 1p36 deletion is an early event in the pathogenesis of meningiomas, the frequency of which increases with increase in tumor grade. Isolated 14q deletion is extremely rare. This molecular alteration gets added on to 1p36 deletion as tumor grade increases. Combined 1p/14q deletion is seen in high-grade convexity meningiomas, while loss of PR immunoreactivity and high MIB-1 LI correlate with increasing tumor grade, suggesting their use as surrogate markers for tumor grading, guiding management and for prognostication of these patients. Majority of meningiomas are immunonegative for p53, and when present, only focal expression is seen. Thus, TP53 mutation possibly does not play a role in the pathogenesis of meningiomas, in contrast to gliomas, which show TP53 mutation in a substantial proportion of cases. Petroclival tumors harboring 1p/14q co-deletion, a genetic marker for aggressive behavior, should be treated with caution, as they may behave more aggressively than those without co-deletion. This is particularly important for incompletely resected tumors, in which 1p/14q co-deletion can be used as a marker for prognostication and for prediction of recurrence.
[Figure 1], [Figure 2]