Clinical and Molecular Features of Patients with Gliomas Harboring IDH1 Non-canonical Mutations: A Systematic Review and Meta-Analysis

The canonical isocitrate dehydrogenase 1 R132 mutation (IDH1 R132) is the most frequent mutation among IDH-mutated gliomas. Non-canonical IDH1 mutations or IDH2 mutations are unusual and their clinical and biological role is still unclear. We performed a systematic review and meta-analysis to assess the clinical role of IDH non-canonical mutations. Overall, we selected 13 of 3513 studies reporting data of 4007 patients with a diagnosis of grade 2 and grade 3 glioma including 3091 patients with a molecularly proven IDH1 or IDH2 mutation. Patients with non-canonical IDH1 mutations were younger and presented a higher DNA methylation level as compared to those with canonical IDH1 R132H alteration. The overall incidence of non-canonical IDH1 mutations was 7.9% (95% CI 5.4–10.7%) in patients with IDH-mutated gliomas. There was no statistical difference in terms of incidence between patients with grade 2 or grade 3 glioma. Patients with non-canonical IDH mutations had a lower rate of 1p19q codeletion (risk difference 31%, 95% CI 23–38%) and presented a significantly prolonged survival (pooled HR 0.47, 95% CI 0.28–0.81) as compared to those with IDH1 R132H mutation. Non-canonical IDH1 mutations occur in 7.9% of IDH-mutated gliomas and identify a specific subgroup of patients with an improved survival despite a lower rate of 1p19q codeletion. Data about the type of IDH mutation should be collected in clinical practice and within interventional trials as this could be a critical variable for improved stratification and selection of patients.

ABSTRACT Introduction: The canonical isocitrate dehydrogenase 1 R132 mutation (IDH1 R132) is the most frequent mutation among IDH-mutated gliomas. Non-canonical IDH1 mutations or IDH2 mutations are unusual and their clinical and biological role is still unclear. Methods: We performed a systematic review and meta-analysis to assess the clinical role of IDH non-canonical mutations. Results: Overall, we selected 13 of 3513 studies reporting data of 4007 patients with a diagnosis of grade 2 and grade 3 glioma including 3091 patients with a molecularly proven IDH1 or IDH2 mutation. Patients with non-canonical IDH1 mutations were younger and presented a higher DNA methylation level as compared to those with canonical IDH1 R132H alteration. The overall incidence of non-canonical IDH1 mutations was 7.9% (95% CI 5.4-10.7%) in patients with IDH-mutated gliomas. There was no statistical difference in terms of incidence between patients with grade 2 or grade 3 glioma. Patients with non-canonical IDH mutations had a lower rate of 1p19q codeletion (risk difference 31%, 95% CI 23-38%) and presented a significantly prolonged survival (pooled HR 0.47, 95% CI 0.28-0.81) as compared to those with IDH1 R132H mutation. Conclusion: Non-canonical IDH1 mutations occur in 7.9% of IDH-mutated gliomas and identify a specific subgroup of patients with an improved survival despite a lower rate of 1p19q codeletion. Data about the type of IDH mutation should be collected in clinical practice and within interventional trials as this could be a critical variable for improved stratification and selection of patients.

INTRODUCTION
The diagnosis of gliomas requires an integrated histological and molecular assessment [1][2][3]. In particular, the 2016 World Health Organization (WHO) Classification of Tumors of the Central Nervous System recognized specific tumor subtypes according to different genomic alterations [3].
The IDH genes are also a critical prognostic factor since patients harboring IDH mutations have significantly longer survival compared to those with IDH wild type (wt) [4][5][6][7].
There is an increasing amount of data suggesting that patients harboring IDH1 mutations different from R132H have specific clinical, radiological, and molecular features [9]. Nonetheless, data about the effective clinical role of non-canonical IDH mutations is still conflicting. Here we investigated the clinical role of IDH non-canonical mutations through a systematic review and meta-analysis. We describe outcomes considering IDH mutations as follows: In the case of multiple publications on the same cohort of patients, we included the most updated version with a longer follow-up. In case of the presence of both abstract and complete published version of the same cohort, we selected the complete publication.
In addition, we collected data about the modality adopted for IDH assessment and age at diagnosis between subgroups with different IDH mutations. Finally, we recorded also the different grades of MGMT methylation according to the type of IDH mutation.

Outcomes of the Meta-Analysis
We were interested to investigate multiple issues (see below). In particular, all these outcomes were focused on patients with WHO grade 2 or grade 3 gliomas excluding all patients with glioblastoma.
The main outcomes of the present analysis were: 1. The overall incidence of non-canonical mutations among patients with gliomas.
In particular, we were interested to assess the overall incidence of IDH non-canonical mutations and IDH1 non-canonical mutations among all grade 2 and 3 gliomas and IDH-mutated gliomas. To assess these issues, we selected studies reporting retrospective and prospective cohorts of patients with complete incidence data. Thus, we did not include case-control studies for this outcome. 2. Incidence of IDH1 non-canonical mutations according to WHO tumor grade among patients with gliomas. For this outcome, we selected only studies reporting complete incidence data according to different tumor grades. For the risk-difference analysis (adopted to estimate the different incidence between patients with grade 2 and grade 3 tumors) we included also case-control studies.

Survival comparison between patients with
IDH and IDH1 non-canonical mutations.
For this outcome, we selected only studies reporting complete survival data. The preferred summarizing tool was hazard ratio (HR) with 95% confidence interval. When available we used the HR provided by the multivariate Cox regression model instead of that provided by the univariate log-rank test. Studies performing a survival comparison but not showing an HR were also reported in the text but not included in the analysis. 4. Difference between 1p19q codeletion incidence among patients with canonical IDH1 mutation or non-canonical IDH mutation. 5. For this outcome, we selected only studies reporting complete incidence data according to the different 1p19q incidence rates. Since we were interested in an incidence ratio between 1p19q patients and patients with IDH1 canonical/non-canonical mutations we included also case-control studies. 6. DNA methylation levels, localization of tumors in the central nervous system (CNS), and age of diagnosis difference between patients with/without non-canonical mutations.
Studies reporting a comparison of methylation grade and age of diagnosis were selected for this outcome.

Statistical Methods
All analyses were performed through R statistical software. Packages adopted for the analysis were tidyverse, dplyr, meta, and metaphor.
In survival analysis, we applied the inverse variance technique for HR assessment reporting both random and fixed effects models.
For incidence analysis, we used the Freeman-Tukey double arcsine transformation of proportions. For estimation of between-study variance, we used inverse variance with the DerSimonian-Laird method.
Finally, the difference between proportions was derived through a risk difference comparison. The inverse variance weighting was used for pooling results.

Risk of Bias
We employed the Newcastle-Ottawa scale (NOS) to assess the risk of bias of studies included in the meta-analysis. Four authors independently reviewed all studies (VDN, EF, AT, LG) rating each selected study. Studies with a score of 7 or more, 4-6, and lower than 4 were considered to have a low, moderate, and high risk of bias, respectively [10].

Compliance with Ethics Guidelines
This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.
The two main reasons for study exclusion were (1) review or case reports, (2) lack of distinction between the type of IDH mutations.

Grade
We investigated the overall incidence of noncanonical mutations according to tumor grade in IDH-mutated gliomas. Five of 13 studies [12][13][14][15]17] reported complete data about the incidence of patients with IDH1 non-canonical mutations while four studies described noncanonical IDH1 mutation incidence among patients with grade 3 IDH-mutated gliomas [11,[13][14][15]. The pooled incidence resulting from analysis was 8.2% (95%CI 4.9-12.0% I 2 57%, Fig. 3a) and 6.7% (95% CI 5.0-8.7%, I 2 0%, Fig. 3b) for patients with grade 2 and grade 3 IDH-mutated gliomas, respectively. No differences in terms of risk difference emerged on the pooled analysis between patients with grade 2 and grade 3 IDH-mutated gliomas (Fig. 3c). These results did not change when adding the case-control study included in our study (Fig. 3d). non-canonical mutations. The pooled risk difference estimated was as 31% (95% CI 23-38%, I 2 6%, Fig. 4), showing a significantly higher possibility to found a 1p19q codeletion among patients with canonical IDH1 mutations compared to those with non-canonical IDH1 mutations [11,14,15]. Fig. 2 a Overall incidence of IDH non-canonical mutations in the overall cohort of gliomas. b Overall incidence of IDH1 non-canonical mutations in the overall cohort of gliomas. c Overall incidence of IDH1 non-canonical mutations in the overall cohort of IDH-mutated gliomas Fig. 3 a Incidence of IDH1 non-canonical mutations among IDH-mutated grade 2 gliomas. b Incidence of IDH1 non-canonical mutations among IDH-mutated grade 3 gliomas. c Risk difference between the incidence of IDH1 non-canonical mutations in grade 2 and grade 3 gliomas. d Risk difference between the incidence of IDH1 non-canonical mutations in grade 2 and grade 3 gliomas with the addition of the case-control study

Survival
Four of 13 studies reported complete survival data for patients with IDH non-canonical (n = 160) and canonical mutations (n = 1019) [9,15,[20][21][22]. The pooled HR of these studies was 0.47 (95% CI 0.28-0.81, I 2 74%, Fig. 5), confirming a possible positive prognostic role for IDH non-canonical mutations. One study reported a prolonged survival for patients with IDH1 non-canonical mutations as compared to IDH canonical mutation [15]. Two studies reported a lack of impact in terms of survival for the presence of IDH1 non-canonical mutations without reporting HR with the confidence interval [11,14].

Age
Three of 13 selected studies [11,13,15] reported a significant difference in IDH1 non-canonical mutations according to age. One study reported a significantly younger age only for patients with IDH1 R132C non-canonical mutations (median age 34.9 vs 42.9 years) [13]. The remaining two studies reported a younger age for patients harboring all types of IDH1 noncanonical mutations as compared to IDH1 canonical mutation (median age 35 vs 43 years and 29 vs 39 years) [11,15]. All studies reported the range between younger and older instead of standard deviance as dispersion index, thus a formal comparison with meta-analysis was not possible.

Methylation
Only one study assessed the level of DNA methylation between patients with noncanonical mutation among three cohorts of patients enrolled on CATNON, TCGA, and TAVAREC studies [9,[20][21][22]. In all these cohorts the grade of methylation was significantly higher in patients with IDH non-canonical glioma as compared to patients with IDH1 canonical mutation.

Familiar Risk of Cancer
Only one of the 13 studies selected reported a possible correlation between the incidence of non-canonical IDH1 mutations and familiar risk of cancer [11]. In particular, the familiar incidence of the tumor was significantly higher as compared to those patients harboring an IDH1 R132H mutation (22.2% vs 5.1%).

Localization of Tumors in the CNS
Only one of the 13 studies reported differences in localization of tumors with IDH1 noncanonical mutations as compared to those with IDH1 R132H-mutated gliomas. This study reported that gliomas with IDH1 non-canonical mutations are more frequently localized on the infratentorial region (5.5% vs 0%) and are more frequently multicentric (4.8% vs 0.9%) as compared to tumors with IDH1 R132H mutation.

DISCUSSION
We carried out a systematic review and metaanalysis to assess clinical and biological features of patients with IDH1 non-canonical mutation [7,9,[11][12][13][14][15][16][17][18][19]. To date, this is the first metaanalysis assessing clinical and pathological characteristics of these patients. In our analysis, we find an overall incidence of IDH1 noncanonical mutations of 7.9% (95% CI 5.4-10.7%, I 2 72%). Patients harboring these alterations had a longer survival (HR 0.47, 95% CI 0.28-0.81, I 2 74%) and presented less often the presence of the 1p19q codeletion (risk difference of 31%, 95% CI 23-38%, I 2 6%). Available data suggest that patients with IDH1 non-canonical mutations are younger and have a higher methylation level compared to those patients with IDH1 canonical mutations. Furthermore, one study identified a strong correlation between IDH1 non-canonical mutations and familial risk of cancer.
According to our findings, patients with non-canonical IDH1 mutations seem to have clinical and pathological features suggesting a specific subgroup of grade 2/3 gliomas.
According to the 2021 WHO classification of CNS tumors, all IDH-mutant tumors are classified as IDH-mutant astrocytoma and graded as WHO grade 2, 3, or 4 [23]. This classification recognized a grade 4 IDH mutant astrocytoma and also tumors harboring CDKN2A/B homozygous deletion. To date, there is unclear distinction between gliomas with canonical or non-canonical IDH mutation. Recently, Tesileanu et al. reported a molecular prognostic assessment among patients with 1p19q noncodeleted IDH mutant astrocytoma enrolled in the CATNON phase III trial [24]. The analysis showed that PDGFRA amplification, CDKN2A/B homozygous deletion, pI3K mutation, and total copy number variation load were independently associated with poor prognosis. Taking all these findings together suggests that the molecular characterization of the disease in terms of identification of the IDH mutation is very important because it is able to identify a specific subgroup of patients with an associated prognosis.
The result of our study suggests that the identification of the IDH mutation could assume critical importance since it is associated with a subgroup of patients with improved clinical outcomes. It appears clear that a genomic assessment of tumor specimens should always be performed whenever feasible. Indeed, immunohistochemical analyses alone are not sufficient to provide a complete molecular assessment of the disease and are also associated with a significant percentage of false negative results in case of tumors harboring non-canonical IDH mutations.
There are some limitations to our study. The incidence rate of IDH1 non-canonical mutations can be underestimated as a result of different techniques of genomic sequencing and molecular assessment adopted [25]. Studies selected are mainly retrospective cohort studies, with only one paper assessing three different cohorts of patients (including two randomized controlled trials) and one case-control study [7,9,[11][12][13][14][15][16][17][18][19][20][21][22]. These studies were not generally focused on assessing clinical outcomes of patients with non-canonical mutations. To reduce this potentially confounding factor, we assessed each study for the specific risk of bias adopting the NOS. We identified a low risk of bias for all studies included. Nonetheless, it should be highlighted that the NOS could present some limitations [10,26].
On the survival analysis, we did not include two studies with a negative result on survival comparison. The study by Poetsch et al. [11] did not find a survival advantage for non-canonical mutations. However, the median overall survival was not reached and only 16% of deaths were observed in patients with IDH1 noncanonical mutations, suggesting a still small number of events. Gravendeel et al. [14] failed to show a survival improvement for patients with IDH1 non-canonical mutations in both univariate and multivariate analyses. However, the authors did not report the median follow-up and the number of events that occurred. We did not include these two negative studies since HR (the summary measure selected) was not available [11,14].
It should be noted that survival data from CATNON, TAVAREC, and TCGA reported a survival comparison between IDH1 canonical and IDH non-canonical mutations [9,[20][21][22]. It has been reported that IDH2 mutations are associated with a higher rate of 1p19q codeletion compared to patients with IDH1 mutations. The positive prognostic role of 1p19q codeletion could partially explain the prolonged survival observed in the subgroup with IDH2 alterations [8]. Nonetheless, one study [15] reported a clear survival benefit for patients harboring no IDH1 non-canonical mutations, suggesting that these mutations alone can be associated with prolonged survival.
The positive survival impact observed in our analysis could reflect an overall high grade of DNA methylation [27] as compared to patients with IDH1 canonical mutations.
The DNA methylation reflects the enzymatic activity of IDH. Indeed, IDH mutant dimers shifted the metabolism of the isocitrate to the production of D-2-hydroxyglutarate (D-2-HG) instead of the a-ketoglutarate (which is the usual product of non-mutated IDH enzyme) [6]. The variable intracellular levels of D-2-HG obtained according to IDH mutations could reflect a different methylation status of DNA; however, this should be still demonstrated [28,29]. An increased level of DNA methylation has been demonstrated in patients with IDH non-canonical mutations as compared to those with IDH canonical ones. Nonetheless, when survival analysis was adjusted also for methylation grade the type of IDH mutation maintained a positive effect on survival, suggesting an independent role as a prognostic factor without confounding effect mediated by methylation grade [9].
Another still unknown issue is the sensitivity of non-canonical IDH1 mutations to IDH inhibitors which are currently under investigation in patients with IDH-mutated primary brain tumors [30].
In conclusion, the results of our meta-analysis identify a new class of gliomas with rare non-canonical IDH1 mutation characterized by young age at diagnosis, high level of DNA methylation, and a possible association with a family history of cancer. Furthermore, compared to gliomas with IDH1 canonical mutations, patients with IDH1 non-canonical mutations have a lower rate of 1p19q codeletion and improved survival. These same gliomas are generally diagnosed at a younger age.

CONCLUSION
In this systematic review and meta-analysis, we found that patients with IDH1 non-canonical mutations have an overall incidence of 7.9% among patients with IDH-mutated gliomas. These patients are generally younger, with a lower rate of 1p19q codeletion, and present an increased survival as compared to gliomas with IDH1 canonical mutation.
Our results reinforced the concept of a complete molecular assessment of the disease. A deep genomic assessment is important as there are several molecular factors able to stratify prognosis of patients with glioma. The IDH mutation type is also important since it can be associated with specific clinical and pathological features.