Location of incidental and symptomatic LGGs. A review of comparative studies Authors: Barbara Tomasino, Ilaria Guarracino, Serena D’Agostini, Asgeir S Jakola, Tamara Ius Published: 2026 ABSTRACT Highlights•We aimed at understanding the anatomical distribution of symptomatic and asymptomatic low grade gliomas.•22 cohort studies and 6 case reports were included.•there is a predominance of iLGGs in the frontal lobes.•there is an equal involvement of the left and right hemisphere.•the number of oligodendroglioma were correlated with the number of frontal lesions.Keywords: incidental low-grade glioma (iLGG), Symptomatic glioma, MRI, Frontal lobe Introduction There is a subgroup of diffuse low grade gliomas (LGGs), with a rare incidence (3.8% to 9.6% of the LGGs) (Potts et al., 2012, Eskandary et al., 2005), that are asymptomatic. This condition has attracted the attention of clinicians, as it represents the clinically silent phase of glioma evolution (see Fig. 1) (Pallud and Mandonnet, 2013, Mandonnet et al., 2014a). Indeed, iLGGs are detected during radiological magnetic resonance imaging (MRI) performed for reasons independent of the lesion itself such as trauma, physical examination, headache, dizziness, otolaryngological symptoms, or volunteer studies (Potts et al., 2012, Eskandary et al., 2005, Ius et al., 2020, Pallud et al., 2010). The schematic natural history of a diffuse low-grade glioma is illustrated (adapted from Madonnet et al. [4]). The neuropsychological status and the silent vs. sympthomatic stage are detailed. The time window addressed in the present study is highlighted in red. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) While the molecular aspects of iLGG (Demetz et al., 2023), along with their clinical bases (Potts et al., 2012, Pallud et al., 2010, Ius et al., 2020, Ius et al., 2022) and the related functional imaging patterns (Yang et al., 2023, Cargnelutti et al., 2023, Cargnelutti et al., 2024a), have been addressed in the literature (see Fig. 2), studies on iLGG and the so-called eloquent areas are limited (e.g., Ius et al., 2020). According to the above-mentioned studies, iLGG are more frequent in non-eloquent areas, but this was not extensively explored. One can question whether the non-eloquence characteristic of iLGGs is related to the fact that iLGGs develop preferentially in certain areas, which are considered less eloquent. At variance, the non-eloquence of iLGG may be due to their smaller volume and to their less distinctive connection with critical areas. For instance, in the past the term “silent area” of the brain used in neurosurgical literature referred to the frontal cortex (Drake, 1979). It has been shown that the involvement of critical eloquent regions is affected by histomolecular status, and is present in a biological gradient where IDH wild-type astrocytomas mostly grow in eloquent regions, while oligodendrogliomas tend to develop in least eloquent regions (Gómez Vecchio et al., 2021). The main clinical and molecular characteristics of iLGGs are summarized. In the current review, we aimed to compare lesion sites in patients with iLGG vs symptomatic LGG (sLGG). The frequency of lesion sites was visualized by MRI lesion mask overlay on lobe level (as reported by authors). As secondary endpoint, we investigated the relation between the iLGG histomolecular status and glioma location. Finally, we tested whether there is an association between the number of oligodendroglioma and the number of lesions in the selected iLGG studies. Methods Study design The aim of this review was to evaluate lesion localization in incidentally detected low-grade gliomas (iLGG) by comparing iLGG and sLGG. The study was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA, Page et al., 2021; see Fig. 3A) statement and registered in PROSPERO (ID 1029685). PRISMA flow diagram for the literature search (A) and results from the risk of bias domains analysis (B). Review question The review questions, according to the PRISMA statement, were formulated following the PICO (P: patients; I: intervention; C: comparison; O: outcomes) scheme as follows: In patients with iLGG (P), how does imaging analysis contribute to the neuroanatomical characterization (O) in terms of frequency, subregion localization, and clinical relevance compared to sLGG (C)? Search strategy The review of the literature was done in PubMed, Ovid Embase, Google Scholar and Scopus from 2000 to 4 March 2025 by using a combination of the following search keywords: “glioma” OR “astrocytoma” OR “oligodendroglioma”) AND (“incidental” OR “asymptomatic” OR “natural history”. Two independent reviewers scrutinized the articles. The exclusion criteria were: i) no English language; ii) abstracts with no data; ii) literature reviews and meta-analysis; iii) chapters, commentaries, forum, letter to the Editor; iv) missing methods and/or clinical result details; v) single case series; vi) studies not relevant for the incidental glioma topic; vii) pediatric cases. The two independent authors agreed on the studies to be included in the final dataset. The reference lists of identified studies were reviewed to identify additional relevant studies. Any disagreement in the screening process were solved by consensus, and in each screening step, the agreement among the authors was evaluated using the interobserver correlation coefficient (Cohen’s k coefficient). Screening process Following the initial search, bibliography data in.nbib format were imported into Endnote (Clarivate Analytics, London, UK) for duplicate removal. After this step, all articles were independently screened by two investigators (B.T. and I.G.) by evaluating their titles and abstracts. When there was no consensus, disagreement was solved by a third independent investigator (T.I.). Following the abstract selection, all full-text papers were retrieved and analyzed. All the collected data were saved in an electronic spreadsheet. The final reference list was then checked by T.I. and A.S.J. Included variables For each study included in the review, we extracted the following data: i) lesion site (the raw number of patients with lesion in the frontal, occipital, parietal, temporal lobes, insular cortex or other areas); ii) lesion side (the raw number of patients with lesions in the left and right hemisphere); iii) reason for MRI assessment leading to the discovery of the iLGG; and iv) clinical data (tumor size histological diagnosis, EOR, postoperative deficits: permanent / transient). Data extraction was performed for iLGG and for sLGG when included as comparative group. Operative definition of eloquence We extracted the definition of eloquence reported in each of them as done by the authors. Bias assessment Cochrane’s tools were used to assess methodological quality. The ROBINS-I tool was applied to evaluate the Risk of Bias (RoB) in non-randomized controlled trials (non-RCTs) measured by the screening process (Sterne et al., 2016). The overall RoB was classified as follows: critical, serious, moderate, low, or no information. Two independent investigators (B.T. and I.G.) separately assessed the RoB. In case of disagreement, a third author (T.I.) was involved. Five main domains were identified to describe the risk of systematic errors that might impair the selected papers: Bias due to confounding; Bias in selection of study participants; Bias in classification of interventions; Bias due to deviations from intended interventions; Bias due to missing data; Bias in measurement of outcomes; Bias in selection of the reported result. Statistical analysis Clinical data Clinical data were analyzed using SPSS 21.0 (SPSS, Inc., Chicago, IL). A Shapiro-Wilk test was performed to test whether the variable “area” showed a normal distribution and to select the proper statistical test. The Shapiro-Wilk test showed that the distribution of the variable “area” departed significantly from normality (W = 0.67, W = 0.63, W = 0.59, W = 0.68, W = 0.75, W = 0.66, all p-values < 0.001). Based on this outcome, a non-parametric approach was used. To verify the difference between “areas” in the iLGG group, a Kruskal-Wallis H test was run. For post-hoc analysis, pairwise comparisons using Dunn's test were performed. A Mann-Whitney U Test was applied to test whether iLGG are more frequent in the left or the right hemisphere. Spearman’s Correlations were used to evaluate the relationship between reason for MRI and lesion area, between clinical variables and lesion area and between outcome data and lesion area (Bonferroni correction for multiple comparisons was used). Mann-Whitney U Test was applied to between-group (iLGG vs sLGG) comparisons. MRI structural data The reviewed studies reported the lesion localization at macroscopical level, e.g., in their tables the number of frontal, temporal, insular, parietal lesions were counted. Accordingly, we cannot perform an analysis at microanatomical level. As exploratory MRI structural analyses were performed on anatomical lesion mask mimicking the patients’ lesion site in the selected studies. To test which brain area was the most frequently damaged region in iLGG vs sLGG, we used the WFU_PickAtlas (https://www.nitrc.org/projects/wfu_pickatlas/) to create volumes of interest (VOI) masks for each macroarea corresponding to the lesion site (frontal lobe, temporal lobe, parietal lobe, occipital lobe or insula) reported by the selected papers. The inclusion criteria for such analysis were i) papers which included both iLGG and control sLGG group. The exclusion criteria were ii) papers that reported unspecified localizations (they reported “other areas”) or those with fewer localizations (e.g. basal ganglia, cerebellum, internal capsule, cingulum) since we cannot create VOIs for such unspecified localization. The MRIcron procedure (https://www.nitrc.org/projects/mricron) was used for each single study to overlap the lesion templates for the iLGG group and for the sLGG group. We subtracted one image from another to compare the difference of intensities. This allows to identify if certain regions are more common for one pathology than another. This % procedure, previously employed (Karnath et al., 2004) allows to control for differences in sample sizes (e.g., N = 35 and 197 for iLGG and sLGG, respectively, by using proportional values. For each voxel the percentage of sLGG patients that have a lesion at the voxel is subtracted from the percentage of iLGG patients that have a lesion at the voxel. The result is a map with the percentage relative frequency difference between these two groups for each voxel. The color code indicates the percentage by which a voxel is more frequently damaged in iLGG group a than in sLGG group. We performed a series of comparisons, one for each individual study, as we could not perform a single analysis on the whole sample of iLGG vs sLGG (n = 141 vs n = 642) because of the limited MRIcron procedure calculation power (i.e., 642 lesion masks are not processed). Results Study Selection Study selection process Using a combination of keywords, MeSH and Emtree hierarchical terms, the investigators found 507 potentially relevant articles, which were saved in a unique Pubmed (.nbib) file, which was then imported into Endnote to identify possible duplicates. Duplicates and papers published before 2003 were excluded and 8 studies were deleted. The remaining 499 studies were screened by title, and subsequently by abstract, leading to the exclusion of further 401 studies (Cohen’s k coefficient = 0.94). Additional 3 investigations were removed according to exclusion criteria (Cohen’s k coefficient = 0.92). 95 studies were selected for full-text reading, 73 studies met the exclusion criteria (Cohen’s k coefficient = 0.97). The selection process is summarized in Fig. 3A following the PRISMA guidelines. The findings of this systematic review are summarized in Table 1. Table 1Details from the scrutinized studies.PercentageLesion areaFrontal: 54.07%Insula: 13.99%Occipital: 1.18%Parietal: 9.42%Temporal: 16.46%Other areas: 6.22%HemisphereLeft: 53.13%Right: 46.86%Eloquent area41.95%Reason for MRIHeadache 30.90%Dizziness 5.57%Head trauma14.28%Investigational protocols/research MRI studies/Health check-up 14.28%Sensitive disorder/Gait disturbance/ENT disorders/diplopia/anosmia/hearing disturbance15.80%Other/Incidental discovery10.43%Neurological change/Mood disorders12.96%AgeMean 38.08 ± 3.04 yearsPreoperative Volume (cm3)Mean 25.99 ± 11.5Extent of ResectionGTR, gross total resection 48.35%STR, subtotal resection; 20.75%sTR, supratotal resection. 17.43%PR, partial resection;10.43%Biopsy 3.01%Postoperative DeficitsTransient21.49%Permanent 4.59% Details from the scrutinized studies. Study characteristics In this systematic analysis, we found 22 articles on the role of neuroanatomy in iLGG, 10 of which were comparative studies that allowed between-group analyses (see Fig. 3A for the PRISMA flow-chart). A summary of non-randomized clinical studies included in the Systematic Review is reported in Table 1. The total number of patients sampled in the selected papers was 1093 patients. Information about lesion location was reported in all the evaluated studies. Overall tumors had a mean of 25.9 ± 11.5 cm3volumes. Dataset about different diagnosis are reported by the selected studies as follows: 38.3% oligodendroglioma, 10.4% astrocytoma, 0.8% mixed glioma, 5.7% oligoastrocytoma, 0.1% ganglioglioma, 7.9% WHO II, 0.9% WHO III, 26.1% astrocytoma IDH1/2 mutated and 0.9% other diagnosis. The analysis showed that among these, there were 8 studies (Ius et al., 2020, Gogos et al., 2020, Cargnelutti et al., 2023, Boetto et al., 2021, Ius et al., 2022, Opoku-Darko et al., 2018, Ng et al., 2020a, Demetz et al., 2023) that included a small group of astrocytoma IDH wild-type: 4/34, 11/113, 3/18, 13/101, 34/267, 7/34, 10/47, 4/23 for a total of 8.4% astrocytoma IDH wild-type. Bias assessment Bias assessment was conducted across all 5 domains according to Cochrane guidelines. Bias evaluation data are summarized in Fig. 3B. Bias was mainly due to confounding, and this was considered serious in all cases since all studies were unbalanced (for example, all lack statistical adjustments or randomization). The bias due to confounding is serious and heavily influences the overall assessment. Operative definition of eloquence We extracted the definition of eloquence reported in each of them. The definition is reported in Table 2. Given that the different studies had heterogeneous criteria for defining eloquence, no further analysis was performed on this variable. In detail, of the 23 included studies, 12 did not provide information about the involvement of the eloquent cortex in iLGG lesions. Nine out of twelve studies reported that iLGG lesions were associated with non-eloquent brain regions (see Table 2). Three studies reported that iLGG lesions were primarily located in the frontal lobe, but unfortunately they did not provide any further details (see Table 2). Table 2Eloquent cortex involvement and Frontal lobe involvement.Eloquent cortexFrontal lobeDefinition of eloquencePallud et al., 2010frontal lobes (68% vs 51%) and non-eloquent brain regions (64%) were associated with iLGG, as compared to their sLGG counterparts.Thus, a combinationof limited mass effect, primarily frontal lobeinvolvement, nondominant location (51% vs 48%), and non-eloquencemay explain the absence of clinical symptoms and the observed good KPS.Not definedPotts et al., 2012were less likely to have tumors in eloquent locations (14.3% vs 61.9%, p < 0.001)−eloquence defined as a tumor involving theinternal capsule; basal ganglia; language, sensory, or motorcortices; or hypothalamusZhang et al., 2014lower frequency of eloquent areas involvement (8.69% vs. 57.14%, P < 0.001)−eloquent areas such as cortices of motor, sensory or language areas as well as basal gangliaWang et al., 2020−−−Ius et al., 2020patients with iLGGhad lower frequency of eloquent areas involvement (11.76% vs. 91.03%, P = 0.0001) compared to those with sLGG−Eloquent areas, such as cortices of motor, sensory, or language areas, likethe insular lobeGogos et al., 2020lower frequency of eloquent areas involvement (P = 0.0001)−Cargnelutti 2020Cargnelutti 2021−−−Zuo-Cheng Yang et al., 2023−−−Birladeanu et al., 2023−Their series had a limited mass effect, primarily frontal lobeReported (in Table 4) the (high) percentage (mean 57.13%) of frontal lobe involvement in tumor localization among 13 studies reviewed, but the authors did not comment on thateloquent brain areas: premotor/SMA cortexBoetto et al., 2020−−−Allison et al., 2022It would be logical to include the predominance of non-eloquentlocations of iLGG when compared with sLGG; however, owing to a reduced number of cases, the differenceswere not significant.70% of incidental LGGs and 44% of sLGG were situated within the frontallobeNot definedIus et al., 2021−−−Opoku-Darko et al., 2018−−−Ng et al., 2021−−−Demetz et al., 2023Higher OS in case of a supramarginal resection reported in the literature. This again favors an early resection of iLGG before an infiltration in eloquent areas occurs.−Not definedDuffau (2012)located in or near eloquent regions−Cortical mapping was then performed todetect eloquent areasCochereau et al., 2016−−−Zeng et al., 2021significantly more tumors wereinitially located adjacent to the functional area in the sLGG (16/26) than in the asymptomatic 11/49group (P < 0.05)−tumor location adjacent to the functional area was defined according to a distance between thetumor border and the brain functional area of less than 1 cm as measured on axial, coronal or sagittal sections of MRI images; otherwise, the location was defined as away from the functional area.Lima and Duffau (2015)In all patients (21/21), the cortical and subcortical eloquentareas were detected by intraoperative stimulationIn the present series,again we observed no intraoperative seizures, despite a mean age at time of surgery of 35 years and despite a majority of patients (16/21) with frontal lobe involvement.intraoperative stimulation defined cortical and subcortical eloquentareasLima et al., 2017All gliomas were located in or near eloquent areas within the left dominant hemisphere (supplementary motor area (n = 4), paralimbic system (n = 5;1 insular, 3 fronto-insular, 1 temporo-insular), frontal lobe (n = 6), temporal lobe (n = 2), and parietal lobe (n = 2).−intraoperative stimulation defined cortical and subcortical eloquentareasOpoku-Darko et al., 2019−−−Ng et al., 2020−−− Eloquent cortex involvement and Frontal lobe involvement. iLGG group by lesion area and hemisphere First, we tested whether there is a more frequent lesion area in the iLGG group. There was a significant difference in the dependent variable between the different lesion areas, χ2(5) = 50.46, p < 0.001 (see Fig. 4A), and post-hoc test using a Bonferroni corrected alpha of 0.0033 showed that Frontal lesions were significantly higher than lesions in other locations (median values: Frontal location 16.5, 2.5 for Insula, 0 for Occipital, 3 for Parietal 3 for Temporal and 0 for other areas. This pattern can also be observed in Fig. 4B showing the localization numbers of each single study group. Then, we tested whether iLGGs are more frequent in the left or the right hemisphere. iLGGs are equally reported in both hemispheres (χ2(1) = -0.704, p = 0.495, see Fig. 4C). Lesion site distribution within the iLGGs group (A-B) and in the comparative iLGGs vs sLGGs study (D-E): means and standard deviations of the different brain areas (A and D respectively); single study inspection of the number of cases for each single lesion area. (B and E). Lesion side distribution: means and standard deviations of left and right hemisphere lesions (C and E respectively). Reason for MRI investigation by lesion area The majority of cases performed MRI because of headache (see Table 3). There was a significant positive relationship between headache and F, insula, P and T lesions (rs (18) = 0.918, rs (18) = 0.959, rs (17) = 0.922, and rs (17) = 0.773, all Ps < 0.001 respectively), between dizziness and F and T lobes, rs (10) = 0.899, p < 0.001 and rs (9) = 0.909, p < 0.001 respectively, between Investigational protocols/research MRI studies/Health check-ups and F and T lobes, rs (15) = 0.759, p < 0.001 and rs (14) = 0.822, p < 0.001 respectively. Head trauma significantly and positively correlated with all the areas (rs (17) = 0.826, rs (16) = 0.870, rs (16) = 0.846, rs (15) = 0.830, all Ps < 0.001). Sensitive disorder/Gait disturbance/ENT disorders/diplopia/anosmia/hearing disturbance significantly and positively correlated with T lobe (rs (13) = 0.848, p < 0.001). Table 3Reason for MRI investigation.Reason% of casesHeadache31Sensitive disorder/Gait disturbance/ENT disorders/diplopia/anosmia/hearing disturbance16Head trauma14Investigational protocols/research MRI studies/Health check-up14neurological change/Mood disorders13other/incidental discovery10Dizziness5 Reason for MRI investigation. Clinical variables by lesion area The pre-operative volume did not correlate with any lobe involvement (all P > 0.05), thus the mean lesion volumes were similar (small) independent of the lesion area (see Table 2). Supratotal resections were positively correlated with frontal and temporal involvement (rs (8) = 0.952, and rs (8) = 0.910, respectively, both P < 0.001). Post-operative deficits 13/22 of the scrutinized studies reported information about post-operative permanent deficits; 36% of the studies reported information about post-operative transient deficits. In the studies, 28% of patients presented with transient deficits and 7% with permanent deficits. The lesion area and the lesion volume did not correlate with the outcome data (all Ps > 0.05). iLGG vs sLGG comparative studies 10/22 studies of the scrutinized studies included both iLGG and sLGG data. There is a significant difference in the number of cases between the iLGG and sLGG for the Frontal lesion area (Z = -2.693, p = 0.006, with 62% frontal iLGG vs 48% frontal sLGG, Fig. 4D)), while for the other lesion area the groups did not differ (Z = -1.034, Z = -0.729, Z = -0.788, Z = -0.525 and Z = -0.555 for the insula, occipital, parietal, temporal and other area, respectively, all Ps > 0.05, Fig. 4D shows the localizations of each single study group). iLGGs and sLGGs are equally reported in both hemispheres (Z = -0.57, p = 0.620 and Z = -0.51, p = 0.620 respectively, see Fig. 4E). MRI structural results By filtering the studies according to the inclusion criteria (see Methods), 4/10 comparative studies were suitable for the MRI lesion analysis in order to compare iLGG vs. sLGG (Potts et al., 2012. Zhang et al., 2014, Ius et al., 2020, Zeng et al., 2021) (see Table 4). Table 4Studies suitable for the MRI analysis according to our criteria (see Methods). Higlighted in grey are the studies included in the MRI subtraction analysis.iLGG/sLGGStudyFinsulaOPTotherPallud et al., 201032/63715/612Potts et al., 201223/994/540/03/165/280/0Zhang et al., 201415/1200/111/54/163/440/0Wang et al., 202057/4590/02/417/7429/2660/842Ius et al., 202021/882/681/02/318/360/0Gogos et al., 202166/27315/1231/010/5216/885/8Yang et al., 20237/610/18Allison et al., 202214/360/30/02/53/191/19Demetz et al., 202312/750/00/00/02/559/8Zeng et al., 202121/133/45/37/313/30/0 Studies suitable for the MRI analysis according to our criteria (see Methods). Higlighted in grey are the studies included in the MRI subtraction analysis. For each study the subtractions between the lesion overlay of the sLGG patient group from the lesion overlay of the iLGG group reveals that the frontal lobe is the area maximally involved by lesions in the iLGG patient group (Fig. 5 A-D), as seen in 3/4 studies. (A-D) Subtraction of sLGG from iLGG lesion masks. The percentage of overlapping lesions is illustrated by different color coding according to increasing frequencies (as indicated in the bar code e.g. a value of 40 reflects that that voxel is damaged 40% more frequently in iLGG patients than in sLGG patients) for the 4 studies selected for the MRI analysis: Potts et al., 2012 1 (A), Zhang 2014 18 (B), Ius 2020 5 (C) and Zeng 2021 21 (D). (E-H) Correlation between the number of oligodendroglioma and the number of lesions in frontal (E), insular (F), parietal (G), and temporal (H) sites in the selected studies. Histomolecular status by area We tested whether there is an association between the number of oligodendroglioma and the number of lesions in the selected studies. We found that the number of oligodendroglioma significantly correlated with the number of frontal (rs (17) = 0.911, p < 0.001), insular (rs (17) = 0.726, p < 0.01), parietal (rs (16) = 0.832, p < 0.001) and temporal (rs (16) = 0.903, p < 0.001) lesions, albeit with a stronger relation with frontal lesions (see Fig. 5E-H). Single case series Our literature search returned 6 single case studies on iLGG. We next report on this case series. Again, the frontal lobe was the most commonly involved lobe (5/6 studies). Interestingly, gliomas occurred more frequently in the right hemisphere (5/6). Further details can be found in Table 5. Table 5Details for the single case series of iLGG.StudyFinsulaOPTotheragesexReason for MRIhistologyhemiCochereau et al., 20161143fheadacheslggrhDimou and Kelly (2020)125fheadachesglioblastomarhGuarracino et al. (2022)130fmigraineoligodendrogliomarhAn et al., 2019151msuspicion of Parkinson diseaseanaplastic oligodendrogliomalhSachdev and Rees (2014)125mmultiple sclerosislow-grade astrocytoma.rhIacoangeli et al., 2012143fvestibular schwannomaoligodendrogliomarh Details for the single case series of iLGG. Discussion The present review assessed the neuroanatomical correlates of iLGG. The within-group analysis demonstrated a predominance of iLGG in the frontal lobe. This pattern was confirmed by the statistical analyses we run on the reviewed literature. There was a heterogeneous definition of eloquence in the reviewed studies. Despite the different operative criteria for defining eloquence, the reviewed studies agree on the fact that iLGG preferentially involve “silent” brain regions. In our study it emerged that this pattern has a neuroanatomical counterpart, namely the frontal lobe. This anatomical predominance is clearly visualized by the lesion-mask subtraction analysis showing that the frontal lobe is a predilection site in the iLGG group vs. sLGG group in 3/4 studies. The frontal lobe was also involved in 5/6 of the iLGG case reports. One may argue that the frontal cortex is commonly involved in gliomas because of its size. However, the comparative analyses showed that the number of frontal lobe lesions is significantly higher in iLGG. We first acknowledge that the classification of lesions by lobes (frontal, parietal, etc.) had some limitation. It is performed in the context of the present large-scale review; obviously it ignores known functional and developmental heterogeneity. Indeed our analysis cannot account for gyral localization, e.g., SFG/SMA vs IFG, which influences eloquence and clinical impact, or mesial vs lateral distinction in the temporal lobe which is critical given their different ontogenetic and functional profiles (e.g., allocortex vs neocortex). If any finer anatomical detail were available one could have postulated a “likelihood of symptoms”; this would have needed also an agreed map of eloquent brain that is likely to be symptomatic when involved by a lesion. Given this theoretical likelihood map, one would have expected that a large part of the frontal lobe would fall in the low likelihood region, while smaller parts of other brain regions would have a low likelihood of symptoms. Following this line of speculation, the distribution of brain iLGGs may have been explained by this map. Future multicentric studies including individual lesion mapping, standardized definitions of eloquence, and connectivity-based analyses will be necessary to address this issue. Nonetheless our findings may have some clinical implications. Our results suggest that it would be more informative to report the brain area mostly involved by iLGG of the presented series instead of reporting “non-eloquent area involvement”. This is because the frontal lobe is neither “non-eloquent” nor “silent” from a functional point of view. Lesions in these patients are discovered incidentally, nonetheless cognitive evidence based on three neuropsychological studies on iLGG (Ng et al., 2020; Cochereau et al., 2016a, Guarracino et al., 2020) is that mild impairments are detected. These are often subclinical (Ng et al., 2020) and likely compensated for (Cargnelutti et al., 2024), and the functional daily life ability is not overwhelmed. iLGG patients show mild decrements mainly on the neuropsychological tests measuring frontal lobe-related functions. For example, mild to borderline difficulties were found (Guarracino et al., 2020) in some patients mainly in terms of executive functions (working memory and short-term memory tests). Another study (Ng et al., 2020b) showed that the most impaired domain was episodic memory, followed by phonological and categorical fluency task all functions related to the frontal lobe. In their study, they found 38.3% of frontal lobe lesions. Interestingly, they reported a significant correlation between presurgical executive functioning and tumor volume. Lastly, it was reported (Cochereau et al., 2016b) that 60% of iLGG patients had cognitive disturbance. Executive functions were altered in 53% of the cases. The most commonly affected tasks were the Stroop task and phonological fluency. In their study, 10/15 of patients had a frontal lesion. In this regard, they stated the importance of the neuropsychological evaluation –which is the sole indication of the patient’s functional status– to determine whether the lesion is epileptically or cognitively silent, especially for patients with iLGG, who do not have epilepsy yet. They argued that an incidental finding does not imply that neuropsychological functioning is preserved. One may think that iLGGs go (cognitively) unnoticed because they occur mainly in the non dominant hemisphere, with less evident symptoms compared to language emerging difficulties that follow dominant hemisphere lesions. However, based on available studies, we found that iLGGs are equally reported for dominant and non dominant hemisphere. This symmetry implies incidental findings occur via routine imaging without the presence of symptoms favoring one side. The lack of hemispheric bias aligns with findings that incidental LGGs are often smaller, expansive rather than infiltrative, and less likely to involve eloquent areas regardless of side. This uniformity supports the view that iLGG onset appears independent of dominance (Demetz et al., 2023). The limitations acknowledged above of the “lobar” approach apply also to the hemisphere issue, considering that, inside the frontal lobe, iLGG could have a different gyral distribution compared to sLGG. It is unquestionable that iLGGs have typically a smaller volume than the general series of sLGG; nonetheless we found no significant correlation between the pre-operative volume and the lesion area. This can help rule out the possibility that frontal iLGGs go cognitively unnoticed simply because they are smaller in volume. Another limitation involves the attribution of symptoms. In secondary analyses of published studies, vague complaints such as headaches cannot be reliably associated with the presence of a tumor. As a result, reported reasons for performing MRI should be viewed as triggers for detection rather than as direct symptoms caused by the lesion. We also found a significant positive relationship because of dizziness and frontal and temporal lobe lesions. It is known that dizziness can reflect frontal lobe seizures (e.g., Hewett and Bartolomei, 2013, Jiang and Zhou, 2020) or frontal white matter damage (Ibitoye et al., 2022) affecting balance and vestibular function, among other causes. Across the relevant articles scrutinized, two categories for MRI reason, namely “other/incidental” and “neurological change/mood disorders” worth investigating. This is because a study reported patients’ subjective complaints – something which no other iLGG study focused on (Cochereau et al., 2016b). The most frequently reported complaints (Cochereau et al., 2016b) were tiredness and concentration/sustained attention difficulties. It is questionable whether beyond the two not clear categories some similar symptoms may be hidden. This would be in line with studies showing that many LGG patients start sick leave 6 months before surgery and before their functional abilities start to objectively decrease (Rydén et al., 2020). Similar data are reported in another LGG study showing that 30% of LGG patients had adjusted workload or tasks as early as 1 year prior to LGG diagnosis (Smits et al., 2015). It may be hypothesized that iLGG patients tested with complex and higher order cognitive tasks would show a more consistently impaired performance when the clinical impact of iLGG is still low. The number of supratotal resections was positively correlated with the number of frontal lobe and temporal lobe lesions. As to the former result, we speculate that this result may be related to the fact that the frontal lobe is traditionally held to be silent. Thus resection is more extensive. However, it should be noted that non-dominant frontal lobe resections are not functionally neutral and that cognitive sequelae (e.g., executive dysfunction, apathy) may emerge even with classical eloquent mapping findings. This can induce more cautious resections. Results are less clear for the temporal lobe, although there are vast experience in performing temporal lobectomies in epilepsy surgery. Further, this be tailored to function thanks to the intraoperative monitoring and DES mapping, in in some cases the extent of resections achieved in the temporal lobe have reached the supratotal level. Further facilitating supratotal resection is that the glioma volumes were relatively small, with subsequent greater chances of a supratotal resection. However, it should be noted that given the distinction between mesial and lateral temporal resections and its functional consequences, a “supratotal” resection in the temporal lobe is anatomically and functionally context-dependent. As regards the histomolecular status, we found that the number of oligodendroglioma strongly related with frontal lesions. This result is in line with earlier studies performed on LGG cohort irrespective of being sLGG or iLGG (Suh et al., 2019, Gómez Vecchio et al., 2021). Unfortunately, our analysis did not allow to stratify the histomolecular status by single lesion area. The slower growth rate of oligodendrogliomas allows for brain adaption and provides a longer time-frame of possible detection prior to symptoms develop. In part, the predilection for oligodendrogliomas has implications for prognosis, and is one possible bias that needs to be addressed in future studies. In general, the interaction between histomolecular status, location and iLGG should be better accounted for. We would like to briefly speculate on how the pattern we observed can be interpreted. The pattern reported in the present study does not reveal any information about plasticity. Nonetheless plasticity could be one of the factors for interpreting the results. To begin with, the frontal lobe is the last part of the brain to fully develop and reaches maturity in the mid-to-late 20 s (Kolk and Rakic, 2022). Studies measuring the plasticity potential of the different brain areas showed that the frontal cortex has indeed a high index of compensation (Herbet et al., 2016). How does this relate to iLGG? Szalisznyo et al. (2013) showed that a patient with iLGG had a preserved cognitive status when the first seizure occurred after 1 year from diagnosis and argued that the good neuropsychological performance was due to plasticity. iLGG showed activation in a contralesional site in the homolog counterpart of the inferior/middle frontal gyrus which was not activated in the sLGG with comparably small glioma (Cargnelutti et al., 2024b). An iLGG is in the so-called clinically silent stage (e.g., Mandonnet et al., 2014b) and compensatory reorganization peaks at the beginning of the symptomatic stage (Szalisznyo et al., 2013). Most changes occurred before the onset of the symptomatic phase (Szalisznyo et al., 2013). Considering that LGG patients are generally young and that iLGG patients are even younger than sLGG ones (e.g.,Pallud et al., 2010), it has been argued that glioma develops and extends at the maximal plasticity potential and while the brain structures are still in maturation (Duffau, 2022) because of the myelin maturation processes. Limitations of the study This study has limitations. As to histomolecular status, data on iLGGs vs sLGGs localization are pooled at the study level, not at the patient level, and this could limit the interpretation of our results. Second, the eloquence definitions in the different studies is heterogeneous, limiting further analysis on this variable aside reporting at a qualitative level the different definitions. Heterogeneity is fund also in the histology classification across studies, limiting the interpretation of our results. A small proportion (about 8%) of iLGG included in this review were reported as histologically low-grade yet IDH-wildtype. Under WHO 2021 criteria, these would not be considered low-grade entities. Their inclusion reflects historical classifications and incomplete molecular profiling in earlier series, and should be interpreted cautiously. With the exception of Demetz et al. (2023), all relevant studies were published before the last WHO CNS 2021 classification and, in most cases, lacked comprehensive molecular profiling—including TERT promoter mutation status, EGFR amplification, and combined chromosome 7 gain/chromosome 10 loss—necessary for definitive reclassification as glioblastoma, IDH-wildtype. Additionally, in the majority of cases, tumor location was reported as pooled counts across histologies, lacking distinct stratification by both anatomical site and histological subtype. This point prevents the selected exclusion of IDH-wildtype patients from the neuroanatomical analysis, raising a recognized bias. Several of these series, however, described an overall clinical course: overall survival events were absent, progression-free survival events were rare and typically occurred only after prolonged follow-up, and adjuvant radiotherapy or chemotherapy was generally not administered. This slow progressive course, especially considering the absence of early tumor progression, makes it improbable that these IDH-wildtype cases exemplified the aggressive phenotype commonly termed “early” or “baby” glioblastoma in the clinical setting of iLGG. Their inclusion therefore signifies historical classification methods used at the time of original studies, rather than contemporary molecular taxonomy. This limitation has been acknowledged, while preserving the integrity and completeness of the original study designs (Rudà et al., 2022, Ius et al., 2022). Moreover, this potential bias is counterbalanced by the consistent reports across all studies indicating proonged survival with minimal or no progression events, indirectly reinforcing thus the idea that these cases are unlikely to represent glioblastoma. Furthermore, the absence of most genotyping data, aside from IDH status, is a striking limitation, as this information is now the basis for the diagnosis of LGG. Lastly, as previously mentioned, we acknowledge a potential anatomical bias due to lobar classification. Accordingly, the anatomical part of this review relies on what is reported in published studies which is heterogenous and commonly restricted to brain lobes. Deeper anatomical assessments, hence, are not possible. Conclusion To conclude, iLGGs have a higher representation in the frontal lobe compared to sLGGs. Future investigations should confirm these findings through prospective, multicenter studies using standardized neuropsychological tests, lesion-symptom mapping, histomolecular analyses and advanced brain connectivity analyses. This will enhance the comprehensive understanding of functional impairments and compensatory neural mechanisms associated with incidental gliomas. Author Contributions B.T. and T.I. designed the study; I.G., B.T and T.I. extracted the data; B.T. analyzed the data; B.T. wrote the paper; T.I, S.d’A and A.J. revised it critically for important intellectual content; all authors edited the paper; all authors revised the final version of the manuscript. CRediT authorship contribution statement Barbara Tomasino: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Data curation, Conceptualization. Ilaria Guarracino: Writing – review & editing, Writing – original draft, Investigation, Data curation. Serena D’Agostini: Writing – review & editing, Methodology, Data curation. Asgeir S. Jakola: Writing – review & editing, Writing – original draft, Validation, Supervision, Investigation, Data curation. Tamara Ius: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This work was supported by the Italian Ministry of Health (Ricerca Corrente 2025 and 2026 to B.T.). Contributor Information Barbara Tomasino, Email: barbara.tomasino@lanostrafamiglia.it. Ilaria Guarracino, Email: ilaria.guarracino@lanostrafamiglia.it. Serena D’Agostini, Email: serena.dagostini@asufc.sanita.fvg.it. Asgeir S. Jakola, Email: jakola.asgeir@gu.se. Tamara Ius, Email: tamara.ius@unipd.it. Data availability Data will be made available on request. References Associated Data Data Availability Statement Data will be made available on request.