From Cellular Origin to Clinical Recurrence: Rethinking Tumor Boundaries in Isocitrate Dehydrogenase-Mutant Gliomas

Abstract

Isocitrate dehydrogenase (IDH)-mutant gliomas are primary malignant brain tumors that predominantly affect young adults, and are characterized by relatively favorable clinical outcomes compared with other brain malignancies. Despite advances in molecular classification and multimodal treatment, tumor recurrence remains inevitable in most patients, raising fundamental questions about the true extent of disease and the biological basis of recurrence. A recent study identified rare IDH1 mutation-bearing cells within histologically and radiologically normal peritumoral cortex, suggesting that the earliest stages of gliomagenesis may extend beyond the conventionally defined tumor margin. This observation suggests that these mutation-bearing cortical cells may contribute to tumor recurrence even after gross total resection and supports the potential value of early therapeutic intervention targeting these latent cells, such as with IDH inhibitors, to delay disease progression. Collectively, these insights suggest that IDH-mutant gliomas may be better understood as a spatially extended biological process rather than a strictly localized lesion. A more integrated approach that combines precise surgical techniques with molecularly informed therapies may improve long-term disease control.

INTRODUCTION

Isocitrate dehydrogenase (IDH)-mutant gliomas are the most common primary malignant brain tumors in young adults under the age of 50 [1]. The 2016 World Health Organization (WHO) classification of Central Nervous System (CNS) tumors incorporated molecular parameters into diagnostic criteria [2]. Since then, gliomas have been defined primarily by their genetic alterations. In the current WHO CNS 5th edition, IDH-mutant gliomas are classified into astrocytoma with intact 1p/19q, spanning grades 2 to 4, and oligodendroglioma with 1p/19q codeletion, ranging from grades 2 to 3 [3]. Compared with IDH-wildtype glioblastoma, these tumors generally exhibit a more indolent clinical course and relatively favorable prognosis, with median survival ranging from 7 to 18.4 years [4,5,6]. Despite these favorable features, however, IDH-mutant gliomas remain characterized by inevitable progression, underscoring the need for improved therapeutic strategies [1,7,8,9].

One of the typical features of these tumors is their diffusely infiltrative growth pattern, which results in poorly demarcated tumor boundaries [10]. In current clinical practice, treatment decisions, including the extent of surgical resection, delineation of treatment fields, and interpretation of residual or recurrent disease, are largely guided by radiographic and histopathological assessments [8,11,12]. However, the extent to which these approaches accurately reflect the true biological distribution of tumor cells remains uncertain. This limitation highlights a fundamental gap between the clinically defined tumor and its underlying biology [13]. In particular, conventional imaging and histological evaluation may fail to capture the earliest stages of gliomagenesis or the full spatial extent of mutation-bearing cells [14,15,16]. As a result, current treatments have largely focused on the MRI-visible tumor, although the underlying biological footprint may extend more diffusely beyond it [11].

Recent advances in genomic and spatial profiling technologies have begun to challenge this tumor-centric view [17]. Emerging evidence suggests that cells harboring glioma-associated driver mutations may exist beyond the radiologically defined tumor mass, within brain regions that appear normal [18,19,20,21]. These observations raise the possibility that IDH-mutant gliomagenesis is not confined to a single lesion, but instead represents a multistep and spatially distributed process [22,23,24,25,26,27,28,29,30,31]. This review discusses recent findings that redefine the spatial and biological boundaries of IDH-mutant gliomas and explore their implications for clinical decision-making.

GLIOMA-INITIATING CELLS EXTEND BEYOND THE TUMOR MARGIN

In our recent study, we sought to determine whether the earliest stages of IDH-mutant gliomagenesis are confined to the radiologically defined tumor mass or extend beyond it into the surrounding brain. To address this, we performed deep targeted sequencing of the peritumoral cortex, defined as histologically and radiologically normal brain tissue located approximately 10 mm beyond the tumor margin, in a cohort of 32 patients with IDH-mutant gliomas, including 29 normal peritumoral cortex samples and 10 normal subventricular zone samples.

Remarkably, we identified rare cells harboring the canonical IDH1^R132H mutation in the peritumoral cortex of 37.9% (11 of 29) of cases [32]. These IDH1 mutation-bearing cells were present despite the absence of overt histopathological abnormalities, indicating that the earliest genetic events in gliomagenesis can precede detectable tumor formation by conventional diagnostic modalities. Importantly, these cells are likely to represent early “seeds” of glioma. That is, they harbor the founding driver mutation, IDH1^R132H, but have not yet acquired the full set of alterations required for malignant transformation. In practical terms, this suggests that the biological origin of the tumor may exist in regions of the brain that appear completely normal on routine imaging and histological examination [32].

To determine the clonal relationship between these peritumoral mutant cells and the tumor mass, we performed phylogenetic analyses based on shared somatic variants. These analyses demonstrated that the mutation-bearing cells in the peritumoral cortex and the tumor cells share a common ancestral origin [32]. Notably, while these cells harbored the founding IDH1 mutation and a subset of passenger mutations, they lacked the full complement of additional genetic alterations typically observed in established tumor tissue. This indicates that they represent early-stage clones along the evolutionary trajectory of IDH-mutant gliomas, rather than infiltrating tumor cells. From a clinical perspective, this distinction is important. These cells are not simply tumor cells that have spread beyond the margin, but rather a biologically earlier population that has the potential to give rise to tumor under appropriate conditions [32]. A representative longitudinal case in our cohort further illustrates this concept (Fig. 1A). In this case, phylogenetic analysis suggested that the recurrent tumor was more closely related to mutation-bearing cells in the peritumoral cortex than to the primary tumor (Fig. 1B). This observation is consistent with an alternative model in which both the primary and recurrent tumors originate from a common ancestral clone harboring IDH1 and TERT promoter mutations with 1p/19q codeletion, which itself may have evolved from an earlier clone in the peritumoral region carrying only the IDH1^R132H mutation [32]. These findings may help explain why recurrence can occur even after apparently complete resection of the visible lesion [33].

Fig. 1. Spatial mismatch between radiological margins and the biological extent of IDH-mutant gliomas. A: Representative MRI from a longitudinal case in our cohort showing the radiologically defined tumor margin and sampling regions. B: Schematic representation of the clonal relationship from this representative case, suggesting a common ancestral origin between tumors and peritumoral cortical cells. C: Conceptual illustration of the spatial distribution of tumor cells and mutation-bearing cells in IDH-mutant gliomas. While the gross tumor mass is confined within the radiologically defined margin, mutation-bearing cells can be present in the surrounding peritumoral cortex. These cells may represent early-stage clones that harbor the founding IDH mutation but lack full malignant features. This spatial discrepancy highlights the potential limitation of imaging-based surgical boundaries and provides a biological basis for tumor recurrence despite gross total resection. IDH, isocitrate dehydrogenase. Created in BioRender. Lee, J. H. L. (2026) https://BioRender.com/h3vgqh8.

To further define the cellular identity of these mutant cells, we employed a multimodal approach integrating cell-type-specific genotyping and spatial transcriptomics. Among the 11 peritumoral cortex samples in which IDH1 mutation-bearing cells were detected, six had sufficient residual tissue available for these downstream analyses. Across six patient samples, IDH1^R132H cells in the peritumoral cortex were consistently found within glial progenitor cell (GPC) populations, including oligodendrocyte progenitor cells (OPCs) [32]. Importantly, these cells were interspersed among normal brain cells and did not exhibit overt morphological atypia, reinforcing the notion that they exist in a latent, non-malignant state despite harboring a canonical driver mutation. Spatially, these cells were scattered throughout the normal parenchyma rather than forming discrete tumor-like mass structures. This pattern further supports the idea that they are not simply extensions of the tumor mass, but instead represent a distributed reservoir of mutation-bearing cells within the brain. Clinically, this suggests that the concept of a sharply defined tumor boundary may not fully capture the true biological extent of disease [11].

To evaluate the functional relevance of these findings, we developed a genetically engineered mouse model in which mouse Idh1^R132H mutations and other drivers were introduced into OPC-lineage cells. Tumors arising in this model recapitulated key molecular and cellular features of human IDH-mutant gliomas, including subtype-specific transcriptional programs and characteristic cellular heterogeneity. These results support the notion that GPCs harboring IDH mutations can serve as a cellular substrate for glioma development [32].

Taken together, these findings indicate that glioma-initiating clones can reside beyond the radiologically defined tumor margin, within brain regions that appear normal on conventional imaging. In other words, the “true” biological field of glioma initiation may extend beyond the lesion that is visible to current diagnostic modalities. This has important implications for how we interpret tumor margins, understand recurrence, and design therapeutic strategies in patients with IDH-mutant gliomas.

RETHINKING SURGICAL MARGINS IN GLIOMA

In many solid tumors, such as those of the stomach, colon, skin, and liver, surgical principles emphasize achieving adequate resection margins beyond the gross tumor to ensure complete removal of microscopic disease [34,35,36,37,38]. In contrast, the surgical management of brain tumors presents a fundamentally different challenge. Due to the functional complexity of the brain, neurosurgeons must balance maximal tumor resection with preservation of neurological function [39,40]. As a result, although the infiltrative biology of these tumors suggests a broader disease extent than is apparent on conventional imaging, the actual extent of resection is constrained by anatomical and functional considerations (Table 1 and Fig. 1C) [41,42]. These considerations highlight the need for more precise intraoperative strategies to define biologically relevant resection margins. The development of real-time, molecularly informed intraoperative monitoring modalities may provide a path toward more accurate and personalized surgical approaches in glioma treatment [15,43,44]. Such approaches may help define biologically relevant disease beyond conventional imaging-based boundaries while preserving neurological function.

Table 1. Representative studies reporting mutation-bearing cells beyond conventionally defined tumor boundaries.

Study	Diagnosis	Key finding	Potential clinical relevance
Lee et al. [26], 2020	Astrocytoma	Rare somatic mosaicism of the same gene (IDH1R132H) in two siblings with glioma	Developmental predisposition to gliomagenesis
Underhill et al. [18], 2024	Astrocytoma, oligodendroglioma, glioblastoma	Mutant cells in unresected peritumoral edema	Residual infiltrative disease beyond radiographic margins
Park et al. [32], 2026	Astrocytoma, oligodendroglioma	IDH1 mutant GPCs in peritumoral area	Possible source of recurrence and disease persistence

IDH, isocitrate dehydrogenase; GPC, glial progenitor cell.

FROM RESIDUAL DISEASE TO RECURRENCE: IMPLICATIONS FOR EARLY INTERVENTION

Tumor recurrence in IDH-mutant gliomas has traditionally been attributed to residual tumor cells that persist following surgical resection [9,45]. However, the presence of IDH1 mutant GPCs in histologically normal peritumoral cortex suggests that recurrence may, in some cases, arise not only from residual tumor cells but also from a broader reservoir of mutation-bearing cells [32]. In this context, early therapeutic intervention targeting latent IDH1 mutation-bearing cell populations could modulate an epigenetically altered, yet non-malignant, cellular population that persists after surgery. However, further preclinical and clinical studies are required before this concept can be generalized across the broader population of patients with IDH-mutant gliomas [22]. Recent clinical studies of IDH inhibitors, such as vorasidenib, have demonstrated the ability to delay disease progression in patients with IDH-mutant low-grade gliomas [7,46]. These findings raise the possibility that targeting the latent mutant cell population at an early stage may influence the trajectory of disease evolution.

However, the optimal timing, patient selection, and therapeutic combinations for IDH-targeted treatment remain to be established. In high-grade disease, where additional genetic alterations and therapy-induced clonal selection pressure shape more aggressive tumor behavior, combination strategies incorporating IDH inhibition alongside conventional modalities may be required [47,48,49].

Taken together, these observations suggest that a deeper understanding of the temporal and spatial dynamics of IDH mutant cell populations may inform more effective, stage-specific therapeutic approaches, although further clinical investigation is needed to validate these concepts.

DISCUSSION

The findings summarized in this review suggest that the biological extent of IDH-mutant gliomas may not be fully captured by conventional radiographic or histopathological assessments. The presence of latent IDH1 mutation-bearing cells in histologically normal peritumoral cortex indicates that glioma initiation can occur beyond the visible tumor margin and may precede clinically detectable disease [32].

These observations highlight a fundamental limitation of current surgical strategies that are guided primarily by imaging-defined boundaries. Although maximal safe resection remains the standard of care, complete removal of all biologically relevant disease may not be achievable based on anatomical margins (Fig. 1C). This may partly explain why tumor recurrence occurs even after apparently complete resection.

In addition, these findings suggest that tumor recurrence in IDH-mutant gliomas may arise from multiple sources, including both residual tumor cells and a broader population of mutation-bearing cells in the peritumoral cortex (Table 1). Previous studies have progressively expanded the concept of glioma-associated mutant cells beyond the conventionally defined tumor boundary. Lee et al. [26] demonstrated developmental somatic mosaicism of IDH1^R132H in two siblings with IDH-mutant astrocytoma, suggesting that predisposing mutant cells may arise during embryonic development. Underhill et al. [18] subsequently identified glioma-associated mutant cells within unresected peritumoral edema across multiple glioma subtypes, highlighting that molecularly abnormal cells may extend beyond radiographic disease. Together, these studies indicate that genetically altered cells can exist outside the gross tumor mass through distinct biological mechanisms. Our recent findings complement these observations by identifying latent IDH1 mutation-bearing GPCs within histologically normal peritumoral cortex [32]. Unlike developmental mosaicism or infiltrative tumor cells, these cells appear to represent an early stage in tumorigenesis. Collectively, these studies support the concept that the biological extent of IDH-mutant gliomas may extend beyond conventionally defined tumor boundaries. This expanded view of disease distribution implies that therapeutic strategies may need to extend beyond the resection cavity. Recent clinical studies of IDH inhibitors have shown activity, particularly in low-grade tumors, raising interest in whether earlier intervention against latent mutation-bearing cell populations could influence disease progression [7,46,47]. However, further studies are required to define the optimal timing, patient selection, tumor grades, and integration with existing treatment modalities [22,48,49,50].

Importantly, IDH1 mutation-bearing cells were detected in only 37.9% of cases in our cohort, and their true prevalence remains uncertain because of spatial sampling limitations and the technical sensitivity of current detection methods. Therefore, additional studies are required before these findings can be generalized across the broader population of patients with IDH-mutant gliomas.

While these findings expand the biological concept of tumor extent, they do not diminish the established importance of maximal safe resection, which remains a cornerstone of glioma treatment [40,46,51]. Rather, they underscore the need to integrate precise surgical techniques with molecularly informed therapeutic approaches. Furthermore, future efforts should focus on developing methods to better define the true biological extent of disease and to translate biological insights into more precise and personalized treatment strategies.

CONCLUSION

Recent evidence indicates that the biological extent of IDH-mutant gliomas may extend beyond radiographically defined tumor margins, with mutation-bearing cells present in the peritumoral brain tissue. These findings provide a potential explanation for tumor recurrence despite gross total resection and highlight the need for more precise, biologically informed therapeutic strategies. Integrating surgical and molecular approaches may improve the management and long-term outcomes of patients with IDH-mutant gliomas.

Availability of Data and Material

No usage of data and materials for this review.