Non-exonuclease domain POLE mutations and their biological significance in colorectal cancer

Abstract

Recent advances in colorectal cancer (CRC) genomics have expanded the understanding of molecular alterations beyond the traditional mismatch repair/microsatellite instability framework. While exonuclease domain mutations (EDMs) in DNA polymerase epsilon (POLE) are established drivers of ultra-mutated, immunogenic tumors with favorable responses to immune checkpoint inhibition, growing evidence suggests that non-EDMs may also carry biological and clinical significance. Emerging data indicate that non-EDM POLE variants are frequently identified in microsatellite stable (MSS) or MS instability (MSI)-low CRC and are often associated with intermediate or high tumor mutational burden, alterations in DNA repair genes, and recurrent co-mutational patterns involving KRAS, phosphatidylinositol 3-kinase catalytic subunit alpha, MutL homolog 3, and MutS homolog 3. These findings challenge the current binary classification of CRC based solely on MSI/mismatch repair status and suggest that a subset of MSS tumors may harbor previously underrecognized genomic instability and immunogenic potential. In this opinion review, we discuss the evolving molecular landscape of non-EDM POLE-mutant CRC, focusing on their interaction with DNA repair pathways, co-mutational profiles, and possible implications for biomarker interpretation and precision oncology. We further address the current limitations of the evidence, including the lack of functional validation for many variants and the absence of prospective clinical data. Although non-EDM POLE mutations should not yet guide treatment decisions in isolation, their integration into broader molecular frameworks may improve biological stratification and support future personalized therapeutic approaches in CRC.

Key Words: Non-exonuclease; Colorectal cancer; Molecular; Microsatellite stable cancers; DNA polymerase epsilon variants

Core Tip: Non-exonuclease domain (EDM) DNA polymerase epsilon (POLE) mutations represent an emerging molecular subgroup in colorectal cancer. Particularly in microsatellite stable tumors, these variants co-occur with elevated mutational burden, altered DNA repair, and co-mutational profiles (KRAS, phosphatidylinositol 3-kinase catalytic subunit alpha, MutL homolog 3, and MutS homolog 3). Rather than acting as isolated biomarkers, non-EDM POLE mutations appear to influence tumor biology through their broader genomic context, highlighting the need for comprehensive biomarker models instead of relying solely on traditional mismatch repair status.

INTRODUCTION

Colorectal cancer (CRC) remains one of the most prevalent malignancies worldwide and a leading cause of cancer-related mortality, reflecting both its biological heterogeneity and the persistent limitations in early detection and effective systemic treatment strategies[1-4]. Over the past decade, advances in molecular characterization have reshaped the classification of CRC from a primarily histopathological disease into a genomically defined entity, enabling more precise prognostic stratification and therapeutic decision-making[5-7]. Central to this paradigm is the distinction between tumors based on mismatch repair (MMR) status and microsatellite instability (MSI), which has become a cornerstone of modern CRC taxonomy.

Tumors exhibiting deficient MMR (dMMR) and high levels of MSI (MSI-H) are characterized by hypermutation, increased neoantigen load, and enhanced immune infiltration, rendering them particularly responsive to immune checkpoint inhibition[8-10]. By contrast, most CRC cases (approximately 85%) are MS stable (MSS) or display low levels of MSI (MSI-L) and are typically associated with lower tumor mutational burden (TMB), reduced immunogenicity, and limited responsiveness to immunotherapy[11-13]. This large subgroup of MSS/MSI-L CRC represents a major unmet clinical need, particularly in younger patients and those with aggressive disease phenotypes, where effective precision medicine strategies remain scarce[14-16].

Beyond MSI status, large-scale genomic and transcriptomic analyses have revealed additional layers of complexity within CRC, including recurrent alterations in key oncogenic pathways such as mitogen-activated protein kinase (MAPK), phosphoinositide 3-kinase (PI3K), WNT, and transforming growth factor beta signaling, as well as distinct molecular subtypes with differing biological behaviors and clinical outcomes[17-19]. Despite these advances, current classification systems fail to fully capture the diversity of DNA replication and repair defects that may underlie tumor evolution, particularly in MSS tumors lacking canonical biomarkers of genomic instability.

Among emerging molecular determinants, mutations in the DNA polymerase epsilon (POLE) gene have gained increasing attention due to their role in maintaining replication fidelity through exonuclease-mediated proofreading activity[20-22]. Pathogenic mutations within the exonuclease domain (EDM) of POLE are well established as drivers of an ultra-mutated phenotype, characterized by exceptionally high TMB and favorable responses to immunotherapy across multiple tumor types, including CRC[23-25]. Consequently, POLE EDM mutations are now recognized as clinically actionable biomarkers in select contexts.

However, growing evidence suggests that alterations outside the EDM-so-called non-EDM-may also contribute to tumorigenesis through alternative mechanisms that do not involve direct loss of proofreading activity[26-28]. These variants are more frequently observed than classical EDMs, particularly within MSS/MSI-L CRC, and may be associated with intermediate or high mutational burdens, co-occurring DNA repair defects, and distinct clinical behaviors[29-31]. Despite their prevalence, non-EDM POLE mutations remain poorly characterized and are not routinely incorporated into current diagnostic or therapeutic frameworks.

This gap in understanding is clinically significant. As most patients with CRC fall into the MSS/MSI-L category, identifying novel biomarkers to refine risk stratification or expand eligibility for targeted or immunotherapeutic approaches is of paramount importance. In this context, non-EDM POLE mutations may represent an underrecognized source of genomic instability and a potential bridge between traditionally “cold” MSS tumors and immunologically active phenotypes.

In this opinion review, we examine the emerging biological and clinical significance of non-EDM POLE mutations in CRC, with a particular focus on their interaction with DNA repair pathways, co-mutational landscapes, and potential implications for precision oncology. By integrating current evidence with critical appraisal, we highlight both the promise and the limitations of this evolving field, and propose a framework for future investigation and clinical translation.

MOLECULAR BIOLOGY OF POLE MUTATIONS

DNA POLE, encoded by the POLE gene, is a key replicative enzyme responsible for high-fidelity DNA synthesis, primarily on the leading strand. It functions as a heterotetrameric complex whose catalytic subunit contains both a polymerase domain and a 3′→5′ EDM, the latter ensuring proofreading and correction of misincorporated nucleotides[32,33]. This proofreading activity substantially enhances replication fidelity and operates in coordination with the MMR system to maintain genomic stability[34,35].

Disruption of this tightly regulated process increases mutation rates and contributes to tumorigenesis. POLE-associated mutagenesis is typically characterized by specific base-substitution patterns that reflect the biochemical properties of polymerase-driven errors[36-38].

Mutations affecting the EDM of POLE impair proofreading, leading to the accumulation of replication errors and an ultra-mutated phenotype, often exceeding 100 mutations per megabase[39]. These tumors exhibit a characteristic mutational signature and are most observed in colorectal and endometrial cancers[40].

Clinically, POLE EDM tumors are associated with high TMB, increased neoantigen load, strong immune infiltration, and favorable responses to immune checkpoint inhibitors. Importantly, these mutations appear to occur early in tumorigenesis, shaping the clonal mutational landscape and contributing to enhanced immunogenicity. Germline EDMs further define polymerase proofreading-associated polyposis, a hereditary cancer predisposition syndrome[30,41].

By contrast, non-EDM POLE mutations are less well characterized but increasingly recognized as potentially relevant, particularly in MSS CRC. These variants do not directly disrupt proofreading activity and likely contribute to genomic instability through indirect mechanisms, including altered polymerase dynamics, replication stress, and impaired coordination with DNA repair pathways[42-44].

Non-EDMs are frequently associated with intermediate or elevated mutational burdens and distinct co-mutational profiles, suggesting a context-dependent role. Rather than acting as primary drivers, they may function as modifiers of genomic instability, particularly when combined with defects in auxiliary repair pathways such as MMR-related genes[43,44]. The interaction between POLE mutations and DNA repair systems, especially MMR, is central to their biological impact. Combined defects in proofreading and MMR result in a synergistic increase in mutation rates and accelerated tumorigenesis[42,44].

In clinical and genomic studies, POLE-mutant tumors exhibit heterogeneous patterns depending on whether MMR deficiency precedes, follows, or co-occurs with POLE alterations. This dynamic interplay helps explain the variability in mutational burden and highlights the importance of considering POLE status within the broader DNA repair context[21,31,43-44].

POLE EDM mutations are increasingly recognized as early, potentially initiating events in colorectal carcinogenesis, detectable even in precursor lesions. This early onset of genomic instability contributes to the accumulation of clonal neoantigens and may underlie the favorable immunological profile of these tumors[42,43].

Whether non-EDM follow a similar temporal pattern remains uncertain, their prevalence in MSS tumors suggests a potential role in early tumor evolution that warrants further investigation. Differences between the EDM and non-EDM POLE mutations are presented in Table 1.

Table 1 Classification of DNA polymerase epsilon mutations in colorectal cancer.

Feature	Exonuclease domain mutations	Non-exonuclease domain mutations
Location	Exonuclease (proofreading) domain	Outside exonuclease domain
Mechanism	Loss of proofreading activity	Indirect effects (replication stress, altered polymerase dynamics)
Tumor mutational burden	Ultra-high (> 100 mut/Mb)	Intermediate to high (context-dependent)
Mutational signature	COSMIC signature 10	Heterogeneous/less defined
MSI status	Often MSI-H or independent	Predominantly MSS/MSI-L
Immunogenicity	High (strong neoantigen load)	Variable, potentially increased in subsets
Clinical relevance	Established biomarker for immunotherapy	Emerging, not yet standardized
Role in tumorigenesis	Early driver event	Likely modifier, context-dependent

MSI-H: High levels of microsatellite instability; MSI-L: Low levels of microsatellite instability; MSS: Microsatellite stable.

GENOMIC LANDSCAPE AND CO-MUTATIONAL PATTERNS IN POLE-MUTANT CRC

The genomic landscape of POLE-mutant CRC is characterized not only by elevated mutational burden but also by distinct and recurrent co-mutation patterns, which may critically influence tumor behavior and therapeutic vulnerability (Table 2). Non-EDM POLE mutations are increasingly recognized in CRC, particularly within MSS or MSI-L tumors, where their biological significance appears to be largely determined by the co-mutational genomic context rather than by isolated effects on polymerase proofreading[21,42,45].

Table 2 Co-mutational landscape of non-exonuclease domain mutations DNA polymerase epsilon-mutant colorectal cancer.

Gene	Pathway/function	Role in CRC	Interaction with POLE non-EDM
MLH3	Mismatch repair	DNA repair	Partial repair deficiency → synergistic mutagenesis
MSH3	Mismatch repair	Indel repair	MSI-L phenotype, replication error accumulation
KRAS	MAPK signaling	Oncogenic driver	Sustains proliferation in unstable genome
BRAF	MAPK signaling	Prognostic/driver	Less frequent, may define subgroups
PIK3CA	PI3K/AKT pathway	Growth and survival	Supports tumor progression alongside instability

CRC: Colorectal cancer; EDM: Exonuclease domain mutation; MAPK: Mitogen-activated protein kinase; MLH3: MutL homolog 3; MSH3: MutS homolog 3; MSI-L: Low levels of microsatellite instability; PIK3CA: Phosphoinositide 3-kinase catalytic subunit alpha; POLE: DNA polymerase epsilon.

A consistent observation across studies is the frequent co-occurrence of non-EDM POLE variants with alterations in DNA repair genes, most notably MutL homolog 3 (MLH3) and MutS homolog 3 (MSH3), which are involved in the processing of insertion–deletion loops and replication-associated errors[28,46]. Although these genes are not typically associated with canonical MMR deficiency, their disruption may lead to partial impairment of repair fidelity and, in combination with altered polymerase function, may produce a synergistic defect in DNA repair capacity. This interaction provides a plausible explanation for the presence of elevated TMB in a subset of MSS tumors that would otherwise be considered genomically stable.

In addition, non-EDM POLE-mutant tumors frequently harbor co-occurring alterations in key oncogenic pathways, including KRAS, PIK3CA, and less commonly, BRAF, suggesting a distinct molecular context characterized by both genomic instability and sustained proliferative signaling (Figure 1)[28,47]. These findings support the concept that non-EDM POLE mutations act as modifiers of tumor evolution within a broader network of genomic alterations, rather than as independent drivers.

Figure 1 Co-mutational patterns on non-exonuclease domain mutations colorectal cancer. EDM: Exonuclease domain mutation; CRC: Colorectal cancer.

From a clinical and interpretative standpoint, this has important implications. The significance of POLE variants cannot be reliably assessed in isolation. We therefore strongly advocate against reporting POLE mutation status as a standalone finding, as omission of co-mutational information-particularly involving DNA repair genes, such as MLH3 and MSH3, may lead to underestimation or misinterpretation of their biological and clinical relevance. Instead, POLE status should be integrated into a composite molecular framework that includes MMR status, TMB, and co-occurring genomic alterations to more accurately reflect tumor biology and guide clinical decision-making (Figure 2).

Figure 2 Integrated biomarker framework for colorectal cancer treatment decisions. CRC: Colorectal cancer; dMMR: Deficient mismatch repair; EDM: Exonuclease domain mutation; pMMR: Proficient mismatch repair; MMR: Mismatch repair; MSI: Microsatellite instability; MSI-H: High levels of microsatellite instability; MSS: Microsatellite stable; TMB: Tumor mutational burden.

CLINICAL IMPLICATIONS

The clinical implications of non-EDM POLE mutations in CRC remain incompletely defined, yet emerging evidence suggests they may represent a clinically relevant, currently underrecognized molecular subgroup. Traditionally, therapeutic stratification in CRC has relied heavily on MMR status and MSI, with immunotherapy largely restricted to MSI-H/dMMR tumors. However, the identification of non-EDM POLE mutations in MSS or MSI-L tumors challenges this paradigm, particularly when these variants are associated with elevated TMB and co-existing defects in DNA repair pathways. In this context, non-EDM POLE-mutant tumors may represent a subset of patients who fall outside current immunotherapy frameworks but still harbor biologically meaningful immunogenic features.

A key clinical implication lies in the interpretation of biomarkers. MSI status alone appears insufficient to capture the full spectrum of genomic instability in CRC. The observation that MSS/MSI-L tumors with non-EDM POLE mutations may exhibit increased mutational burden suggests that POLE status-particularly when integrated with co-mutational profiles and DNA repair alterations-could serve as an auxiliary biomarker for identifying patients who might benefit from immune checkpoint inhibition[11-13]. While current evidence remains largely retrospective and hypothesis-generating, it is increasingly difficult to justify a strictly binary MSI-based approach given the molecular complexity. We therefore believe that selected patients with MSS with non-EDM POLE mutations and evidence of increased mutational or neoantigen burden should be considered for clinical trial enrollment and, where appropriate, exploratory immunotherapy strategies.

Beyond immunotherapy, the co-mutational landscape of non-EDM POLE tumors may open avenues for combination treatment approaches. The frequent coexistence of alterations in pathways such as MAPK (KRAS, BRAF) and phosphoinositide 3-kinase PI3K (phosphoinositide 3-kinase catalytic subunit alpha [PIK3CA]) suggests that these tumors may benefit from rationally designed strategies that combine targeted therapies with immune modulation[47]. Although such approaches remain investigational, they highlight the need to move beyond single-biomarker decision models toward integrated, pathway-informed therapeutic strategies.

From a diagnostic perspective, these findings have immediate implications for molecular testing. Restricting POLE analysis to EDM hotspots risks overlooking a substantial proportion of potentially relevant alterations, particularly in MSS tumors. While full-length POLE sequencing and tumor neoantigen burden assessment may not yet be feasible in all clinical settings, a tiered testing strategy appears both pragmatic and necessary[48-50]. In our view, expanded POLE analysis should be prioritized in selected scenarios, including early-onset CRC, MSS/MSI-L tumors with unexpectedly high TMB, or cases harboring concurrent DNA repair gene alterations. At a minimum, non-EDM POLE variants should be systematically reported, even when classified as variants of uncertain significance, as their contextual interpretation may evolve with accumulating evidence.

At the same time, caution is warranted. The current body of evidence is limited by small cohort sizes, retrospective analyses, and population-specific data, and the functional consequences of many non-EDM POLE variants remain incompletely understood[42,51-56]. We strongly emphasize that the presence of a POLE mutation, particularly outside the EDM, should not, in isolation, dictate treatment decisions or be interpreted as a surrogate marker of high immunogenicity. Clinical implementation must be guided by an integrative assessment that includes tumor stage, mutational burden, immune microenvironment, and treatment response data (Table 3).

Table 3 Clinical implications and proposed integration of non-exonuclease domain DNA polymerase epsilon mutations in colorectal cancer.

Clinical domain	Current evidence	Key supporting findings	Proposed interpretation/recommendation
Immunotherapy eligibility	MSS CRC generally resistant to ICIs	MSS/pMMR tumors show poor response rates to ICIs (systematic reviews)	Do not exclude all MSS tumors; consider molecular subgroups
	POLE mutations linked to immunogenicity	POLE EDM tumors show high TMB and strong immune infiltration[25,30]	POLE status should be considered alongside MSI
	Non-EDM POLE may confer benefit	Non-exonuclease POLE mutations associated with ICI response in selected cases[26,31]	Selected MSS + non-EDM + high TMB → potential ICI candidates (clinical trials)
TMB	High TMB predicts ICI response	POLE-mutant tumors show elevated TMB even in MSS context[28,42]	TMB should be interpreted with POLE status, not MSI alone
	MSS tumors usually low TMB	Subset of MSS + POLE non-EDM show intermediate/high TMB	Identify “hidden hypermutated” MSS subgroup
Biomarker strategy	MSI/MMR used as binary biomarker	MSI classification fails to capture full genomic instability spectrum	Move toward composite biomarker model
	POLE not routinely integrated	Non-EDM variants often underreported or labeled VUS	Include POLE (full-length or extended panels) in selected cases
Molecular testing	Hotspot POLE testing (EDM-focused)	Non-EDMs more frequent than EDMs[42]	Expand testing beyond exonuclease domain in high-risk scenarios
	TMB/TNB testing limited	Requires WES or large panels[55]	Use tiered strategy: Prioritize selected patients
Co-mutation-informed therapy	Targeted therapy used independently	KRAS, PIK3CA, BRAF commonly co-occur with POLE[42,46]	Consider combination approaches (ICI + targeted therapy)
	Limited integration of pathways	Co-mutations reflect pathway activation (MAPK, PI3K)	Adopt pathway-informed therapeutic strategies
Prognostic implications	MSI-H → favorable prognosis	POLE EDM tumors associated with improved outcomes	Prognostic role of non-EDM unclear
	MSS → poorer prognosis	Some POLE non-EDM tumors show aggressive features[28]	Avoid overinterpreting mutation alone without context
Reporting practices	POLE often reported alone or omitted	Co-mutations (MLH3, MSH3) influence biological behavior	Always report POLE with co-mutational profile
	Variants labeled as VUS	Functional impact unclear but context-dependent	Do not ignore non-EDMs-interpret within genomic context
Clinical trial design	MSI-based stratification	POLE-mutant tumors span MSI categories	Future trials should stratify by POLE + co-mutations
Clinical caution	Biomarker-driven decisions increasing	Many non-EDM variants lack functional validation	Do not use POLE alone for treatment decisions
	Evidence still emerging	Small cohorts, retrospective data	Integrate molecular + clinical parameters (stage, TME, response)

CRC: Colorectal cancer; EDM: Exonuclease domain mutation; ICI: Immune checkpoint inhibitor; MAPK: Mitogen-activated protein kinase; MMR: Mismatch repair; MSI: Microsatellite instability; MSI-H: High levels of microsatellite instability; MSS: Microsatellite stable; PIK3CA: Phosphoinositide 3-kinase catalytic subunit alpha; pMMR: Proficient mismatch repair; POLE: DNA polymerase epsilon; TMB: Tumor mutational burden; TME: Tumor microenvironment; TNB: Tumor neoantigen burden; VUS: Variant of uncertain significance; WES: Whole exome sequencing.

In this regard, future research should prioritize prospective clinical trials stratifying patients based on POLE status and co-mutational profiles[57,58], rather than MSI alone, as well as functional validation studies to clarify the biological impact of non-EDM variants. Until such data are available, non-EDM POLE mutations should be viewed not as definitive biomarkers, but as promising signals within a broader molecular context, requiring careful interpretation[59,60]. Balancing innovation with caution will be essential to avoid premature clinical translation while still recognizing the potential of this emerging subgroup to expand the boundaries of precision oncology in CRC.

CONCLUSION

Recognition of non-EDM colorectal tumors may be a significant prognostic tool and aid in personalized patient treatment. This exciting new frontier in colorectal surgical oncology should not lead us astray. We should be very cautious about overinterpreting high mutation prevalence in the absence of biological validation. In other words, we should not rush to change treatment decisions or label patients as high-risk based on the frequency of a mutation alone, but only after further research has been conducted. Prognostic assessment in non-EDM POLE–mutant CRC should integrate molecular findings with clinical parameters, including disease stage, tumor location, immune microenvironment features, treatment response, and longitudinal outcomes, rather than relying on mutation frequency alone.