Screening programs improved the early detection of colorectal cancers (CRCs) (1–5) but raised a new clinical issue: the management of endoscopically resected pT1 colorectal tumors (pT1 CRC).

According to the Tumor Node Metastasis (TNM) classification of malignant tumors, pT1 CRC is defined by the presence of submucosal invasion and represents the earliest stage with metastatic potential (6). Specifically, nodal metastasis in this setting has been reported in up to 15% of pT1 CRC cases (7, 8).

Following pT1 CRC endoscopic resection, two different approaches can be undertaken: either conservative endoscopic follow-up (9, 10) or surgical resection enabling nodal status assessment (11–13). This choice is guided by the risk of node metastasis which is estimated according to specific histopathological parameters which stratify these tumors in low- and high-risk pT1 CRC (14, 15). The ultimate clinical aim is to balance staging/therapeutic benefits with surgical risks and consequences (16, 17).

Regarding the analyzed histopathological features, tumor budding, lymphovascular invasion, tumor grading, and micro staging repeatedly resulted to be significantly associated with nodal involvement: currently, in the presence of even one of these morphological features, surgical resection is proposed, thus favoring lymph node metastases identification but also increasing the percentages of cases submitted to surgery with no lymph node involvement (7, 18).

Based on this observation, a significant subgroup of patients is overtreated (19, 20). In these cases, colectomy with lymphadenectomy is stadiative rather than curative, and it exposes patients to surgery-related morbidity and mortality risks that could be avoided through improved patient selection. Improvement of the risk estimation of nodal metastasis in this setting is an unmet clinical need and would help avoid these risks. In particular, it has to be considered that most of the patients without evidence of node metastasis after surgery were surgically treated because of the presence of a single high-risk parameter, thus novel approaches should focus on developing combined multiparametric scores (7, 8, 18, 21).

Fostered by this state of the art, this study aims to improve pT1 CRC risk stratification by proposing a novel and effective clinicopathological predictor score of nodal metastases, established through the analysis of a comprehensive range of clinical and histopathological features in a large consecutive series of pT1 CRC, surgically treated at a tertiary referral institution.

This study aimed to enhance the clinical management of endoscopically resected pT1 CRC by identifying and combining metastasis-correlated clinicopathological features in a predictive score of nodal metastasis.

Our series of retrospective consecutive pT1 CRC presented a percentage of nodal metastasis (8,7%) in agreement with the literature range of 4.9–16.9% (7, 8, 19, 21). This data shows that, to date, a large number of cases are potentially overtreated and exposed to significant surgical-related risks without certain benefits except for achieving definitive pathological tumor staging.

Until now, the prediction of CRC pT1 nodal status has been based on histopathological features alone assessed on endoscopically-resected specimens. According to the current international guidelines, the presence of even one single histopathological parameter among poorly differentiated tumors (G3), high-grade tumor budding, lymphovascular invasion, and submucosal invasion ≥1 mm depth or ≥ 4 mm width is routinely used to assign a high risk of node metastasis, aiming for identifying any potential metastatic case but leading to the surgical overtreatment of patients with no metastatic disease (11, 13, 28–33). Of note, these features proved to correlate with node metastasis although a somewhat disappointing interobserver agreement (34, 35). In our study, we confirmed lymphovascular invasion and high-grade tumor budding, and additionally identified absent/mild TILs and younger age (cut-off: 65 years) as relevant features related to nodal involvement and, therefore, we combined them to build our risk score.

Among these, tumor budding and lymphovascular invasion have been thoroughly studied and their association with nodal metastasis is well recognized (8, 18, 33, 36). Recent evidence also suggested that these features are intertwined and could represent two sides of the same coin, ultimately leading to lymph node metastasis (21) since cells composing the tumor “buds” were found to be also responsible for the invasion of nearby stromal lymphovascular vessels. These cells exhibited epithelial and mesenchymal markers suggesting the activation of the epithelial-mesenchymal transition process, an aggressive tumor phenotype associated with poor survival and resistance to therapy (37–41). Our data support their strong reciprocal intertwining and their influence on lymph nodes metastasis development.

Our results also demonstrate that TILs are significantly correlated with the risk of harboring a nodal metastasis, a finding in agreement with the established role of the immune system in influencing CRC growth and survival (42–45). However, it should be noted that so far, studies evaluating the role of TILs in CRC mainly focused on advanced disease settings. The few reports addressing pT1 CRC provided conflicting results. (43, 44, 46–48). A univariate analysis of 102 endoscopically resected pT1 CRCs reported an increased number of CD3 positive TILs in the metastatic group, but this finding was not confirmed by multivariate analysis (21). No correlations were observed analyzing the CD8 positive TILs (21). Differently, a recent study by Kang et al. analyzed the distribution and mean numbers of CD3, CD8, and FOXP3 positive lymphocytes in pT1 CRC and identified a significant association between lymph node metastasis and a lower number of CD8 positive lymphocytes located within the tumor core (49). In the present study, we demonstrated a correlation between absent/mild TILs and a higher risk of nodal involvement. This association seems to be particularly relevant since none of our cases with high TILs presented nodal metastasis, while the four score 1 cases presenting nodal metastasis received this score exactly due to the presence of absent/mild TILs. We believe that TILs could become a significant histopathological feature of pT1 CRC, representing a morphological counterpart of tumor immunobiology and immune system activation against tumor growth. In this regard, our results support the relevance of this parameter in shaping tumor invasiveness even at early disease stages, but further studies should specifically investigate TILs in early CRC to confirm their predictive significance in terms of nodal metastasis and provide additional insights on their mechanistic role and develop specific guidelines regarding their assessment.

Lastly, the predictive significance of age at diagnosis is of particular interest since it shows the importance of clinical variables even in this early oncological setting. Moreover, whereas the assessment of histopathological features alone can be hampered by inter-observer variability (50), age at diagnosis is not. Younger age is also known to be related to more aggressive CRC lesions (51–54).

We did not confirm micro staging (≥1 mm depth and ≥4 mm width of invasion) nor tumor grading as potential risk factors of nodal involvement. Similar evidence has been recently reported in the literature: in particular, micro staging reliability in pT1 CRC has been questioned for the significant influence of technical processing and lesion morphology on its interpretation; similarly, the importance of tumor grading as a predictive marker has been found to be overall limited compared with other morphological features (55–59).

In the practical daily routine, some pT1 CRC cases, although close to the recommended cut-off, did not meet the required number of harvested lymph nodes. Indeed, this is a limitation that actually occurs, potentially impacting the pT1 CRC staging (60, 61). Aiming to develop a score that can be applied in a “real-world” pT1 CRC series, we preferred to also maintain these cases and test our score in both scenarios, thus considering either all cases (including the ones with <12 lymph nodes) or the subgroup of N+ and N0 >12 only cases. Indeed, our score maintains the same level of statistic and clinical significance, further confirming its practical relevance.

We acknowledge that our study presents some potential limitations, including its retrospective nature and the limited, although considerable, sample size. Moreover, future studies should investigate the efficacy of the here proposed approach on endoscopic resection samples.

Ultimately, our score improves the overall identification of patients who would benefit from a surgical approach following endoscopically resected pT1 CRC compared to the evidence reported so far in the literature (62–64). It should also be noted that balancing the risks due to over-and undertreatment is extremely challenging when dealing with an early tumor stage. Identifying subgroups of patients, no matter how small, with very low or very high metastatic risks is particularly important since it allows to confidently propose a therapeutic option to these patients. Our score enabled this result since no score 0 patient showed nodal metastasis, while all score 4 had metastatic disease. Although the sample sizes of these groups were particularly limited, these promising results warrant further validation. Among patients with intermediate-risk scores (score 1–3), we observed a progressive increase of nodal metastasis (score 1: 3.4%, score 2: 16.7%, score 3: 60.0%), a piece of information that would not have been available by employing the conventional dichotomized approach. This stratified risk assessment could help tailor patients' management by multidisciplinary boards.

In conclusion, despite some limitations, including the retrospective and single-institutional nature, our study allowed us to develop a novel multiparametric score combining evidence-based histopathological features with age at diagnosis as a clinical variable. Our score represents an effective system to estimate the lymph node metastasis risk, superior to the current single parameter-based risk assessment. Further multi-centric and ideally prospective studies are recommended to confirm our findings and enable the adoption of the proposed score within multidisciplinary CRC tumor boards to guide patients' management.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Institutional Review Board (or Ethics Committee) of the University of Turin (approval number: DSM-ChBU n° 03/2020, approved in October 2020). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

LB and PC: conceptualization and funding acquisition. AG, EF, GB, SO-A, RS, IC, LB, and PC: methodology and writing—review and editing. AG, EF, GB, SO-A, and LB: formal analysis. AG and EF: writing—original draft preparation. All authors contributed to the article and approved the submitted version.

This research was funded by the Rete Oncologica del Piemonte e della Valle d'Aosta (Oncology Network of Piedmont and Aosta Valley-Italy; Grant to LB and PC, no specific grant number available).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.