Total neoadjuvant therapy (TNT) is increasingly recognized as the preferred approach for high-risk locally advanced rectal cancer, with studies showing higher response rates and improved treatment compliance compared with conventional chemoradiotherapy [3, 4]. Recent evidence also suggests improved oncological outcomes and greater potential for organ preservation with TNT [5, 9]. Reflecting this progress, the NCCN incorporated TNT into its 2022 guidelines for both CRM-threatened and CRM-negative tumors [10]. Although TNT offers advantages in terms of treatment response and watch-and-wait feasibility, tumor regrowth remains a significant concern and necessitates careful surveillance.

In this study, we analyzed 32% of patients with LARC in our cohort who underwent TNT, achieved cCR, and were selected for NOM. In this cohort, the regrowth rate was 23%, reflecting the significant challenge of pursuing NOM after initial cCR.

The findings of this study align with data from the International Watch & Wait Database, which reported a 2-year incidence of local regrowth of 25.2% in patients achieving cCR after TNT without surgery [11]. As younger patients in our cohort had a marginally higher incidence of regrowth, early onset LARC could be considered a potential predictor of regrowth, although this association did not reach statistical significance. Nonetheless, tumor regrowth remains a critical challenge, particularly in younger patients, as it may affect their long-term health outcomes and quality of life. Patients experiencing regrowth are known to be at an elevated risk of distant metastasis [12], and local management of regrowth is complex [13]. While all patients in our study had clear proximal and distal margins, one patient exhibited a positive circumferential margin and another had near-complete mesorectal excision, reflecting the complexity of the resection. Furthermore, one patient developed distant lung metastasis, indicating the potential for disease progression despite an initial cCR. These findings emphasize the importance of identifying regrowth predictors to refine patient selection for NOM and facilitate the early detection of local recurrence.

Based on existing studies, some potential predictors of tumor regrowth have been identified that were not replicated in our current study. For example, one previously identified risk factor for local regrowth is the total radiation dose, where patients who received less than 50.4 Gy have been reported to have a higher likelihood of developing local regrowth [12]. However, in our cohort, most patients underwent long-course radiation therapy, which may explain the lack of significant differences in regrowth observed between patients who received different radiation doses in our analysis. The InterCoRe consortium previously reported the risk of local regrowth to be approximately 20% for cT2, 30% for cT3, and nearly 40% for cT4 tumors, respectively [14]. In our cohort, most patients had cT3 disease; however, no statistical differences in local regrowth rates were observed between the T stages. In contrast, Habr-Gama et al. found that radiological nodal status was not associated with the risk of local regrowth or surgery-free survival in these patients [15]. Our findings align with this observation, as regrowth rates were comparable across the different N stages in patients with LARC. It is important to note that most of these potential predictors have been identified in patients who pursued NOM following chemoradiation alone rather than TNT. This distinction highlights the need for further research to evaluate predictors specifically in the context of TNT protocols.

Although no definitive predictors of regrowth have been identified following TNT, certain patterns have emerged. In the OPRA trial, tumor regrowth was observed in 44% and 29% of patients in the induction and consolidation chemotherapy arms, respectively [11]. Different chemotherapy regimens were used within these arms, leading to variations in the timing of chemotherapy completion, which in turn affected the intervals between chemotherapy completion and final treatment response assessment. Notably, an extended interval between the initiation of neoadjuvant therapy and surgery may have contributed to the increased risk of local failure. These findings suggest that treatment intervals and sequencing of chemotherapy (induction versus consolidation) may play a role in influence the likelihood of regrowth, highlighting the need for further studies to optimize treatment strategies. Similarly, findings from the RAPIDO trial, which compared TNT with consolidation chemotherapy versus standard preoperative nCRT, further highlighted the interplay between treatment timing and outcomes [3]. In RAPIDO, patients undergoing TNT had a longer interval between radiotherapy and surgery (22–24 weeks) than those in the standard nCRT arm (6–10 weeks). Although the pathological complete response (pCR) rate was higher in the TNT arm (28% vs. 14%; p < 0.0001), local failure was also more frequent in this group (12% vs. 8%; p = 0.07). This paradoxical result may reflect the repopulation of radio-chemoresistant clonogens during the extended waiting period, further emphasizing the need to refine predictive markers and optimize intervention strategies to minimize regrowth risks.

Identifying effective predictors of regrowth is complicated by the potential tumor biological changes induced by TNT or subsequent NOM. Residual tumor cells surviving chemoradiation may adapt by repairing damage from consolidation chemotherapy, selecting for more aggressive chemoresistant clones characterized by epithelial-to-mesenchymal transition [16]. Additionally, chemoradiation may promote local immune evasion in residual cells with specific genomic alterations, increasing the likelihood of both local regrowth and distant metastasis [17]. Any therapeutic benefit achieved through induction or consolidation chemotherapy may, in patients with cCR, inadvertently drive the selection and dissemination of small clusters of viable chemo-resistant tumor cells that may remain scattered within the area formerly occupied by the tumor [18]. This underscores the critical need to define robust predictors of regrowth to optimize patient selection and screening during NOM, which could potentially be enabled through recent advances in artificial intelligence and machine learning, which leverage routinely acquired imaging data [17].

This study has several limitations inherent to its design. First, the retrospective and single-center nature of the study may restrict the generalizability of our findings to broader patient populations. The most important limitation is the small sample size, which reduces the statistical power of the analyses and increases the risk of a type II error, potentially leading to false acceptance of the null hypothesis. Due to the limited sample size (n = 28), multivariable modeling could not be performed, which restricted our ability to control for potential confounders. This methodological limitation should be considered when interpreting these results. Consequently, the findings of this study should be regarded as exploratory and hypothesis-generating, rather than confirmatory. To more reliably identify predictors of tumor regrowth in patients undergoing NOM after TNT, larger multi-institutional and collaborative studies are essential to validate and strengthen these results. Although we evaluated the histologic grade, MMR status, and gene mutations, the inclusion of additional variables that better reflect tumor biology could enhance predictive analyses. Furthermore, variations in treatment sequencing and time intervalsstemming from the individualized TNT protocols at our centermay have influenced our results. Our relatively short follow-up period also posed a limitation, as 25% of the patients included in the NOM cohort were enrolled within the past year, potentially restricting the ability to assess long-term outcomes. Despite these limitations, our study is among the few that specifically investigates predictive factors for rectal tumor regrowth after TNT, emphasizing the clinical significance and utility of identifying such predictors to optimize patient selection and NOM.