Colorectal surgery has had a notable shift from traditional open surgery to minimally invasive techniques over recent years. Initially, laparoscopic surgery gained prominence as a less invasive alternative, demonstrating clear benefits over open surgery, including faster return of bowel function, less postoperative pain and opioid use, and shorter length of stay. However, a trend toward adopting robotic-assisted surgery emerged, providing high-resolution three-dimensional imaging and superior ergonomics, with potential to overcome technical limitations of laparoscopy.

In this study, there was a pronounced preference for minimally invasive surgical approaches (91%), reflecting the current standard in colorectal cancer management. Notably, laparoscopic procedures were more frequently utilized for colonic interventions, whereas robotic-assisted surgeries predominated in rectal cases. This distribution aligns with the prevailing literature, where the most comprehensive studies on robotics in colorectal cancer have specifically emphasized its application in rectal cancer treatment [9, 10].

At the beginning of the study period, there was an increase in the use of laparoscopy, which was progressively substituted by an increase in robotic surgery. At the same time, as expected, a decrease in the proportion of open surgery was observed in the first years of the study period. The biggest concern when transitioning from open approach to MIS is the increase of intra- and postoperative complications during the learning period. Our study found that during the implementation of both laparoscopic and robotic systems, the increase in CD ≥ 3 complications did not change. As a matter of fact, at the end of the study period (2023) twice the number of operations were performed by the surgical team with the same absolute value of complications.

We believe that the main reason for these results is the application of modular, systematized, and permanently audited learning. This method is performed by the most experienced senior surgeon and has been advocated as the method of choice when learning advanced surgical techniques, having no deleterious effects on clinical outcomes. This goal-oriented, targeted teaching method enables a single experienced surgeon to prepare several senior colorectal surgeons for independent MIS practice at the same time. A major benefit of modular training is its applicability across a wide spectrum of patients. With this module, it is possible to provide at least one module of training during 98% of all colorectal resections even though most patients have adverse factors regarding conversion/technical difficulty [11].

Total mesorectal excision completeness has become a marker for good surgical technique in rectal cancer and predicts the likelihood of local recurrence of the cancer in the pelvis [12]. In the population presented, surgery was performed according to the mesorrectal plane in 87% of patients, which is comparable to the results observed in ACOSOG and COLOR II Trial (85%) and higher than those in the ROLLAR (75%) and COREAN Trials (73.6%) [10, 13,14,15]. Although this rate is lower than in the REAL Trial (93%) [9], there are patients submitted to laparoscopic and open treatment in our population and a higher rate of lower tumors, which may explain this difference.

Concurrent with the high quality of mesorectal excision, a high rate of R0 resections was observed in our study (95%), suggesting that patients in our center receive effective treatment regarding complete tumor removal with clear margins and minimal risk of recurrence, aligning closely with the standards established by several widely recognized RCTs in the colorectal field.

These results collectively provide evidence showing the efficacy and success of the implementation of MIS in colorectal cancer in this center regarding oncological outcomes.

The conversion rate in our study was significantly lower compared to the literature (11–25%) [14, 16,17,18]. According to the CLASICC Trial, the most common causes for conversion in colon cancer were excessive tumor fixity, uncertainty of tumor clearance, and obesity, while in the COST Trial they were advanced disease, no visualization of critical structures, and adhesions [17]. As a tertiary center, we managed to treat patients with more advanced disease, maintaining a low conversion rate. Notably, in our cohort of rectal cancer patients, conversion occurred only in mid-rectal cancers. In the COLOR II trial, mid-rectal cancers also had the highest risk of conversion [14]. Although the explanation is still not clear, it is known that in laparoscopy there is a limited of angle of operation that can make the resection of middle and low rectal cancer difficult.

Concerning the robotic approach, in rectal cancer there were no conversions to open surgery, which aligns with the advantages of robotic surgery in rectal cancer described in the literature [7,8,9,10]. Globally, the conversion rate was higher in laparoscopic compared to robotic surgery in both colon and rectal cancers, similar to results provided in the ROLARR and REAL Trials [9, 10].

The overall low conversion rate in this study highlights not only the effective approach taken towards MIS in this center but also the high level of safety of these procedures. Conversion to open surgery is often used to assess the MIS learning curve, as conversion rate is mainly associated with the experience of the surgeon, hospital volume, and case complexity.

Our results demonstrate a 32% total number of complications. This overall rate is lower than the literature rate of 40–58% [13, 14, 18]. Similarly, in our study, severe complications (Clavien-Dindo ≥ 3) occurred in only 11% of all colorectal cancer resections, which is lower than the rate in the available literature [13, 19] when compared to well established units.

The anastomotic leak rate was 4.2%, which is lower compared to 8.5% inthe CLASICC Trial, 12% in the COLOR II Trial, and 11% in the ROLAR Trial [10, 14, 18]. Several factors can contribute to the risk of anastomotic leak [20], namely patient’s comorbidities, preoperative chemotherapy, type of anastomosis, tumor stage and location, use of protective stoma, surgeon’s experience, and postoperative care, although some of these are still controversial in the literature.

Regarding the length of stay (LOS) after surgery, the previously cited studies indicate a mean of 5 to 13 days, whereas the median hospital stay is 6 days. However, the hospital stay may be influenced not only by time to recovery and quality of postoperative care but also by social factors such as differences in medical fees, medical insurance, and medical systems among countries [21].

The low rate of postoperative complications, mainly severe complications (CD ≥ 3) and anastomotic leak, is a positive indicator of quality and safety, demonstrating that there can be a shift from open to MIS with minimal risks to the patient. Furthermore, the short LOS suggests faster recovery and reduced risk of hospital-acquired complications, supporting that the implementation of MIS in colorectal cancer is not only safe but also efficient and has socioeconomical advantages over open surgery.

The future directions of MIS will bring the integration of artificial intelligence (AI), which is expected to provide real-time decision support and predictive analyses during surgery. In addition, robotic platforms are expanding to allow for more complex procedures, which could lead to greater use in advanced and recurrent colorectal cancer. These advances should be designed to provide better access training and standardization as well as cost-reduction to facilitate generizability.

One of the primary strengths of this study is its external validity owing to the utilization of population-based data that mirror daily clinical practice. This ensures that the findings are widely applicable and reflective of real-world scenarios. Additionally, the integrity of the dataset is supported by the minimal number of missing data, which contributes to the robustness and reliability of the results presented. Due to the minimal number of missing data concerning our main outcomes, a list-wise deletion approach was preferred.

Limitations of this study included that this was a single-institution retrospective analysis, with inherent selection bias. There is a limited generalizability to other healthcare settings, as patient populations and surgical practices can vary between institutions. Furthermore, this study primarily focuses on short-term outcomes and perioperative measures. Long-term oncological outcomes and patient survival data may not be available or sufficiently analyzed in this study, which limits the ability to draw conclusions about the long-term effectiveness of MIS.

Despite these potential drawbacks, it is reassuring that our results are similar to those previously published in the more representative RCTs in this field.