Robotic transanal surgery: perspectives for application

Transanal minimally invasive surgery (TAMIS) is a surgical technique which allows the local excision of rectal benign tumors and early stage cancers measuring up to 4 cm and lying within 6-8 cm from the anal verge. It is performed by means of a disposable transanal platform and conventional laparoscopic instruments, proving to be effective and easily available. Hence, TAMIS soon became a valid alternative to other transanal resective procedures, especially transanal endoscopic microsurgery, and rapidly spread. Moreover, soon after its introduction, TAMIS started to be performed also using robotic technologies, but no clear advantages were found to date. This review is intended to provide a general overview on TAMIS, with a special focus on its association with robotic systems and the perspectives of this approach.


INTRODUCTION
Benign tumors and early stage cancers of the rectum measuring less than 4 cm and lying within 6-8 cm from the anal verge represented for a long time an indication for conventional transanal excision (TAE) [1] . is technique was performed under direct view using a rigid anoscope and conventional surgical instruments. However, inadequate exposure was quite usual, especially when lesions were located in the middle or upper rectum, which frequently appeared inaccessible [1,2] . at o en compromised the quality of surgical resection [1,2] , with a rate of positive margins higher than 10% even in the series of the most experienced surgeons [3,4] .
The attempt to overcome the limitations of TAE stimulated the development of transanal endoscopic microsurgery (TEM). Introduced in the early 1980s by Buess et al. [5] , this technique involved the use of three main components: a speci c rigid proctoscope, a dedicated camera, and modi ed laparoscopic instruments. In particular, the rigid proctoscope was xed to the operating table, oriented by the surgeon, and provided with several ports for pneumorectum creation, smoke evacuation, the camera, and instruments. Operative steps were quite similar to TAE, and the surgical wound resulting from full-or partial-thickness wall resection could be le open or closed by several techniques (such as sutures or clips).
Compared to TAE, TEM allowed an easier resection of rectal tumors lying in the middle or upper rectum, making excision possible even in some cases of lower sigmoid colon lesion [6] . Moreover, several studies showed better outcomes after TEM [3,[7][8][9][10] . Notably, a systematic review with meta-analysis published by Clancy et al. [8] in 2015 reported significant differences in terms of negative resection margins, specimen fragmentation, and local recurrence in favor of TEM, whereas the postoperative complication rate was similar between conventional TEM and TAE.
However, TEM showed to have some important limitations. Notably, the dedicated surgical equipment was designed for an up-to-down approach to rectal lumen. at made the resection of anterior rectal lesions quite challenging, requiring to place the patient in a prone position and to use speci c split-leg operating tables. Moreover, the necessary surgical material was expensive and the learning curve long. Hence, the implementation of TEM remained limited [1,2] .
To provide an alternative to TEM for local excision of rectal tumors, in 2009, an American surgical team from Florida introduced a new technique called transanal minimally invasive surgery (TAMIS) [11] . e present study is intended to provide a general overview on TAMIS, summarizing its most important aspects and focusing on the association with robotic technology and its implementation.

TAMIS: TECHNICAL ASPECTS AND INDICATIONS
TAMIS is a surgical technique introduced by Atallah, Albert, and Larach in 2009 to provide an alternative to TEM for local excision of rectal neoplasia [11] .
It combines the use of a disposable multichannel port placed transanally with conventional laparoscopic equipment. Notably, the pneumorectum is achieved using common laparoscopic systems in ating CO 2 , and the endoluminal pressure ranges between 15 and 25 mmHg [1] . A 30°-or 45°-angled 5-mm laparoscope is preferable [12,13] , whereas conventional laparoscopic instruments are used for manipulation. Initially, singlesite multichannel ports conceived for laparoscopic abdominal surgery were adapted to a transanal use. Later, several devices were speci cally designed [12][13][14] .
TAMIS is indicated for the local excision of a number of benign and premalignant tumors of the rectum located at up to 15 cm from the anal verge [15][16][17][18][19] . is technique may also represent a curative treatment for selected patients with rectal cancer. Notably, according to the National Comprehensive Cancer Network guidelines 2018 [20] , transanal local excision may be an appropriate therapeutic option in case of early-stage T1 tumors with small size (< 3 cm), well to moderate di erentiation, location within 8 cm from the anal verge, and extension to less than 30% of rectal circumference. The resection must be full-thickness and assure more than 3-mm negative margins. e specimen must be oriented and pinned before xation. In the case of negative pathological features, such as positive resection margins, lymphatic and vascular invasion, poor di erentiation, or Kikuchi sm3 level, a more radical resection is needed. Moreover, TAMIS could also be indicated in the case of more advanced T-stage rectal cancer in order to provide an excisional biopsy for a more precise pathologic examination or to treat patients at high surgical risk in association with other treatments [13,14] . e principles of local resection are similar in TEM and TAMIS. First, the lesion must be marked around its circumference. Benign tumors may be excised limiting the dissection to the submucosal layer, without need to close the surgical wounds [13] . On the contrary, malignant tumors require a full-thickness resection of the rectal wall [13] . Moreover, in the case of posterior tumors, a small amount of perirectal fat may be excised en bloc to ensure a complete excision and allow lymph node analysis [13] .
When rectal tumors are located posteriorly, it would not be necessary to close a full-thickness defect because of its extraperitoneal position, as suggested by several authors reporting no higher complication rate after leaving surgical wounds open [13,21] . However, this time-consuming practice is recommended to cover an eventual peritoneal entry [13,22] . In particular, the wall defect is generally closed transversely, using separated or running sutures, clips, or other devices [1,13] . Peritoneal entry represents a well-known complication of TAMIS and it is described more frequently when tumors are located anteriorly in the middle or upper rectum [12] . When it occurs (1% of cases [12] ), it is recommended to use a steep Trendelenburg position to facilitate wall repair and to convert to laparoscopy if it is impossible to maintain an adequate pneumorectum [22][23][24] . With this regard, some authors suggest placing the patients in a prone position to limit the amount of CO 2 passing into the peritoneal cavity [25] . Among the postoperative complications of TAMIS, a rectovaginal fistula can also occur, besides common general surgery complications, such as bleeding (2.8%) [12] , fever, urinary tract infections, and atrial brillation.

TAMIS: OUTCOMES
In 2014, Martin-Perez et al. [12] published a systematic review of the literature about TAMIS, including 24 retrospective studies and 9 case reports. e authors reported that up to eight alternative abbreviations were used to indicate the same technique, and included overall 390 patients, undergoing TAMIS for malignant lesions in 209 cases (53.5%), adenomas and high-grade dysplasia in 152 cases (39%), and other pathology in 29 cases (7.5%). Among these latter 29 cases, 23 patients (79.3%) were operated for neuroendocrine lesions, 3 patients (10.3%) for fibrosis, 1 patient (3%) for GIST, 1 patient (3%) for mucocele, and 1 patient (3%) for melanoma. Surgical procedures were performed using eight different TAMIS platforms, among which SILS port was the most commonly used (66.7% of all studies included). e mean size of resected lesions was 3.1 cm (range: 0.8-4.7 cm), whereas the mean distance from the anal verge was 7.6 cm (range: 3-15 cm). e authors reported only full-thickness excisions in 22 studies (60.6%), only partial thickness excisions in 3 studies, both full-thickness and partial thickness excisions in 8 studies (24.2%), and no precision about the extent of resection in 5 studies (15.2%). Conversion rate to TAE, TEM, or abdominal laparoscopic surgery was 2.31% (9/390). e mean operative time was 76 min (range 25-162 min), the complication rate was 7.4%, and the mean hospital stay was two days. Considering the publications reporting speci c information about surgical resection quality, the rates of positive resection margins, specimen fragmentation, and recurrence were 4.36% (12/275), 4.1% (4/97), and 2.7% (7/259), respectively.

TAMIS VS. TEM
In the literature, several studies compared the outcomes of TAMIS and TEM [26][27][28] . Among them, it is important to consider the large multi-institutional matched prospective study published by Lee et al. [28] in 2017. It included 181 patients undergoing TAMIS and 247 patients undergoing TEM, and showed that TAMIS was associated with shorter operative time (mean: 70 min vs. 108 min), shorter length of hospital stay (median: zero days vs. one day), and lower blood loss (median: 10 mL vs. 30 mL) compared to TEM. However, no signi cant di erence was found in terms of positive margins (7% for TAMIS vs. 6% for TEM), specimen fragmentation (4% for TAMIS vs. 3% for TEM), postoperative complications (9% for TAMIS vs. 11% for TEM), and recurrence after resection for rectal malignant lesion (7% for TAMIS vs. 7% for TEM). Overall, the authors concluded that, given the absence of signi cant di erences in terms of resection quality and postoperative morbidity, the choice of the technique should be based on surgeon's preference, availability of surgical materials, and costs.
Two further comparative studies were both published in 2016 [26,27] . Notably, compared to TEM, TAMIS had shorter median hospital stay (four days vs. ve days) but lower full-thickness resection rate (85% vs. 100%) according to Mege et al. [26] and higher specimen volume (mean: 5.6 cm 3 vs. 15.9 cm 3 ) according to Melin et al. [27] . No other signi cant di erence between TAMIS and TEM was found in terms of operative, pathological, and survival outcomes.
In 2019, Van den Eynde et al. [29] published another retrospective comparative study, including 68 patients in the group TAMIS and 53 patients in the group TEM. No conversion was reported in both groups. Operative time was again signi cantly shorter for TAMIS (median: 45 min vs. 65 min), whereas lesion surface area was larger for TEM (median: 21 cm 2 vs. 14 cm 2 ). e di erence in operative time persisted a er correction for lesion surface area. No other signi cant di erence was found in terms of quality of resection, morbidity, and hospital stay. Finally, the authors concluded in favor of TAMIS, whose shorter hospital stay was explained by an easier set-up and a greater versatility of the transanal platform. Moreover, the main advantage of TAMIS was identi ed in the fact that all procedures could be performed in lithotomy position, the whole rectal circumference being accessible with this technique.
To be noted, several studies also analyzed the learning curves of TAMIS and TEM, reporting an improvement of operative e ciency a er 14-24 [30] and 18-31 [31] procedures for TAMIS, and a er 16 [32] procedures for TEM. However, no comparison was performed in this regard.
Overall, the available studies do not show a clear superiority of TAMIS over TEM, especially in terms of resection quality, which appear similar. However, several technical advantages making TAMIS preferable are reported in the literature [1,2,13] . In particular, the use of a shorter sha and a more exible platform allows surgeons to reach all quadrants of the rectum, and to perform surgery in lithotomy position also in the case of anterior or lateral lesions. Moreover, TAMIS allows a faster set-up (2 min vs. up to 30-45 min [13] ) and a 360° visibility (220° for TEM), involving the use of less expensive and more easily available equipment, such as conventional laparoscopic instruments.

ROBOTIC TRANSANAL SURGERY
The technical advantages provided by surgical robotic systems, such as stable 3D view and ameliorated manipulation, appeared to overcome some ergonomic limitations of TAMIS. erefore, the same authors who had previously introduced this technique described the rst combined use of robotic technology and TAMIS platforms for local transanal excision in a cadaveric model in 2011 [33] , and then in a real patient in 2012 [34] .
Notably, the rst patient undergoing a procedure of robotic transanal surgery or robotic TAMIS (R-TAMIS) was a 58-year-old woman with a 3-cm tubulovillous rectal adenoma with focal intramucosal carcinoma. e tumor was located at 7 cm from the anal verge in the le anterolateral quadrant. e patient was placed in a lithotomy position and a GelPOINT platform was used, in association with a 5-mm laparoscope introduced through one 5-mm port and two robotic arms introduced through two 8-mm ports. A full thickness resection was performed and the wall defect was closed using a running barbed suture. However, already in this rst case report, the authors underlined the high direct cost of the procedure, $1500, and suggested the use of (R-TAMIS) in complex cases where TEM or conventional TAMIS, o en indicated as laparoscopic TAMIS (L-TAMIS) [2] , were not possible.
Some useful data are also provided by the second largest series found about R-TAMIS, published by Liu et al. [36] in 2018 and including 34 patients. All cases reported were performed using a Da Vinci Xi technology and a GelPOINT platform. Surgery was performed for benign lesions in 22 cases (64.7%) and for malignant lesions in 11 cases (32.3%). In 94% of cases (n = 32), patients were placed in a lithotomy position. e mean distance of rectal lesions from the dentate line was 8.6 cm (range: 2-15 cm), with a mean maximum diameter of 2.6 cm (range: 0.5-4.5 cm). e overall operative time was 100 ± 70 min (mean ± SD), including a docking time of 25 ± 14 min (mean ± SD) and a console time of 76 ± 67 min (mean ± SD). No intraoperative complication and no conversion were reported. The postoperative complication rate was 3% and the mean hospital stay was 1.18 (± 0.83) days. Full-thickness R0 resection was achieved in 97% of patients (n = 33), three patients were upstaged to T2-stage and underwent anterior resection, and one patient was staged T3. Moreover, the univariate analysis of operative time predictors showed that severe obesity [body mass index (BMI) > 35] was a positive predictor of total operative time and console time, probably because of a narrow space between the legs, whereas specimen size was a positive predictor of total operative time. Finally, the authors concluded that R-TAMIS was safe for lesions located within up to 15 cm from the dentate line and sizing up to 5.5 cm, and that a BMI > 35 was a signi cant predictor of a longer and more challenging operation.
At least three further series with more than 10 patients were also published, one by Huang et al. [41] in 2019 including 23 patients, one by Gómez Ruiz et al. [37] in 2017 including patients, and one by Hompes et al. [39] in 2014 including 16 patients In the first study, robotic procedures were performed in a prone jackknife position, whereas, in the latter two, all procedures were performed using a lithotomy position. Note that Hompes et al. [39] reported the use of a glove port, made up of a surgical glove positioned transanally with ports inserted through the glove's ngers.
Moreover, in 2019, Lee et al. [2] published a study including 19 patients undergoing R-TAMIS and reported that patients with anterior and lateral lesions (83.3% of overall included patients) were operated in a prone position in order to reduce the con ict among robotic arms, whereas the lithotomy position was used only for posterior lesions.

L-TAMIS VS. R-TAMIS
Robotic transanal surgery might offer several advantages compared to conventional TAMIS. Notably, it might increase the possibility of resecting rectal lesions located in difficult sites, reducing the need for proctectomy, as well as it might make surgical wound repair easier.
However, only a few studies designed to compare the outcomes of conventional and robotic TAMIS are available in the literature. The most relevant of these is a retrospective study published by Lee et al. [2] in 2019, comparing the results of 21 patients undergoing conventional TAMIS (indicated as L-TAMIS) and 19 patients undergoing R-TAMIS. Overall, no significant difference was found in terms of perioperative outcomes (notably: total operative time, blood loss, postoperative complications, and length of hospital stay) and pathologic ndings. e only signi cant di erence was represented by direct costs, which were higher in the robotic group ($3562 for L-TAMIS vs. $4441 for R-TAMIS, P = 0.04).

TRANSANAL TOTAL MESORECTAL EXCISION
The latest development of the transanal approach to rectal cancer is represented by transanal total mesorectal excision (TaTME). Indeed, laparoscopic rectal surgery may be challenging because of patientand tumor-related factors. Male obese patients o en show a very limited surgical eld [42,43] , anterior rectal tumors appear to have a higher rate of positive resection margins [43,44] , and determining the distance between rectal tumor and distal staple line is o en di cult [42] . TaTME is a hybrid surgical technique employed for low rectal cancer combining a laparoscopic dissection for colonic mobilization and a mesorectal excision performed using a transanal approach through a GelPOINT platform or a rigid proctoscope. These approaches may be sequential or simultaneous, the specimen is generally extracted through the anus, and a manual coloanal anastomosis is nally performed [45] . Several clinical and pathological factors are considered to indicate a transanal approach in the case of rectal cancer: male gender, narrow and/or deep pelvis, BMI > 30 kg/m 2 , prostatic hypertrophy, tumor located at less than 12 cm from the anal verge, tumor size > 4 cm, and tissue alteration following radiotherapy [46] .
Sylla et al. [47] performed the first clinical case of TaTME in 2010. In 2016, Deijen et al. [48] published a systematic review including overall 794 patients with rectal cancer undergoing this technique. e mean operative time was 244 min and the TME specimen was complete in 87% of procedures. Major postoperative complications occurred in 11.5% of patients and anastomotic leak in 5.7%. Local recurrence rate was 8.1% at 18.9 months. A volume effect was also shown, with high-volume centers (> 30 cases) having higher TME quality, lower rate of positive circumferential resection margin (CRM), and lower rate of major complications.
In 2017, Penna et al. [49] analyzed TaTME short-term outcomes using the data of an international registry including 720 patients. TME was complete in 85% of cases, had minor defects in 11% of cases, and had major defects in 4%. e rate of R1 resection was 2.7%. Among the risk factors of suboptimal TME, perforation, and/or R1 resection, the authors also identi ed a laparoscopic transabdominal posterior dissection to less than 4 cm from the anal verge.
In 2018, Jiang et al. [50] published a systematic review of studies comparing TaTME and laparoscopic TME. They included 762 patients overall operated for cancer of the middle and lower rectum, and showed a signi cant di erence in terms of positive CRM in favor of TaTME, whereas positive distal resection margin and TME quality were similar.

ROBOTIC TATME
TaTME has also been performed using robotic technology, being indicated as robotic TaTME or roboticassisted transanal TME (RATS-TME). However, the studies on this subject currently available in the literature are characterized by a limited number of patients [51 -55] .
Notably, one of the largest was published in 2020 by Hu et al. [56] and included only 20 patients. Eleven of them (55%) were operated for mid rectal cancer and nine (45%) for low rectal cancer. The distance between tumor to anal verge was 6.0 ± 2.7 cm (mean ± SD), the operative team was 172.3 ± 24.2 min (mean ± SD), and the blood loss 82.0 ± 107.1 mL (mean ± SD). No conversion was reported, the postoperative complication rate was 35%, and the length of hospital stay was 8.8 ± 4.2 days. In 90% of cases, TEM was complete, and the number of lymph nodes harvested per specimen was 18.7 ± 6.3 (mean ± SD). e distal margin length was 3.1 ± 1.3 cm. Positive CRM was found in 15% of cases (n = 3).

CONCLUSION
TAMIS represents a valid therapeutic option in the surgical treatment of benign and malignant lesions of the rectum compared to TEM. Robotic transanal surgery soon developed a er the introduction of TAMIS and consists in associating the use of robotic surgical systems to transanal disposable platforms. It proved to be safe and e ective, but the data currently available in the literature do not show clear statistical advantages compared to conventional TAMIS. erefore, further studies will be necessary.