Lacy et al. BMC Cancer (2020) 20:677 https://doi.org/10.1186/s12885-020-07171-y RESEARCH ARTICLE Open Access Three-year outcome after transanal versus laparoscopic total mesorectal excision in locally advanced rectal cancer: a multicenter comparative analysis F. B. de Lacy1* , S. X. Roodbeen2, J. Ríos3, J. van Laarhoven4, A. Otero-Piñeiro1, R. Bravo1, T. Visser5, R. van Poppel5, S. Valverde1, R. Hompes2, C. Sietses5, A. Castells6, W. A. Bemelman2, P. J. Tanis2 and A. M. Lacy1 Abstract Background: For patients with mid and distal rectal cancer, robust evidence on long-term outcome and causal treatment effects of transanal total mesorectal excision (TaTME) is lacking. This multicentre retrospective cohort study aimed to assess whether TaTME reduces locoregional recurrence rate compared to laparoscopic total mesorectal excision (LapTME). Methods: Consecutive patients with rectal cancer within 12 cm from the anal verge and clinical stage II-III were selected from three institutional databases. Outcome after TaTME (Nov 2011 - Feb 2018) was compared to a historical cohort of patients treated with LapTME (Jan 2000 - Feb 2018) using the inverse probability of treatment weights method. The primary endpoint was three-year locoregional recurrence. Results: A total of 710 patients were analysed, 344 in the TaTME group and 366 in the LapTME group. At 3 years, cumulative locoregional recurrence rates were 3.6% (95% CI, 1.1–6.1) in the TaTME group and 9.6% (95% CI, 6.5– 12.7) in the LapTME group (HR = 0.4; 95% CI, 0.23–0.69; p = 0.001). Three-year cumulative disease-free survival rates were 74.3% (95% CI, 68.8–79.8) and 68.6% (95% CI, 63.7–73.5) (HR = 0.82; 95% CI, 0.65–1.02; p = 0.078) and three-year overall survival 87.2% (95% CI, 82.7–91.7) and 82.2% (95% CI, 78.0–86.2) (HR = 0.74; 95% CI, 0.53–1.03; p = 0.077), respectively. In patients who underwent sphincter preservation procedures, TaTME was associated with a significantly better disease-free survival (HR = 0.78; 95% CI, 0.62–0.98; p = 0.033). Conclusions: These findings suggest that TaTME may improve locoregional recurrence and disease-free survival rates among patients with mid and distal locally advanced rectal cancer. Keywords: Rectal cancer, Total mesorectal excision, Locoregional recurrence, TaTME * Correspondence: borjalacy@gmail.com This article was not based on a previous communication to a society or meeting. 1 Department of Gastrointestinal Surgery, Institute of Digestive and Metabolic Diseases, Hospital Clinic, IDIBAPS, Centro de Investigación Biomédica en Red en Enfermedades Hepáticas y Digestivas (CIBERehd), University of Barcelona, Centro Esther Koplowitz, and Cellex Biomedical Research Center, Barcelona, Catalonia, Spain Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Lacy et al. BMC Cancer (2020) 20:677 Page 2 of 12 Background with mid- and low rectal cancer from 2014 onwards. All Each year, 125,000 new cases of rectal cancer are diag- patients with histologically proven rectal adenocarcin- nosed in the European Union [1] and mortality ranges oma treated by TaTME were prospectively registered in between 4 and 10/100,000 per year. The therapeutic ap- a local standardised database or in the International proach to rectal cancer is eminently multidisciplinary, TaTME Registry [11]. Consequently, a multicenter data- but surgery remains the main cornerstone for cure. For base was created, which included the TaTME cohort mid- and low rectal tumours, total mesorectal excision and a cohort of patients treated by LapTME between (TME) along embryological-specific planes is the stand- January 2000 and February 2018, through a retrospective ard surgical treatment [2]. An intact specimen based on analysis of clinical records. All three hospitals used the principles of TME grading, the circumferential resec- TaTME as a standard procedure for patients with tion margin, and the distal resection margin have be- mid- and low rectal cancer until their recent partici- come the most critical factors in predicting the risk of pation in the COLOR III trial, a randomised study in locoregional recurrence and long-term survival [3–5]. which participants are allocated to either TaTME or Primary rectal cancer surgery can be performed LapTME [12]. through an open, laparoscopic, robotic, or transanal ap- For this analysis, adult patients with a solitary locally proach. The oncological superiority of one approach advanced rectal adenocarcinoma, according to the ACO- over the other is still a topic of debate. A recent meta- SOG Z6051 definition (cT3/cT4, or cN1/cN2 with any analysis showed that the risk of a suboptimal TME spe- cT) detected by magnetic resonance imaging (MRI) with cimen is higher with laparoscopy compared to open sur- or without transrectal ultrasonography, within 12 cm of gery [6]. The transanal TME (TaTME) was developed to the anal verge treated with TaTME or LapTME were in- improve patient outcomes and the quality of the dissec- cluded [13]. The exclusion criteria were: patients with tion, which is believed to be especially significant in cTisN0 or cT1-2 N0; pelvic malignancy within 5 years; mid- and low rectal tumours. Studies published to date severe, incapacitating disease, i.e. American Society of have reported mesorectal excision completeness rates as Anaesthesiologists (ASA) classification IV-V; procedures high as 92.5 to 96%, and an even significantly higher rate performed in an emergency setting; tumours previously of complete and near-complete mesorectal excisions treated by local excision; unknown cT or cM; metastatic compared to conventional laparoscopic TME (LapTME) tumours (M1); synchronous tumours; active Crohn’s or [7–9]. These clinical findings suggest that TaTME is a ulcerative colitis; familial risk-colorectal cancer syn- highly promising technique, although the translation of dromes; and patients with 30-day mortality when it was this data into improved mid- and long-term oncological judged to have occurred as a direct result of a major ac- outcomes has yet to be proven. Unexpectedly, a recent tive postoperative complication, which is not of primary study reported a pattern of rapid and multifocal locore- interest. The Institutional Ethics Committees (Comité de gional recurrence after TaTME [10]. Therefore, this Ética de la Investigación con Medicamentos, Beoor- multicenter international study was designed with the dlingscommissie Wetenschappelijk Onderzoek, and goal of comparing the three-year oncological outcome of Medisch Ethische Toetsings Commissie AMC) approved patients with primary locally advanced rectal cancer the TaTME and LapTME techniques years prior to this treated with TaTME and LapTME in three high-volume study in the three institutions noted, and the current rectal cancer centers. study protocol was assessed and accepted by the local Institutional Review Boards. Patients provided written Methods informed consent. Study population Data were obtained from one Spanish center, the Hos- Endpoints pital Clinic of Barcelona, and two Dutch centers, the The primary endpoint was three-year locoregional recur- Gelderse Vallei Hospital and the Amsterdam UMC at rence. Secondary endpoints included systemic recur- AMC. LapTME was first introduced at the Hospital rence, disease-free survival, and overall survival. Clinic in 1994. In November 2011, TaTME became the standard of care for all patients presenting with rectal Procedures and definitions cancer that did not require abdominoperineal resection The specific staging, classification methods, and surgical or pelvic exenteration. In February 2017, the transanal procedures have been described in more detail previ- approach became standard for patients requiring an ously [14–16]. Tumours were considered high if the dis- abdominoperineal resection. Gelderse Vallei Hospital is tal border of the tumour was > 10 cm from the anal a high-volume rectal cancer institution in which TaTME verge, mid if it was between 5 and 10 cm, and low in was first used in 2012. At the Amsterdam UMC TaTME case of a distal border < 5 cm. Patients were eligible for became the standard procedure for patients presenting neoadjuvant therapy in cases of cT3b-d/cT4 or cN- Lacy et al. BMC Cancer (2020) 20:677 Page 3 of 12 positive tumours below the peritoneal reflection, or if was defined as any recurrence in the pelvic area and had the circumferential resection margin was threatened or to be confirmed at least on imaging. involved, although other factors such as extramural ven- ous invasion were also taken into account and discussed Statistical analysis by a multidisciplinary team. The indication to receive Qualitative variables were expressed as absolute frequen- radiotherapy alone or in combination with chemother- cies and percentages. Quantitative variables were re- apy was given depending on the institution-specific pro- ported as means or medians with their 95% confidence tocols. Short-course one-week radiotherapy was intervals (CI), except for follow-up periods, which were administered by 25 Gy in five daily fractions. Neoadju- expressed as median with range. To allow for an un- vant long-course chemoradiotherapy was administered biased comparison, an inverse probability of treatment by continuous 5-fluorouracil (5-FU) infusion (225 mg/ weights approach was used [18]. A logistic regression m2 for 5 days per week) or capecitabine (825 mg/m2 model was applied, including demographic and clinical twice daily for 5 days per week), and a total dosage of 45 preoperative variables such as hospital, age, gender, ASA Gy, by a weekly dose of 9 Gy divided in 5 days each classification, body mass index (BMI), the distance of the week, for a total of 5 weeks. The interval between com- tumour from the anal verge, cT and cN stage, and base- pletion of long-course chemoradiotherapy and surgery line threatened or involved circumferential resection was 5 to 7 weeks at the beginning of the LapTME cohort margin. As the goal was to develop a balanced popula- recruitment, and, in accordance with current guidelines, tion that was independent of the outcome assessment, was subsequently extended up to 12 weeks and associ- postoperative variables were not included, except for ated with appropriate restaging [1]. relevant variables that were assessable only after surgery The mesorectal specimen was analysed on the basis of (i.e., pT and pN stage, and pathological response to neo- four major pathological factors: the integrity of the mesor- adjuvant therapy). The covariate radiologic extramural ectum, graded following the Quirke method: complete, venous invasion was excluded from the model due to a near-complete, or incomplete [5]; the circumferential re- significant amount of missing data because this informa- section margin, considered to be positive when the dis- tion was not routinely reported until recent years. The tance between the deepest portion of the tumour and the covariate type of surgery (categorised as sphincter-saving resection margin was ≤1 mm, or in the case of a positive or abdominoperineal resection) could not be included in lymph node at ≤1 mm of the radical dissection plane; the the propensity score calculation due to the small num- distal resection margin, considered to be positive if ber of abdominoperineal resections in the TaTME tumour cells were present ≤1 mm from the lower border group. of the tumour to the cut edge of the specimen; and the A well-balanced distribution of the covariates in the number of lymph nodes harvested. Pathological tumour weighted sample was confirmed by means standardised response to neoadjuvant therapy was scored by the Ryan differences meeting a standard objective of ±0.10 [19]. tumour regression grade (three-point TRG): TRG 1, no vi- The only exception was the covariate hospital, in which able cancer cells, or single cells or small groups of cancer a standardised difference of 0.12 was achieved. Since cells; TRG 2, residual cancer outgrown by fibrosis; and some authors consider the cut-off point for standardised TRG 3, significant fibrosis outgrown by cancer, or no fi- differences to be ±0.20, and given the extensive homo- brosis with extensive residual cancer [17]. geneity of the rest of covariates (with standardised differ- ences of less than ±0.02), this was finally accepted [20]. Follow-up The estimation of the survival functions was carried Until follow-up was completed after 5 years, patients vis- out using the Kaplan-Meier method. The estimation of ited every 3 to 6 months during the first 2 years and every the effect of surgical procedure was performed using 6 to 12 months during the remaining 3 years. Visits in- Cox hazard models weighted by the inverse probability cluded a history, physical evaluation with digital rectal of treatment weight adjusting by preservation of the examination, and determination of the carcinoembryonic sphincter (sphincter-saving surgery or abdominoperineal antigen level. In Barcelona, imaging studies with thoracic resection) with a cut-off at 3 years. Additional analyses and abdominopelvic CT scans were requested every 6 using Accelerated Failure Time models were used for months during the first 2 years and annually during the the analysis of time to event data in order to estimate remaining 3 years. In both Gelderse Vallei and Amsterdam the time ratio (TR) for the effect of the surgical proced- UMC, imaging study was based on liver ultrasound, and ure on acceleration in the time to the event [21]. CT scan was performed in case of suspicion of local recur- Statistical tests were two-sided with a 5% type I error. rence or distant metastasis. Pelvic MRI and/or transrectal All the analyses were carried out using SPSS version 25 ultrasound-guided needle biopsy were requested when (IBM) or SAS version 9.4 (SAS Institute, Inc., Cary, pelvic recurrence was suspected. Locoregional recurrence North Carolina). Lacy et al. BMC Cancer (2020) 20:677 Page 4 of 12 Results TaTME or LapTME, with standardised differences be- Between 2000 and 2018, 863 patients with primary lo- fore and after the inverse probability of treatment cally advanced rectal adenocarcinoma met the inclusion weight. criteria and were eligible. Of these, 153 (17.7%) were ex- In both groups, similar rates of neoadjuvant therapy cluded because data were missing in the covariates se- administration were observed (71.2% vs. 77.0%; p = lected for the inverse probability of treatment weight. 0.086). Short-course radiotherapy was given to 59 This occurred mainly at the beginning of the LapTME (24.5%) patients in the TaTME cohort and to 79 (28.1%) cohort, because either systematization in reporting all patients in the LapTME cohort (p = 0.408), while most the information was scarce, or the necessary investiga- patients in both groups received long-course chemora- tions to obtain that data were not always performed in diotherapy: 180 (52.3%) vs. 203 (55.4%) (p = 0.371). that period. Therefore, the final study population con- TaTME was associated with a significant reduction in sisted of 710 patients, of whom 344 (48.5%) underwent abdominoperineal resection rates: 2.9% vs. 25.6%, p < TaTME and 366 (51.5%) LapTME (Fig. 1). The median 0.001. Similar rates of 30-day postoperative complica- follow-up in the TaTME and LapTME cohorts was 28.4 tions were observed (31.9% vs. 35.2%; p = 0.382). (range 0.1–83.6) and 61.1 (range 1.1–205.7) months, re- spectively. After truncation at 3 years, more than 30% of Histopathological outcomes and adjuvant therapy the patients remained at risk in both groups. Table 1 Despite an absence of significant differences in the ori- shows the selected covariates of patients treated with ginal multi-category mesorectal specimen variable, a Fig. 1 Flow diagram of study population. Abbreviations: TME, total mesorectal excision; AMC, Amsterdam University Medical Centers; ASA, American Society of Anesthesiologists; BMI, body mass index; CRM, circumferential resection margin; NAT, neoadjuvant therapy; LapTME, laparoscopic total mesorectal excision; TaTME, transanal total mesorectal excision Lacy et al. BMC Cancer (2020) 20:677 Page 5 of 12 Table 1 Selected covariates of patients treated with laparoscopic or transanal total mesorectal excision for rectal cancer, with standardised differences before and after inverse probability of treatment weighting Variable Surgical approach Standardised differences TaTME, No. (%) LapTME, No. (%) Before IPTW After IPTW Total, No. 344 366 Center Barcelona 194 (56.4) 212 (57.9) 0.1189 0.1238 Gelderse Vallei 90 (26.1) 79 (21.5) AMC 60 (17.4) 75 (20.4) Age, years (mean, 95% CI) 66.5 (65.2–67.7) 66.4 (65.2–67.6) 0.0071 < 0.0001 Gender Female 104 (30.2) 137 (37.4) 0.1526 −0.007 Male 240 (69.7) 229 (62.5) ASA I 44 (12.7) 68 (18.5) 0.0682 0.0002 II 239 (69.4) 227 (62.0) III 61 (17.7) 71 (19.4) BMI, kg/m2 (mean, 95% CI) 25.5 (25.1–25.9) 26.4 (25.7–27.0) −0.161 −0.017 Distance from AV, cm (mean, 95% CI) 7.2 (6.9–7.5) 6.5 (6.1–6.8) 0.2487 0.0024 Clinical T-stage cT1 0 (0.0) 4 (1.1) 0.0726 0.0047 cT2 27 (7.8) 41 (11.8) cT3 289 (84.0) 266 (76.6) cT4 28 (8.1) 35 (10.0) Clinical N-stage cN0 148 (43.0) 113 (30.8) −0.288 0.0005 cN1 155 (45.0) 181 (49.4) cN2 41 (11.9) 72 (19.6) Baseline threatened/involved CRM 94 (27.3) 101 (27.6) −0.006 −0.009 Pathologic response to NATa TRG 1 93 (38.2) 103 (36.5) −0.144 −0.003 TRG 2 76 (31.2) 85 (30.1) TRG 3 74 (30.4) 94 (33.3) Pathological T-stage −0.144 −0.003 pT0 40 (11.6) 35 (9.5) −0.144 −0.003 pTis 1 (0.2) 0 (0.0) pT1 26 (7.5) 20 (5.4) pT2 85 (24.7) 95 (25.9) pT3 179 (52.0) 179 (48.9) pT4 13 (3.7) 37 (10.1) Pathological N-stage pN0 251 (72.9) 243 (66.3) −0.144 −0.003 pN1 57 (16.5) 76 (20.7) pN2 33 (9.5) 47 (12.8) Variable Surgical approach Standardised differences TaTME, No. (%) LapTME, No. (%) Before IPTW After IPTW Lacy et al. BMC Cancer (2020) 20:677 Page 6 of 12 Table 1 Selected covariates of patients treated with laparoscopic or transanal total mesorectal excision for rectal cancer, with standardised differences before and after inverse probability of treatment weighting (Continued) Variable Surgical approach Standardised differences TaTME, No. (%) LapTME, No. (%) Before IPTW After IPTW pN3 0 (0.0) 0 (0.0) pN1c 3 (0.8) 0 (0.0) b Sphincter saving surgery 334 (97.0) 272 (74.3) NA NA Abbreviations: TaTME, transanal total mesorectal excision; LapTME, laparoscopic total mesorectal excision; IPTW, inverse probability of treatment weighting; AMC, Amsterdam University Medical Centers; ASA, American Society of Anesthesiologists; BMI, body mass index; AV, anal verge; CRM, circumferential resection margin; NAT, neoadjuvant therapy; TRG, tumour regression grade; NA, not applicable a Including only patients treated with NAT. The TRG system developed by Ryan et al. was used17 b Not included in the IPTW calculation due to the large differences between groups. It was used as an adjustment cofactor in Cox models significant higher rate of complete or near-complete was Within 3 years after primary rectal cancer surgery, 35 observed in the TaTME cohort: 98.5% vs. 93.5%, p = locoregional recurrences were observed in the LapTME 0.0003. The rate of circumferential resection margin in- cohort, with a corresponding number of 12 in the volvement and incidence of intra-operative rectal perfor- TaTME cohort. Of those patients, 25 (six in the TaTME ation were also lower in the TaTME group (Table 2). group and 19 in the LapTME group) presented with at These findings translated into an overall better compos- least one of the following risks factors: T4 tumour, N2 ite endpoint of poor pathological outcome for TaTME. disease, incomplete mesorectal specimen, or positive cir- cumferential resection margin. The median time to locoregional recurrence could not be calculated since Survival and recurrence analyses the event rate was less than 50%. However, the Acceler- Three years after surgery, the rates of locoregional recur- ated Failure Time analysis identified a longer time-ratio rence were 3.6% in the TaTME group and 9.6% in the in the TaTME group TR = 2.3 (CI, 1.34–4.00; p = 0.026). LapTME group Hazard Ratio (HR) = 0.4 (95% CI, 0.23– No multifocal pattern of recurrence was diagnosed. 0.69; p = 0.001) (Fig. 2). After stratifying for sphincter- saving surgery or abdominoperineal resection, a lower Discussion rate of locoregional recurrence was maintained in those In this multicenter cohort of 710 patients with clinical patients who underwent TaTME with sphincter preser- stage II-III rectal adenocarcinoma, TaTME provided a vation HR = 0.42 (95% CI, 0.24–0.73; p = 0.002). No dif- three-year 60% risk reduction for locoregional recur- ference was observed in patients with low rectal cancer rence compared to LapTME. In patients undergoing sur- HR = 0.9 (95% CI, 0.28–2.93; p = 0.866). In patients with gery with sphincter preservation, the three-year disease- cancer of the mid rectum, the rates of locoregional re- free survival rate was higher for patients treated with currence were 5.3% in de TaTME group and 12.3% in TaTME than for patients treated with LapTME. These the LapTME group HR = 0.39 (95% CI, 0.2–0.76; p = benefits could be explained by an improved quality of 0.006). Systemic metastases were reported in 16.4% of the mesorectal specimen, with fewer positive resection the patients in the TaTME group and in 19.8% of the margins, and lower rate of rectal perforations [4, 22, 23]. patients in the LapTME group HR = 0.93 (95% CI, 0.7– The performance of an optimal TME is technically de- 1.24; p = 0.615). manding, and the histopathological equivalence of the At 3 years, the disease-free survival rates were 74.3% in laparoscopic and open approaches has been recently the TaTME group and 68.6% in the LapTME group questioned [13, 24]. Fleshman et al. compared LapTME HR = 0.81 (95% CI, 0.65–1.02; p = 0.078) (Fig. 3). How- to open TME and included similar patients as the ever, when the analysis was limited to patients with current study, except for the fact that every patient re- sphincter preservation, an improved disease-free survival ceived neoadjuvant therapy. They found that LapTME was observed in patients who underwent TaTME HR = did not meet the criteria for noninferiority in a compos- 0.78 (95% CI, 0.62–0.98; p = 0.033). The overall survival ite score of complete or near-complete TME, and nega- rates were 87.2% in the TaTME group and 82.2% in the tive circumferential and distal resection margins [13]. It LapTME group HR = 0.74 (95% CI, 0.53–1.03; p = is important to note that Fleshman et al. did not find 0.076). Significant differences in overall survival could any difference in survival within 2 years after surgery not be demonstrated in patients who underwent TaTME [25]. Still, the study was not designed as an equivalence with sphincter preservation HR = 0.73 (95% CI, 0.52– trial for survival and recurrence, and the absence of nu- 1.02; p = 0.068). The survival and recurrence subgroup merical distinctions might not be indicative of any analyses are shown in Fig. 4. dissimilarity. Lacy et al. BMC Cancer (2020) 20:677 Page 7 of 12 Table 2 Pathologic and adjuvant therapy outcomes after inverse probability of treatment weighting Variable Surgical approach P Value TaTME, No. (%) LapTME, No. (%) Total, No. 344 366 AJCC pathological stage 0 38 (11.0) 32 (8.7) 0.8616 I 90 (26.1) 92 (25.1) II 123 (35.7) 119 (32.5) III 93 (27.0) 123 (33.6) IV 0 (0.0) 0 (0.0) Mesorectal specimen Complete 318 (93.2) 242 (89.3) 0.1678 Near-complete 20 (5.8) 13 (4.8) Incomplete 3 (0.8) 16 (5.9) Distance to CRM, mm (median, 95% CI) 10.0 (10.0–12.0) 7.5 (6.0–10.0) 0.0131 CRM involvement 32 (9.5) 56 (16.2) 0.0038 Distance to DRM, mm (median, 95% CI) 20.0 (20.0–25.0) 19.5 (15.0–20.0) 0.248 DRM involvement 6 (1.8) 7 (2.0) 0.6135 Rectal perforation 2 (0.8) 8 (3.2) 0.0262 Composite poor pathological outcomea 35 (10.6) 69 (24.7) < 0.001 Perineural invasion 44 (13.0) 47 (18.3) 0.0109 Lymphovascular invasion 68 (21.4) 44 (17.0) 0.0182 Budding no 155 (82.8) 38 (52.7) 0.0002 low 23 (12.3) 32 (44.4) moderate 2 (1.0) 0 (0.0) high 7 (3.7) 2 (2.7) Differential grade 0.5589 good 20 (6.3) 15 (4.8) 0.5589 moderate 254 (80.3) 240 (77.4) poor 17 (5.3) 22 (7.1) Number of lymph node harvested (median, 95% CI) 15.0 (15.0–16.0) 14.0 (14.0–15.0) 0.0133 Adjuvant chemotherapy 42 (12.2) 61 (17.1) 0.0508 Adjuvant radiotherapy 4 (1.1) 15 (4.2) 0.0002 Abbreviations: TaTME, transanal total mesorectal excision; LapTME, laparoscopic total mesorectal excision; AJCC, American Joint Committee on Cancer; CRM, circumferential resection margin; DRM, distal resection margin a Complete or near-complete TME, and negative CRM and DRM Nevertheless, several clinical studies and meta-analyses The reported improvement in locoregional recurrence have reported improved histopathological outcomes with rates was at the expense of patients with cancer of the TaTME compared to LapTME [26, 27]. However, the mid rectum, but unexpectedly this was not confirmed in translation of these potentially advantageous data into patients with cancer of the low rectum. This apparent improved mid- and long-term oncological outcomes is disparity could be explained by a high proportion of pa- still scarce. The outcomes of the present study indicate tients, not estimable due to retrospective access to the that TaTME not only allows patients to benefit from the data, in the LapTME group with low rectal tumours who short-term advantages of minimally invasive surgery, but underwent open perineal dissection using the transanal might also be superior in ensuring mid-term locoregio- way to facilitate the most challenging part of the proced- nal recurrence and disease-free survival in selected ure. According to a randomised trial, transanal perineal patients. dissection has been shown to decrease the risk of Lacy et al. BMC Cancer (2020) 20:677 Page 8 of 12 Fig. 2 Three-year locoregional recurrence between TaTME and LapTME in patients with locally advanced rectal cancer. Abbreviations: TaTME, transanal total mesorectal excision; LapTME, laparoscopic total mesorectal excision; HR, hazard ratio; CI, confidence interval circumferential resection margin involvement by more surgical instruments or through TaTME in the low rec- than 4-fold compared to a purely abdominal TME [28]. tum. Given the difficulties of approaching mid rectal Although further investigation is required, these com- cancers using an open transanal technique (the Transa- bined data suggest that the surgical therapy of mid and nal Abdominal Transanal (TATA) procedure) and the low rectal cancer should include a perineal approach. inferiority of a pure transabdominal laparoscopic ap- This might be performed with conventional open proach as suggested by the present data, TaTME might Fig. 3 Three-year disease-free survival between TaTME and LapTME in patients with locally advanced rectal cancer. Abbreviations: TaTME, transanal total mesorectal excision; LapTME, laparoscopic total mesorectal excision; HR, hazard ratio; CI, confidence interval Lacy et al. BMC Cancer (2020) 20:677 Page 9 of 12 Fig. 4 Subgroup analyses of survival and recurrence among patients with locally advanced rectal cancer treated with TaTME or LapTME. Abbreviations: HR, hazard ratio; CI, confidence interval; APR, abdominoperineal resection; TaTME, transanal total mesorectal excision; LapTME, laparoscopic total mesorectal excision be the preferred technique when the tumour is in the consecutive patients undergoing TaTME with a mid rectum. Robot assisted transabdominal laparoscopy complete and minimum follow-up of 3 years, report- might potentially achieve similar results, but this has still ing three- and five-year local relapse as low as 2 and to be confirmed. 4%, respectively [31]. Although knowledge in the literature is still scarce However, a recent study questioned the oncologic val- and follow-up periods are relatively short, the re- idity of the transanal approach. Larsen et al. reported, on ported data that TaTME might be associated with a behalf of the Norwegian Colorectal Cancer Group, that lower risk of locoregional recurrence are substantiated 9.5% of the 110 patients who underwent TaTME pre- by several observational studies. Tuech et al. analysed sented with an unexpected pattern of early locoregional 56 consecutive patients with low rectal cancer treated recurrence, characterised by rapid, multifocal growth in with TaTME and reported a locoregional recurrence the pelvic cavity and sidewalls [10]. The authors sug- rate of 1.7% with a median follow-up of 29 months gested that this atypical pattern of relapse may be a con- [29]. Veltcamp Helbach et al. analysed 80 patients sequence of transanal pursestring failure, with spillage of with mid- or low rectal cancer who underwent the malignant cells that are aerosolised by the transanal TaTME, and the locoregional recurrence rate after 2 insufflator. However, neither in the present study nor years was 2.5% [16]. With a median follow-up of 31.9 any of the published studies to date has revealed an un- months, Lelong et al. reported a 0% locoregional re- expected pattern of recurrence. Moreover, the Acceler- lapse rate [30]. More recently, Hol et al. analysed 159 ated Failure Time analysis of our study identified Lacy et al. BMC Cancer (2020) 20:677 Page 10 of 12 TaTME as a protective factor on patients’ locoregional among patients with locally advanced rectal cancer. recurrence time, suggesting a longer time to pelvic re- Further investigation in a randomised clinical trial is lapse in that group compared to patients treated with warranted. LapTME. This clinical research study was based on real- Supplementary information world clinical practice and involved several groups Supplementary information accompanies this paper at https://doi.org/10. 1186/s12885-020-07171-y. of surgeons to enhance external validity. However, a significant limitation is its nonrandomised design. Additional file 1: Supplementary file 1. DF_AC (ICMJE Form from Dr. Observational studies are more susceptible to Antoni Castells). biases, even more so with surgical interventions Additional file 2: Supplementary file 2. DF_AML (ICMJE Form from where the risk of treatment assignment partiality is Dr. Antonio M. Lacy). increased during the early phase of the learning Additional file 3: Supplementary file 3. DF_AO (ICMJE Form from Dr. curve. To avoid this allocation bias, we decided to Ana Otero). use the inverse probability of treatment weight Additional file 4: Supplementary file 4. DF_CS (ICMJE Form from Dr. Colin Sietses). method, which has been shown to deliver results Additional file 5: Supplementary file 5. DF_FBL (ICMJE Form from more comparable to an RCT than other techniques Dr. F. Borja de Lacy). such as propensity score stratification and, unlike Additional file 6: Supplementary file 6. DF_JR (ICMJE Form from Mr. matching, retains most participant data [32]. After José Ríos). weighting, the covariates of the sample obtained Additional file 7: Supplementary file 7. DF_JVL (ICMJE Form from were well-balanced and independent of treatment Dr. Jacqueline van Laarhovene). assignment. The only exception was the covariate Additional file 8: Supplementary file 8. DF_PJT (ICMJE Form from Dr. Pieter J. Tanis). type of surgery, which was applied as an adjustment Additional file 9: Supplementary file 9. DF_RB (ICMJE Form from Dr. cofactor in the Cox models. Nevertheless, we were Raquel Bravo). unable to correct for unknown cofounders, and the Additional file 10: Supplementary file 10. DF_RH (ICMJE Form from analysis by subgroup depending on the type of sur- Dr. Roel Hompes). gery should be interpreted with caution due the low Additional file 11: Supplementary file 11. DF_RvP (ICMJE Form from number of abdominoperineal resections in the Dr. Roy van Poppel). TaTME cohort. Additional file 12: Supplementary file 12. DF_SV (ICMJE Form from Dr. Silvia Valverde). Another limitation is the inherent retrospective Additional file 13: Supplementary file 13. DF_SXR (ICMJE Form from design which used historical controls. Besides, the Dr. Sapho X. Roodbeen). extramural venous invasion variable could not be Additional file 14: Supplementary file 14. DF_TV (ICMJE Form from included in the propensity score model due to the Dr. Tjaakje Visser). large amount of missing data. This occurred pre- Additional file 15: Supplementary file 15. DF_WAB (ICMJE Form dominantly in the LapTME group because the from Dr. Willem A. Bemelman). radiological extramural venous invasion has recently begun to be described. Variables such as budding, Abbreviations TME: Total mesorectal excision; TaTME: Transanal total mesorectal excision; perineural and lymphovascular invasion displayed a LapTME: Laparoscopic total mesorectal excision; MRI: Magnetic resonance heterogeneous distribution. However, the increased imaging; ASA: American Society of Anaesthesiologists; M1: Metastatic risk of recurrence that this may carry seems to be tumours; TRG: Tumour regression grade; BMI: Body mass index; CI: Confidence interval; TR: Time ratio; HR: Hazard ratio; TATA: Transanal offset across the groups. Nonetheless, despite our abdominal transanal procedure extensive corrections, the presence of cofounders that might bias the results cannot be excluded. Fi- Acknowledgements The authors are grateful to Georgina Casanovas and Gemma Domenech, nally, a representative non-selected group of pa- who assisted our senior statistician (JR) in performing the statistical analysis. tients who were treated for locally advanced rectal Thanks also to the Research, Innovation, and Education Departments of the tumours was included. However, the surgical teams Hospital Clinic of Barcelona for funding the study. have extensive experience performing transanal pro- Authors’ contributions cedures, and the results may not be generalised to All authors have read and approved the manuscript. Furthermore, the other clinics that have recently started to perform specific contributions from each author were: Study conception and design: FBL, SXR, JR, JVL, RH, AC, AML. Acquisition of data: FBL, SXR, AOP, TV, RVP, SV. TaTME. Analysis of data: FBL, JR. Interpretation of data: FBL, JR, SXR, RH, CS, AC, WAB, PJT, AML. Drafting of manuscript: FBL, JR, SXR. Critical revision: JVL, AOP, RB, Conclusions TV, RVP, SV, RH, CS, AC, WAB, PJT, AML. The results of this multicenter observational trial sup- Funding port a possible role for TaTME in improving locore- This work has been funded by the Resident Award “Emili Letang,” granted gional recurrence and disease-free survival rates by Hospital Clinic of Barcelona, Research, Innovation, and Education Lacy et al. BMC Cancer (2020) 20:677 Page 11 of 12 Departments (Grant number: 25_delacyoliverb_250709_cgicm_pfr2018). The study of lateral tumour spread and surgical excision. Lancet. 1986;2(8514): funder of this study supported the data collection and database 996–9. management. The funder of this study had no influence in the design of the 4. Nagtegaal ID, Quirke P. What is the role for the circumferential margin in study, the collection, analysis and interpretation of the data, or writing of the the modern treatment of rectal cancer? J Clin Oncol. 2008;26(2):303–12. report. The corresponding author had full access to all the data in the study 5. Nagtegaal ID, van de Velde CJ, van der Worp E, Kapiteijn E, Quirke P, van and had responsibility for the decision to submit for publication. Krieken JH, et al. Macroscopic evaluation of rectal cancer resection specimen: clinical significance of the pathologist in quality control. J Clin Availability of data and materials Oncol. 2002;20(7):1729–34. The datasets used and analysed, together with the generated syntaxes 6. Martinez-Perez A, Carra MC, Brunetti F, de’Angelis N. Pathologic outcomes (coding), during the current study are available from the corresponding of laparoscopic vs open mesorectal excision for rectal cancer: a systematic author and Mr. José Ríos on reasonable request. review and meta-analysis. JAMA Surg. 2017;152(4):e165665. 7. Buchs NC, Wynn G, Austin R, Penna M, Findlay JM, Bloemendaal AL, et al. A Ethics approval and consent to participate two-centre experience of transanal total mesorectal excision. Color Dis. The Institutional Ethics Committees approved the TaTME and LapTME 2016;18(12):1154–61. techniques years prior to this study in the three institutions noted, and all 8. Chen CC, Lai YL, Jiang JK, Chu CH, Huang IP, Chen WS, et al. Transanal total patients signed informed consent. The Ethics Committee of the Hospital mesorectal excision versus laparoscopic surgery for rectal cancer receiving Clinic of Barcelona (named Comité Ético de Investigación Clínica), where the neoadjuvant chemoradiation: a matched case-control study. Ann Surg Principal Investigator (FBL) belongs, approved the feasibility of the study Oncol. 2016;23(4):1169–76. (reference HCB/2016/0640). The current study protocol was assessed and 9. Velthuis S, Nieuwenhuis DH, Ruijter TE, Cuesta MA, Bonjer HJ, Sietses C. accepted by the local Institutional Review Boards. Transanal versus traditional laparoscopic total mesorectal excision for rectal carcinoma. Surg Endosc. 2014;28(12):3494–9. Consent for publication 10. Larsen SG, Pfeffer F, Korner H. Norwegian colorectal cancer G. Norwegian Not applicable. moratorium on transanal total mesorectal excision. Br J Surg. 2019;106(9): 1120–1. Competing interests 11. Penna M, Hompes R, Arnold S, Wynn G, Austin R, Warusavitarne J, et al. Dr. F. Borja de Lacy, Drs. Sapho Xenia Roodbeen, Mr. Jose Ríos, Dr. Jacqueline Incidence and risk factors for anastomotic failure in 1594 patients treated by van Laarhoven, Dr. Ana Otero-Piñeiro, Dr. Raquel Bravo, Dr. Tjaakje Visser, Dr. transanal total mesorectal excision: results from the international TaTME Roy van Poppel, and Dr. Silvia Valverde have no conflicts of interest or finan- registry. Ann Surg. 2018;269:700–711. cial ties to disclose. Dr. Roel Hompes reports an educational grant from Stry- 12. Deijen CL, Velthuis S, Tsai A, Mavroveli S, de Lange-de Klerk ES, Sietses C, ker, personal fees from Applied Medical outside the submitted work. Dr. et al. COLOR III: a multicentre randomised clinical trial comparing transanal Colin Sietses reports personal fees from Medtronic, personal fees from Olym- TME versus laparoscopic TME for mid and low rectal cancer. Surg Endosc. pus, and personal fees from AFS medical, outside the submitted work. Dr. 2016;30(8):3210–5. Antoni Castells reports personal fees from Amadix, Goodgut and Universal 13. Fleshman J, Branda M, Sargent DJ, Boller AM, George V, Abbas M, et al. Diagnostics, and grants from SAF2014 and AECC, outside the submitted Effect of laparoscopic-assisted resection vs open resection of stage II or III work. Dr. Willem A. Bemelman reports grants from VIFOR, grants from Med- rectal cancer on pathologic outcomes: the ACOSOG Z6051 randomized tronic, and grants from Braun, outside the submitted work. Dr. Pieter J. Tanis clinical trial. JAMA. 2015;314(13):1346–55. reports personal fees from Johnson & Johnson, personal fees from Olympus, 14. Delgado S, Momblan D, Salvador L, Bravo R, Castells A, Ibarzabal A, et al. and personal fees from B Braun, and research grant from Life Cell, outside Laparoscopic-assisted approach in rectal cancer patients: lessons learned the submitted work. Dr. Antonio M. Lacy reports personal fees from Medtro- from >200 patients. Surg Endosc. 2004;18(10):1457–62. nic, personal fees from Olympus, personal fees from Applied Medical, and 15. Lacy AM, Tasende MM, Delgado S, Fernandez-Hevia M, Jimenez M, De Lacy personal fees from Conmed, outside the submitted work. B, et al. Transanal total mesorectal excision for rectal cancer: outcomes after 140 patients. J Am Coll Surg. 2015;221(2):415–23. Author details 16. Veltcamp Helbach M, Deijen CL, Velthuis S, Bonjer HJ, Tuynman JB, Sietses 1 C. Transanal total mesorectal excision for rectal carcinoma: short-term Department of Gastrointestinal Surgery, Institute of Digestive and Metabolic Diseases, Hospital Clinic, IDIBAPS, Centro de Investigación Biomédica en Red outcomes and experience after 80 cases. Surg Endosc. 2016;30(2):464–70. en Enfermedades Hepáticas y Digestivas (CIBERehd), University of Barcelona, 17. Ryan R, Gibbons D, Hyland JM, Treanor D, White A, Mulcahy HE, et al. Centro Esther Koplowitz, and Cellex Biomedical Research Center, Barcelona, Pathological response following long-course neoadjuvant Catalonia, Spain. 2Department of Surgery, Amsterdam University Medical chemoradiotherapy for locally advanced rectal cancer. Histopathology. 2005; Centers, University of Amsterdam, Cancer Center Amsterdam, Meibergdreef 47(2):141–6. 9, Amsterdam, The Netherlands. 3Medical Statistics Core Facility, August Pi 18. Austin PC. An introduction to propensity score methods for reducing the and Sunyer Biomedical Research Institute (IDIBAPS); Biostatistics Unit, Faculty effects of confounding in observational studies. Multivar Behav Res. 2011; of Medicine, Universitat Autònoma de Barcelona, Villarroel 170, 08036 46(3):399–424. Barcelona, Catalonia, Spain. 4Department of General Surgery, Jeroen Bosch 19. Lee BK, Lessler J, Stuart EA. Improving propensity score weighting using Ziekenhuis, ‘s Hertogenbosch, The Netherlands. 5Department of Surgery, machine learning. Stat Med. 2010;29(3):337–46. Gelderse Vallei Hospital, Ede, The Netherlands. 6Department of 20. Austin PC. Balance diagnostics for comparing the distribution of baseline Gastroenterology, Institute of Digestive and Metabolic Diseases, Hospital covariates between treatment groups in propensity-score matched samples. Clinic, IDIBAPS, Centro de Investigación Biomédica en Red en Enfermedades Stat Med. 2009;28(25):3083–107. Hepáticas y Digestivas (CIBERehd), University of Barcelona, Barcelona, 21. Patel K, Kay R, Rowell L. Comparing proportional hazards and accelerated Catalonia, Spain. failure time models: an application in influenza. Pharm Stat. 2006;5(3):213– 24. Received: 12 May 2020 Accepted: 12 July 2020 22. Kitz J, Fokas E, Beissbarth T, Strobel P, Wittekind C, Hartmann A, et al. Association of plane of total mesorectal excision with prognosis of rectal cancer: secondary analysis of the CAO/ARO/AIO-04 phase 3 randomized References clinical trial. JAMA Surg. 2018;153(8):e181607. 1. Glynne-Jones R, Wyrwicz L, Tiret E, Brown G, Rodel C, Cervantes A, et al. 23. Kusters M, Marijnen CA, van de Velde CJ, Rutten HJ, Lahaye MJ, Kim JH, Rectal cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment et al. Patterns of local recurrence in rectal cancer; a study of the Dutch TME and follow-up. Ann Oncol. 2017;28(suppl_4):iv22–40. trial. Eur J Surg Oncol. 2010;36(5):470–6. 2. MacFarlane JK, Ryall RD, Heald RJ. Mesorectal excision for rectal cancer. 24. Stevenson AR, Solomon MJ, Lumley JW, Hewett P, Clouston AD, Gebski VJ, Lancet. 1993;341(8843):457–60. et al. Effect of laparoscopic-assisted resection vs open resection on 3. Quirke P, Durdey P, Dixon MF, Williams NS. Local recurrence of rectal pathological outcomes in rectal cancer: the ALaCaRT randomized clinical adenocarcinoma due to inadequate surgical resection. Histopathological trial. JAMA. 2015;314(13):1356–63. Lacy et al. BMC Cancer (2020) 20:677 Page 12 of 12 25. Fleshman J, Branda ME, Sargent DJ, Boller AM, George VV, Abbas MA, et al. Disease-free survival and local recurrence for laparoscopic resection compared with open resection of stage II to III rectal cancer: follow-up results of the ACOSOG Z6051 randomized controlled trial. Ann Surg. 2019; 269(4):589–95. 26. Ma B, Gao P, Song Y, Zhang C, Zhang C, Wang L, et al. Transanal total mesorectal excision (taTME) for rectal cancer: a systematic review and meta- analysis of oncological and perioperative outcomes compared with laparoscopic total mesorectal excision. BMC Cancer. 2016;16:380. 27. Wu Z, Zhou W, Chen F, Wang W, Feng Y. Short-term outcomes of transanal versus laparoscopic total mesorectal excision: a systematic review and meta-analysis of cohort studies. J Cancer. 2019;10(2):341–54. 28. Denost Q, Adam JP, Rullier A, Buscail E, Laurent C, Rullier E. Perineal transanal approach: a new standard for laparoscopic sphincter-saving resection in low rectal cancer, a randomized trial. Ann Surg. 2014;260(6): 993–9. 29. Tuech JJ, Karoui M, Lelong B, De Chaisemartin C, Bridoux V, Manceau G, et al. A step toward NOTES total mesorectal excision for rectal cancer: endoscopic transanal proctectomy. Ann Surg. 2015;261(2):228–33. 30. Lelong B, Meillat H, Zemmour C, Poizat F, Ewald J, Mege D, et al. Short- and mid-term outcomes after endoscopic transanal or laparoscopic transabdominal total mesorectal excision for low rectal cancer: a single institutional case-control study. J Am Coll Surg. 2017;224(5):917–25. 31. Hol JC, van Oostendorp SE, Tuynman JB, Sietses C. Long-term oncological results after transanal total mesorectal excision for rectal carcinoma. Tech Coloproctol. 2019;23:903. 32. Torres F, Rios J, Saez-Penataro J, Pontes C. Is propensity score analysis a valid surrogate of randomization for the avoidance of allocation bias? Semin Liver Dis. 2017;37(3):275–86. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.