Short‐Term Outcomes of Transanal Total Mesorectal Excision for Rectal Cancer: A Multicenter Propensity Score‐Matched Analysis Comparing Patients With and Without Preoperative Neoadjuvant Therapy Authors: Takafumi Shima, Hiroki Hamamoto, Takeru Matsuda, Ichiro Takemasa, Keitaro Tanaka, Toshiki Mukai, Jun Watanabe, Sang‐Woong Lee, Seiichiro Yamamoto, Takeshi Naitoh, Akihiro Kondo, Akihiro Kondo, Akinobu Taketomi, Akiyoshi Kanazawa, Atsushi Ogura, Chikayoshi Tani, Gunpei Yoshimatsu, Hideki Endo, Hideya Kashihara, Hirokazu Suwa, Junichiro Kawamura, Kay Uehara, Kei Kimura, Keiichi Okano, Keiichiro Okuyama, Kenji Okita, Kiyoshi Maeda, Kohei Shigeta, Koji Okabayashi, Koya Hida, Masafumi Inomata, Masahiko Watanabe, Masaji Tani, Masataka Ikeda, Masatsune Shibutani, Masayoshi Iwamoto, Masayuki Ishii, Nobuki Ichikawa, Shigenori Honma, Shinichiro Mori Published: 2026 ABSTRACT ABSTRACTAimTo evaluate the association between preoperative neoadjuvant therapy and short‐term outcomes of transanal total mesorectal excision for rectal cancer.MethodsWe retrospectively analyzed patients who underwent transanal total mesorectal excision at 26 Japanese institutions between 2012 and 2019. Patients who received neoadjuvant therapy were matched 1:1 with patients who did not. Perioperative, postoperative, and pathological outcomes were compared.ResultsAfter matching, 294 patients (147 per group) were analyzed. Operative time was longer in the neoadjuvant therapy group than in the non‐neoadjuvant therapy group (median, 510.0 vs. 382.0 min), with an adjusted mean difference of 129.25 min (95% CI, 87.90–170.59; p < 0.001). Clavien–Dindo grade ≥ III complications occurred in 18/147 patients (12.2%) without neoadjuvant therapy and 16/147 (10.9%) with neoadjuvant therapy (adjusted OR, 0.90; 95% CI, 0.44–1.85; p = 0.769). Full‐cohort conventional and Firth logistic analyses and marginal standardization likewise did not identify a statistically significant association. Neoadjuvant therapy remained associated with longer operative time after adjustment for operative procedure, lateral pelvic lymph node dissection, and abdominal approach. Resection‐margin involvement did not differ significantly between groups.ConclusionNeoadjuvant therapy was associated with longer operative time during transanal total mesorectal excision, whereas no statistically significant association with severe postoperative morbidity was detected. Sensitivity analyses likewise did not identify a statistically significant association, supporting the technical feasibility of transanal total mesorectal excision after neoadjuvant therapy in experienced centers. Abbreviations abdominoperineal resection American Society of Anesthesiologists physical status body mass index carbohydrate antigen 19–9 Clavien–Dindo carcinoembryonic antigen confidence interval chemoradiotherapy interquartile range intersphincteric resection low anterior resection lateral pelvic lymph node dissection lower urinary tract dysfunction neoadjuvant chemotherapy neoadjuvant therapy odds ratio pathological complete response propensity score standardized mean difference surgical‐site infection transanal total mesorectal excision total mesorectal excision total neoadjuvant therapy total pelvic exenteration Introduction Total mesorectal excision (TME) is crucial for achieving local control of tumors in the mid‐ and lower rectum, and transanal TME (TaTME) was developed to overcome the technical challenges of laparoscopic pelvic dissection for rectal cancer, particularly in patients who are male or obese or who have a narrow pelvis or mid‐ to low‐rectal tumors [1, 2, 3]. Its improved visualization and maneuverability may facilitate precise dissection in the deep pelvis. The treatment options for locally advanced rectal cancer include neoadjuvant chemotherapy (NAC) and neoadjuvant chemoradiotherapy (NACRT); in recent years, total neoadjuvant therapy (TNT) has been adopted to increase the pathological complete response (pCR) rate and improve systemic disease control [4, 5, 6]. As the use of these neoadjuvant treatment modalities expands, understanding their influence on the technical feasibility, perioperative safety, and pathological outcomes of TaTME has become increasingly important. A network meta‐analysis of 29 randomized controlled trials involving 6237 patients found broadly comparable perioperative and oncological outcomes among open, laparoscopic, robotic, and transanal approaches. Although laparoscopic surgery was associated with a higher risk of circumferential resection‐margin involvement than transanal surgery, this analysis compared surgical approaches and did not specifically evaluate the influence of NAT on TaTME outcomes [7]. In addition, a retrospective comparative study reported comparable short‐term outcomes between TaTME and robotic TME in patients with low rectal cancer following neoadjuvant chemoradiotherapy. However, its retrospective design and single‐centre setting limit the generalizability of those findings [8]. Nevertheless, evidence regarding the effects of NAT on the short‐term outcomes of TaTME remains limited, particularly in multicentre populations. In the parent nationwide Ta‐Ta‐Mi study, 706 patients who underwent TaTME at 26 Japanese institutions were evaluated. The 3‐year cumulative local recurrence rate was 6.82%, and oncological outcomes were considered acceptable [9]. However, that analysis focused primarily on local recurrence and did not specifically assess the influence of NAT on short‐term surgical outcomes. Therefore, in this study, we evaluated the impact of NAT on the short‐term perioperative, pathological, and complication outcomes of TaTME for rectal cancer using a propensity score‐matched cohort derived from 26 institutions. Patients and Methods Study Design and Reporting This retrospective multicenter cohort study was conducted in accordance with the STROBE statement [10] for observational research and was conducted across 26 institutions affiliated with the Japan Society of Laparoscopic Colorectal Surgery. Patients diagnosed with primary rectal cancer who underwent TaTME at one of the participating institutions between January 2012 and December 2019 were included. The inclusion criteria were histologically diagnosed adenocarcinoma or adenosquamous carcinoma with the distal edge of the tumor located within 12 cm of the anal verge. Patients with recurrent disease, multiple cancers, or ulcerative colitis and those who underwent concomitant surgery for another disease were excluded. Hartmann's procedure or total pelvic exenteration were excluded a priori (Figure 1). The choice of NAT regimen was not standardized by the study protocol and was determined according to clinical practice at each participating institution; detailed institution‐specific eligibility criteria and the interval between completion of NAT and surgery were not prospectively collected. Similarly, criteria for no, unilateral, or bilateral LLND were not standardized across institutions. Tumor height was collected in the parent Ta‐Ta‐Mi study, in which the median distance from the anal verge was 4 cm, and has been previously reported [9], whereas systematic MRI‐based assessments of mesorectal fascia involvement and extramural venous invasion were not included in the present database. Detailed information regarding sphincter or levator ani muscle invasion, the preoperatively planned surgical procedure, and institution‐specific eligibility criteria for ISR or other sphincter‐preserving procedures was not prospectively collected. CRM involvement was defined as a tumor‐to‐circumferential resection margin distance of ≤ 1 mm, and a distance > 1 mm was classified as CRM negative. Pathological assessment was performed locally at each institution without centralized pathological review, and detailed specimen‐processing and CRM measurement procedures were not standardized by the study protocol. Study flow diagram. NAT, neoadjuvant therapy; TPE, total pelvic exenteration. The study was approved by the Institutional Review Board and Ethics Committee of Kobe University Graduate School of Medicine (IRB reference code: B220092). The need for informed consent was waived in view of the anonymity of the study data. However, in accordance with the Japanese ethical regulations, Japanese Personal Information Protection Law, and instructions of the Ethics Committees of each institution, consent was secured via the opt‐out route. Information regarding the purpose and methodology of the study was disclosed on the hospital noticeboard and/or website and opportunities to opt out were provided to the extent possible. Endpoints The pre‐specified primary endpoint was postoperative complications of Clavien–Dindo (CD) grade ≥ III, classified according to the original CD classification [11]. Secondary endpoints included (i) CD grade ≥ II complications; (ii) specific complications—anastomotic leakage, postoperative ileus, postoperative bleeding, lower urinary tract dysfunction (LUTD), and surgical‐site infection (SSI); (iii) operative time, intraoperative blood loss, number of harvested lymph nodes, and length of hospital stay; and (iv) pathological findings (resection margins, pT, pN, pM, lymphatic invasion, vascular invasion, perineural invasion, tumor budding, and lateral pelvic lymph‐node metastasis). Statistical Analysis Categorical variables are presented as numbers and percentages and were compared using the chi‐square test or Fisher's exact test, as appropriate. Continuous variables are presented as median and interquartile range (IQR) and were compared using the Mann–Whitney U test. To reduce treatment selection bias and potential confounding, propensity score (PS) matching was performed using the MatchIt package [12, 13]. Patients in the neoadjuvant therapy group were matched 1:1 to those in the non‐neoadjuvant therapy group using nearest‐neighbor matching without replacement with a caliper width of 0.2 standard deviations of the logit of the propensity score [12]. Age, sex, body mass index, American Society of Anesthesiologists physical status, carcinoembryonic antigen (CEA), carbohydrate antigen 19–9 (CA19‐9), previous abdominal surgery, and clinical tumor stage (cT, cN, and cM; TNM classification of the 8th edition of the American Joint Committee on Cancer) were included in the propensity score model. Exact matching was imposed on categorized age (≥ 70 vs. < 70 years), BMI (≥ 25 vs. < 25 kg/m2), ASA‐PS (1–2 vs. 3–4), CEA (≥ 5 vs. < 5 ng/mL), CA19‐9 (≥ 37 vs. < 37 U/mL), cT (0–3 vs. 4), and cN. After matching, covariate balance was assessed using standardized mean differences (SMDs); an SMD < 0.1 was considered indicative of good balance. After matching, continuous BMI showed only minor residual imbalance and was therefore additionally adjusted for in the matched outcome models. Continuous outcomes were analyzed using weighted linear regression and binary outcomes using weighted logistic regression with cluster‐robust standard errors by matched subclass. For sparse binary outcomes, absolute risk differences were additionally estimated using weighted linear probability models with matched‐subclass cluster‐robust standard errors; risk differences are reported as NAT minus non‐NAT in percentage points. For the primary endpoint, sensitivity analyses included conventional multivariable logistic regression in the complete‐case full cohort with institution‐cluster robust standard errors, Firth logistic regression, and marginal standardization. Additional analyses evaluated NAT according to specific regimen (NAC, CRT, TNT, and radiotherapy alone), a post‐matching operative‐time model adjusted for BMI, operative procedure, LLND, and abdominal approach, a no‐LLND subgroup analysis, and NAT × LLND interaction testing. Anastomotic leakage was additionally analyzed among patients undergoing LAR or ISR, with adjustment for BMI and diverting stoma. All tests were two‐sided, and p < 0.05 was considered statistically significant. Results Patient Characteristics A total of 679 patients with rectal cancer who underwent TaTME at the 26 participating institutions were eligible for inclusion. Complete propensity‐score covariate data were available for 676 patients. After 1:1 propensity score matching, 294 patients (147 per group) were included in the matched analysis (Table 1). After matching, good balance was achieved for the prespecified covariates, except for minor residual imbalance in continuous BMI (absolute SMD = 0.113; Table 1). Continuous BMI was therefore additionally adjusted for in the matched outcome models. TABLE 1Patient characteristics.FactorCategoryBefore matchingAfter matchingNon‐neoadjuvant therapies group (N = 375)Neoadjuvant therapies group (N = 304)SMDNon‐neoadjuvant therapies group (N = 147)Neoadjuvant therapies group (N = 147)SMD n (%) n (%) n (%) n (%)Age (years)< 70225 (60.0)218 (71.7)−0.260104 (70.7)104 (70.7)0.097≥ 70150 (40.0)86 (28.3)43 (29.3)43 (29.3)SexMale259 (69.1)211 (69.4)−0.007103 (70.1)101 (68.7)0.030Female116 (30.9)93 (30.6)44 (29.9)46 (31.3)BMI (kg/m2)< 25268 (71.5)221 (72.7)−0.028109 (74.1)109 (74.1)0.113≥ 25107 (28.5)83 (27.3)38 (25.9)38 (25.9)ASA‐PSI, II340 (90.7)289 (95.1)−0.203145 (98.6)145 (98.6)0.000III, IV35 (9.3)15 (4.9)2 (1.4)2 (1.4)CEA (ng/mL)< 5273 (72.8)162 (53.3)0.39199 (67.3)99 (67.3)0.012≥ 5102 (27.2)142 (46.7)48 (32.7)48 (32.7)CA19‐9 (U/mL)< 37345 (92.0)262 (86.2)0.169141 (95.9)141 (95.9)0.010≥ 3730 (8.0)42 (13.8)6 (4.1)6 (4.1)Previous abdominal surgery−303 (80.8)238 (78.3)0.061115 (78.2)115 (78.2)0.000+72 (19.2)66 (21.7)32 (21.8)32 (21.8)cT0–3333 (89.0)221 (73.2)0.414125 (85.0)125 (85.0)0.000441 (11.0)81 (26.8)22 (15.0)22 (15.0)cN0262 (69.9)106 (34.9)0.73476 (51.7)76 (51.7)0.000≥ 1113 (30.1)198 (65.1)71 (48.3)71 (48.3)cM0359 (95.7)254 (83.6)0.329135 (91.8)135 (91.8)0.000116 (4.3)50 (16.4)12 (8.2)12 (8.2) Note: Clinical T category was unavailable in one non‐NAT patient and two NAT patients before matching; these patients were excluded from the propensity‐score analysis. Age, BMI, CEA, and CA19‐9 are presented as categories, their SMDs were calculated using continuous values.Abbreviations: ASA‐PS, American Society of Anesthesiologists physical status; BMI, body mass index; CA19‐9, carbohydrate antigen 19–9; CEA, carcinoembryonic antigen; NAT, neoadjuvant therapy; SMD, standardized mean difference. Patient characteristics. Note: Clinical T category was unavailable in one non‐NAT patient and two NAT patients before matching; these patients were excluded from the propensity‐score analysis. Age, BMI, CEA, and CA19‐9 are presented as categories, their SMDs were calculated using continuous values. Abbreviations: ASA‐PS, American Society of Anesthesiologists physical status; BMI, body mass index; CA19‐9, carbohydrate antigen 19–9; CEA, carcinoembryonic antigen; NAT, neoadjuvant therapy; SMD, standardized mean difference. Perioperative Outcomes Perioperative outcomes are shown in Tables 2 and 3. After matching, the overall distribution of operative procedure no longer differed significantly between groups, and diverting stoma use was also similar (94/147 (63.9%) in the non‐NAT group vs. 102/147 (69.4%) in the NAT group; p = 0.387). Differences in LLND and abdominal approach remained clinically relevant and were therefore examined in additional sensitivity analyses rather than being incorporated into the propensity score, because these procedural variables may occur downstream of NAT and treatment response. TABLE 2Perioperative surgical procedures after propensity score matching.VariableCategoryNon‐NAT (N = 147), n (%)NAT (N = 147), n (%) p Operative procedureLAR38 (25.9%)31 (21.1%)0.314ISR73 (49.7%)86 (58.5%)APR36 (24.5%)30 (20.4%)Abdominal approachLaparoscopic127 (86.4%)91 (61.9%)< 0.001Robotic19 (12.9%)49 (33.3%)Open1 (0.7%)7 (4.8%)Diverting stoma−53 (36.1%)45 (30.6%)0.387+94 (63.9%)102 (69.4%)Lymph node dissectionD214 (9.5%)4 (2.7%)0.029D3133 (90.5%)143 (97.3%)LLND−105 (71.4%)67 (45.6%)< 0.001Unilateral11 (7.5%)16 (10.9%)Bilateral31 (21.1%)64 (43.5%)Autonomic nerve resection−141 (95.9%)133 (90.5%)0.079Unilateral5 (3.4%)7 (4.8%)Bilateral1 (0.7%)7 (4.8%)Combined resection of other organs−138 (93.9%)135 (91.8%)0.651+9 (6.1%)12 (8.2%)Transanal approach teamsOne‐team34 (23.1%)52 (35.4%)0.029Two‐team113 (76.9%)95 (64.6%)Transanal approach gradeI34 (23.1%)35 (23.8%)0.651II101 (68.7%)104 (70.7%)III12 (8.2%)8 (5.4%) Note: p values compare the distribution of each procedural characteristic between matched groups; Pearson chi‐square tests were used for multi‐category variables and continuity‐corrected/fisher‐type 2 × 2 comparisons for binary variables.Abbreviations: APR, abdominoperineal resection; ISR, intersphincteric resection; LAR, low anterior resection; LLND, lateral pelvic lymph node dissection; NAT, neoadjuvant therapy. Perioperative surgical procedures after propensity score matching. Note: p values compare the distribution of each procedural characteristic between matched groups; Pearson chi‐square tests were used for multi‐category variables and continuity‐corrected/fisher‐type 2 × 2 comparisons for binary variables. Abbreviations: APR, abdominoperineal resection; ISR, intersphincteric resection; LAR, low anterior resection; LLND, lateral pelvic lymph node dissection; NAT, neoadjuvant therapy. TABLE 3Perioperative quantitative outcomes after propensity score matching.OutcomeNon‐NAT (N = 147) median (IQR)NAT (N = 147) median (IQR)BMI‐adjusted mean difference95% CI p Operative time (min)382.0 (289.5–474.0)510.0 [374.5–667.0]129.2587.90 to 170.59< 0.001Intraoperative blood loss (g)50 (20–120)40 [10–112]25.44−44.99 to 95.880.479Number of lymph nodes dissected17 (12–29)22 [13–38]3.09−0.26 to 6.440.071Length of hospital stay (days)17 (14–23)18 [14–24]−0.11−3.90 to 3.680.954Abbreviations: CI, confidence interval; IQR, interquartile range; NAT, neoadjuvant therapy. Perioperative quantitative outcomes after propensity score matching. Abbreviations: CI, confidence interval; IQR, interquartile range; NAT, neoadjuvant therapy. Operative time was longer in the NAT group than in the non‐NAT group (median, 510.0 min (IQR, 374.5–667.0) vs. 382.0 min (IQR, 289.5–474.0), respectively), with an adjusted mean difference of 129.25 min (95% CI, 87.90–170.59; p < 0.001) (Table 3). In the subgroup of patients who did not undergo LLND, NAT remained associated with longer operative time than non‐NAT, with an adjusted 13.9% increase in operative time (95% CI, 1.4%–27.9%; p = 0.029; Table 4). There was no significant interaction between NAT and LLND extent (global P for interaction = 0.715; Table 5). In a post‐matching sensitivity model additionally adjusting for BMI, operative procedure, LLND, and abdominal approach, NAT remained associated with a 14.0% longer operative time (95% CI, 5.3%–23.4%; p = 0.001; Table S1). TABLE 4Operative time in patients without lateral lymph node dissection.OutcomeNon‐NAT (N = 105)NAT (N = 67)Adjusted % change95% CI p Operative time, min, median (IQR)335 (266–422)382 (296–491)13.91.4 to 27.90.029 Note: Estimates were derived from an analysis restricted to patients who did not undergo LLND.Abbreviations: CI, confidence interval; IQR, interquartile range; NAT, neoadjuvant therapy. Operative time in patients without lateral lymph node dissection. Note: Estimates were derived from an analysis restricted to patients who did not undergo LLND. Abbreviations: CI, confidence interval; IQR, interquartile range; NAT, neoadjuvant therapy. TABLE 5NAT × LLND interaction analysis.TermComparison/interpretationAdjusted % change95% CI p RoleNATNAT vs. Non‐NAT within No‐LLND14.21.6–28.40.027Main effectUnilateral LLNDUnilateral LLND vs. No LLND within Non‐NAT35.515.9–58.30.000Main effectBilateral LLNDBilateral LLND vs. No LLND within Non‐NAT52.537.5–69.30.000Main effectNAT × unilateral LLNDModification of NAT association with unilateral LLND4.3−13.6–26.10.660InteractionNAT × bilateral LLNDModification of NAT association with bilateral LLND6.6−8.6–24.40.417InteractionGlobal NAT × LLND interaction test p for interaction = 0.715 Note: Estimates were derived from the interaction model including NAT, LLND extent, and NAT × LLND interaction terms.Abbreviations: CI, confidence interval; IQR, interquartile range; LLND, lateral pelvic lymph node dissection; NAT, neoadjuvant therapy. NAT × LLND interaction analysis. Note: Estimates were derived from the interaction model including NAT, LLND extent, and NAT × LLND interaction terms. Abbreviations: CI, confidence interval; IQR, interquartile range; LLND, lateral pelvic lymph node dissection; NAT, neoadjuvant therapy. For the primary endpoint, CD grade ≥ III complications occurred in 18 of 147 patients (12.2%) in the non‐NAT group and 16 of 147 patients (10.9%) in the NAT group (BMI‐adjusted matched‐pair OR, 0.90; 95% CI, 0.44–1.85; p = 0.769) (Table 6). Among patients undergoing LAR or ISR, anastomotic leakage occurred in 13 of 111 patients (11.7%) in the non‐NAT group and 4 of 117 patients (3.4%) in the NAT group. In the complete‐case full cohort (n = 676), the adjusted OR was 1.16 (95% CI, 0.67–2.03; p = 0.595) using conventional multivariable logistic regression with institution‐cluster robust standard errors and 1.16 (95% CI, 0.70–1.94; p = 0.562) using Firth logistic regression. Marginal standardization yielded adjusted risks of 11.5% for non‐NAT and 13.1% for NAT, corresponding to an adjusted risk difference of +1.59 percentage points (95% CI, −4.46 to +7.64; p = 0.606; Table S2). TABLE 6Postoperative complication outcomes after propensity score matching.OutcomeNon‐NAT, n/N (%)NAT, n/N (%)MeasureAdjusted estimate (95% CI) p CD grade ≥ II complications54/147 (36.7%)48/147 (32.7%)OR0.86 (0.52–1.44)0.571CD grade ≥ III complications18/147 (12.2%)16/147 (10.9%)OR0.90 (0.44–1.85)0.769Anastomotic leakage a 13/111 (11.7%)4/117 (3.4%)OR0.28 (0.09–0.88)0.029Ileus12/147 (8.2%)12/147 (8.2%)OR1.08 (0.45–2.56)0.866Postoperative bleeding1/147 (0.7%)1/147 (0.7%)RD, pp−0.01 (−1.88 to 1.86)0.993LUTD7/147 (4.8%)7/147 (4.8%)OR0.99 (0.32–3.09)0.991Surgical site infection7/147 (4.8%)6/147 (4.1%)RD, pp−0.90 (−5.78 to 3.99)0.719 Note: For sparse outcomes, adjusted risk differences were estimated using weighted linear probability models with matched‐subclass cluster‐robust standard errors. RD is expressed as NAT minus non‐NAT in percentage points.Abbreviations: CD, Clavien–Dindo; CI, confidence interval; ISR, intersphincteric resection; LAR, low anterior resection; LUTD, lower urinary tract dysfunction; NAT, neoadjuvant therapy; OR, odds ratio; pp., percentage points; RD, risk difference. a Anastomotic leakage was analyzed only among patients undergoing LAR or ISR and was adjusted for BMI and diverting stoma. Postoperative complication outcomes after propensity score matching. Note: For sparse outcomes, adjusted risk differences were estimated using weighted linear probability models with matched‐subclass cluster‐robust standard errors. RD is expressed as NAT minus non‐NAT in percentage points. Abbreviations: CD, Clavien–Dindo; CI, confidence interval; ISR, intersphincteric resection; LAR, low anterior resection; LUTD, lower urinary tract dysfunction; NAT, neoadjuvant therapy; OR, odds ratio; pp., percentage points; RD, risk difference. Anastomotic leakage was analyzed only among patients undergoing LAR or ISR and was adjusted for BMI and diverting stoma. In the exploratory multivariable analysis of the complete‐case full cohort (n = 676), NAT remained associated with longer operative time after adjustment for demographic, clinicopathological, and operative variables, corresponding to an 8.3% increase in operative time (95% CI, 2.3%–14.7%; p = 0.006). Male sex, higher BMI, previous abdominal surgery, cT2–4 disease compared with cT0, LLND, APR, and an open surgical approach were also associated with longer operative time, whereas the two‐team transanal approach was associated with shorter operative time (Table 7). TABLE 7Main multivariable analysis of factors associated with operative time.VariableComparison/codingApproximate change in operative time (%)95% CI p Neoadjuvant therapyNon‐NATReference——NAT+8.32.3–14.70.006SexMaleReference——Female−7.4−12.0 to −2.70.003BMIPer 1‐kg/m2 increase+1.30.7–1.9< 0.001ASA‐PSI, IIReference——III, IV−6.4−14.4 to 2.40.147CEAlog (x + 1)+0.8−1.9 to 3.50.578CA19‐9log (x + 1)+2.3−0.1 to 4.60.058Previous abdominal surgery−Reference——++6.00.3–12.00.038cT0Reference——1+12.9−2.9 to 31.40.1152+16.81.2–34.90.0343+21.15.9–38.50.0054+29.010.8–50.20.001cN−Reference——++1.5−4.0 to 7.40.599cM−Reference——++5.0−3.1 to 13.80.236LLND−Reference——Unilateral+24.314.8–34.7< 0.001Bilateral+54.645.7–64.1< 0.001Operative procedureLARReference——ISR+2.7−3.4 to 9.30.392APR+7.91.4–14.80.017Surgical approachOpenReference——Minimally invasive surgery a −16.6−35.8 to −0.10.049Transanal approach teamsOne‐teamReference——Two‐team−17.6−21.6 to −13.4< 0.001 a Abbreviations: APR, abdominoperineal resection; ASA‐PS, American Society of Anesthesiologists physical status; BMI, body mass index; CA19‐9, carbohydrate antigen 19–9; CEA, carcinoembryonic antigen; CI, confidence interval; ISR, intersphincteric resection; LAR, low anterior resection; LLND, lateral lymph node dissection; NAT, neoadjuvant therapy. a Minimally invasive surgery includes laparoscopic and robotic approaches. Main multivariable analysis of factors associated with operative time. Abbreviations: APR, abdominoperineal resection; ASA‐PS, American Society of Anesthesiologists physical status; BMI, body mass index; CA19‐9, carbohydrate antigen 19–9; CEA, carcinoembryonic antigen; CI, confidence interval; ISR, intersphincteric resection; LAR, low anterior resection; LLND, lateral lymph node dissection; NAT, neoadjuvant therapy. Minimally invasive surgery includes laparoscopic and robotic approaches. In the regimen‐specific analysis, operative time, blood loss, and postoperative complications were compared among the NAC, CRT, TNT, and radiotherapy‐alone groups. CD grade ≥ III complications occurred in 20/180 (11.1%), 17/102 (16.7%), 3/12 (25.0%), and 2/8 (25.0%) patients receiving NAC, CRT, TNT, and radiotherapy alone, respectively, without a statistically significant overall difference (p = 0.164; Table S3). Operative time was longest in the radiotherapy‐alone group. However, when NAC was used as the reference in the adjusted analysis, operative time did not differ significantly among CRT, TNT, and radiotherapy alone (Table S4). Pathological Findings Pathological findings are summarized in Table 8. Distal margin positivity was rare, occurring in 1 of 146 evaluable patients (0.7%) in the non‐NAT group and 1 of 147 patients (0.7%) in the NAT group. The NAT group had significantly lower frequencies of lymphatic invasion (23.8% vs. 43.5%; adjusted OR, 0.41; 95% CI, 0.24–0.69; p < 0.001) and venous invasion (38.8% vs. 64.6%; adjusted OR, 0.34; 95% CI, 0.20–0.56; p < 0.001) than the non‐NAT group. In contrast, no significant between‐group differences were observed in p/ypT category (p = 0.607), p/ypN status (p = 0.848), perineural invasion (35/147 (23.8%) vs. 27/147 (18.4%); adjusted OR, 0.67; 95% CI, 0.38–1.17; p = 0.159), or tumor differentiation (differentiated histology: 138/147 (93.9%) vs. 130/147 (88.4%); adjusted OR, 1.95; 95% CI, 0.83–4.58; p = 0.127). Lateral pelvic lymph‐node metastasis was numerically more frequent in the NAT group (16/147 (10.9%) vs. 8/147 (5.4%); adjusted OR, 2.06; 95% CI, 0.92–4.63; p = 0.079), although the difference did not reach statistical significance. TABLE 8Pathological findings after propensity score matching.VariableCategoryNon‐NAT (N = 147), n (%)NAT (N = 147), n (%)Adjusted effect95% CI p Distal marginNegative145 (98.6%)146 (99.3%)RD +0.07 pp−1.64 to 1.780.935Positive1 (0.7%)1 (0.7%)Unknown1 (0.7%)0 (0.0%)CRMNegative (> 1 mm)130 (88.4%)132 (89.8%)OR 0.710.28 to 1.810.472Positive (≤ 1 mm)11 (7.5%)8 (5.4%)Unknown6 (4.1%)7 (4.8%)LPN metastasisNegative139 (94.6%)131 (89.1%)OR 2.060.92 to 4.630.079Positive8 (5.4%)16 (10.9%)Lymphatic invasionNegative83 (56.5%)112 (76.2%)OR 0.410.24 to 0.69< 0.001Positive64 (43.5%)35 (23.8%)Venous invasionNegative52 (35.4%)90 (61.2%)OR 0.340.20 to 0.56< 0.001Positive95 (64.6%)57 (38.8%)Perineural invasionNegative93 (63.3%)108 (73.5%)OR 0.670.38 to 1.170.159Positive35 (23.8%)27 (18.4%)Unknown19 (12.9%)12 (8.2%)BuddingBD158 (39.5%)74 (50.3%)——0.006BD210 (6.8%)8 (5.4%)BD310 (6.8%)0 (0.0%)Unknown69 (46.9%)65 (44.2%)p/ypT0–2/Tis72 (49.0%)75 (51.0%)OR 1.110.75 to 1.650.6073–475 (51.0%)72 (49.0%)p/ypNNegative98 (66.7%)100 (68.0%)OR 0.950.58 to 1.560.848Positive49 (33.3%)47 (32.0%)Pathological MNegative136 (92.5%)140 (95.2%)RD −3.05 pp−6.38 to 0.280.074Positive11 (7.5%)7 (4.8%)Tumor differentiationDifferentiated138 (93.9%)130 (88.4%)OR 1.950.83 to 4.580.127Others9 (6.1%)17 (11.6%) Note: For sparse outcomes, adjusted risk differences were estimated using weighted linear probability models with matched‐subclass cluster‐robust standard errors. RD is expressed as NAT minus non‐NAT in percentage points. Percentages are based on the total matched cohort; effect estimates for distal margin and CRM were calculated among patients with evaluable margin data.Abbreviations: BD, tumor budding grade; CI, confidence interval; CRM, circumferential resection margin; LPN, lateral pelvic lymph node; NAT, neoadjuvant therapy; OR, odds ratio; RD, risk difference. Pathological findings after propensity score matching. Note: For sparse outcomes, adjusted risk differences were estimated using weighted linear probability models with matched‐subclass cluster‐robust standard errors. RD is expressed as NAT minus non‐NAT in percentage points. Percentages are based on the total matched cohort; effect estimates for distal margin and CRM were calculated among patients with evaluable margin data. Abbreviations: BD, tumor budding grade; CI, confidence interval; CRM, circumferential resection margin; LPN, lateral pelvic lymph node; NAT, neoadjuvant therapy; OR, odds ratio; RD, risk difference. Discussion This propensity‐score‐matched analysis investigated the impact of NAT on short‐term outcomes following TaTME for rectal cancer. The principal finding was that NAT was associated with longer operative time. No statistically significant association between NAT and CD grade ≥ III postoperative complications was detected in the matched analysis. Sensitivity analyses using full‐cohort conventional logistic regression, Firth logistic regression, and marginal standardization likewise did not identify a statistically significant association between NAT and severe postoperative complications. However, effect estimates varied across analyses, and the confidence intervals remained wide; therefore, clinically relevant benefit or harm could not be excluded. These results should not be interpreted as demonstrating equivalence or the complete absence of an effect. With regard to morbidity, these results align with recent evidence suggesting that NAT, including chemotherapy, chemoradiation, and TNT, does not increase the incidence of serious complications following rectal‐cancer surgery in general [5, 6, 14, 15, 16, 17, 18, 19]. The COLOR III trial reported comparable short‐term morbidity between TaTME and laparoscopic TME, although the trial did not specifically address NAT [20]. Similarly, the Liu et al. randomised comparison of TaTME versus laparoscopic TME likewise documented similar severe morbidity in both arms, again without directly randomising NAT exposure [21]. A 2019 network meta‐analysis pooling open, laparoscopic, robotic, and transanal mesorectal‐excision trials found no significant differences in intra‐operative morbidity, conversion rate, or Grade III/IV complications across approaches [7]. A recent systematic review of NAC alone similarly found no excess in postoperative morbidity compared with NACRT [22]. The overall CD grade ≥ III complication rate in the matched cohort was 11.6% (34/294), which falls within the 10%–15% range reported in previous multicentre trials and supports the external consistency of the present findings. The longer operative time observed after NAT is compatible with increased operative complexity related to treatment‐associated tissue edema, fibrosis or less distinct anatomical planes, but it may also reflect differences in tumor characteristics, LLND, operative procedure, abdominal approach, institutional practice, and surgeon experience [15, 23, 24]. Importantly, the association persisted in the post‐matching sensitivity model adjusting for BMI, operative procedure, LLND, and abdominal approach. LLND itself was strongly associated with longer surgery, but NAT remained associated with longer operative time even among patients who did not undergo LLND, and no significant NAT × LLND interaction was detected. These findings suggest that LLND and other procedural factors contribute to operative complexity but do not fully explain the observed association between NAT and operative duration. Residual confounding nevertheless remains possible. By contrast, the two‐team transanal approach was independently associated with shorter operative time, consistent with previous reports suggesting greater procedural efficiency with simultaneous abdominal and transanal dissection [25]. Although operative time was longest in the radiotherapy‐alone group in the descriptive analysis, adjusted regimen‐specific comparisons did not identify significant differences among NAT regimens. The indications for NAC, CRT, and TNT were not standardized by the study protocol; detailed institution‐specific selection criteria were not collected, and the interval between completion of NAT and surgery was unavailable. Accordingly, regimen‐specific findings should be considered exploratory. The pathological findings showed significantly lower rates of lymphatic and venous invasion in the NAT group than in the non‐NAT group (lymphatic invasion, 23.8% vs. 43.5%; venous invasion, 38.8% vs. 64.6%; both p < 0.001), consistent with the known treatment‐related pathological response to NAT reported in randomized trials and meta‐analyses [14, 15, 16, 17, 18, 19]. In contrast, no significant between‐group differences were observed in the grouped p/ypT or p/ypN categories. TNT, in particular, has been associated with higher pathological complete response rates [16, 17], although regimen‐specific pathological effects could not be definitively assessed in the present study. Importantly, no significant difference in resection‐margin involvement was detected. Distal margin positivity was observed in 1 of 146 evaluable patients (0.7%) in the non‐NAT group and 1 of 147 patients (0.7%) in the NAT group (adjusted risk difference, +0.07 percentage points; 95% CI, −1.64 to 1.78; p = 0.935). CRM involvement, defined as a margin distance of ≤ 1 mm, was observed in 11 of 141 evaluable patients (7.8%) in the non‐NAT group and 8 of 140 patients (5.7%) in the NAT group (adjusted OR, 0.71; 95% CI, 0.28–1.81; p = 0.472). These findings are clinically relevant because fibrosis and tissue edema after NAT may obscure the dissection plane and potentially increase the risk of an inadequate resection margin. Nevertheless, in this multicentre cohort treated at experienced centres, TaTME after NAT was not associated with an increased rate of short‐term pathological margin involvement. An unexpected finding was that anastomotic leakage was less frequent in the NAT group than in the non‐NAT group among patients undergoing LAR or ISR (3.4% vs. 11.7%, respectively). Several factors may explain this difference. Tumor regression after NAT may facilitate distal rectal transection and pelvic dissection in selected patients, potentially reducing technical difficulty during reconstruction. In addition, differences in patient selection, institutional practice, anastomotic technique, perfusion assessment, and other perioperative management factors that were not captured in the database may have contributed to this finding. Although the analysis was restricted to patients with an anastomosis and adjusted for diverting stoma, the numerically lower leakage rate in the NAT group was observed in both patients with and without a diverting stoma. However, the number of leakage events was small, particularly among patients without a diverting stoma, and the timing of surgery after NAT and detailed technical factors related to anastomosis were unavailable. Therefore, this finding should be regarded as exploratory and should not be interpreted as evidence that NAT itself protects against anastomotic leakage. Several limitations should be acknowledged. First, this was a retrospective observational study, and assignment to NAT was non‐randomised. Moreover, the NAT group included only patients who ultimately proceeded to TaTME. Patients who developed disease progression, severe treatment toxicity, or another reason for not undergoing surgery were not captured. The findings therefore apply to selected surgical patients and not to the safety or effectiveness of the complete NAT‐to‐surgery treatment strategy. Second, NAT regimen selection, LLND criteria, and sphincter‐preserving surgical indications were not standardized across participating institutions, and the interval between completion of NAT and surgery was not available. In addition, group‐specific tumor height and MRI‐based assessments of mesorectal fascia involvement and extramural venous invasion were unavailable, limiting our ability to fully account for anatomical tumor characteristics. Third, although propensity score matching substantially improved measured baseline balance, residual and unmeasured confounding remains possible. Fourth, pathological assessment was performed locally without centralized review or a standardized detailed CRM processing and measurement protocol. Fifth, several clinically important outcomes were rare, resulting in limited statistical precision despite the use of risk differences and Firth sensitivity analyses. Finally, all participating centers were experienced Japanese institutions, which may limit generalizability. Conversely, the nationwide multicenter design and the availability of detailed clinical, operative, and pathological data represent important strengths. This study therefore provides clinically relevant evidence regarding the impact of NAT on the short‐term outcomes of TaTME. Conclusions In this multicentre retrospective cohort of selected patients who underwent TaTME, NAT was associated with a longer operative duration. No statistically significant association between NAT and Clavien–Dindo grade ≥ III postoperative complications was detected in either the matched cohort or full‐cohort sensitivity analyses, although clinically relevant benefit or harm cannot be excluded because of the limited precision of the estimates. These findings support the technical feasibility of TaTME after NAT in experienced centres. Author Contributions Takafumi Shima: conceptualization, methodology, investigation, formal analysis, writing – original draft. Hiroki Hamamoto: conceptualization, investigation, methodology, writing – review and editing, formal analysis, data curation. Takeru Matsuda: conceptualization, supervision, project administration, writing – review and editing. Ichiro Takemasa: conceptualization, supervision, project administration, writing – review and editing. Keitaro Tanaka: supervision, writing – review and editing. Toshiki Mukai: supervision, writing – review and editing. Jun Watanabe: supervision, writing – review and editing. Sang‐Woong Lee: supervision, writing – review and editing. Seiichiro Yamamoto: supervision, writing – review and editing. Takeshi Naitoh: supervision, writing – review and editing. Funding The authors have nothing to report. Ethics Statement The study was approved by the Institutional Review Board and Ethics Committee of Kobe University Graduate School of Medicine (IRB reference code: B220092). The need for informed consent was waived in view of the anonymity of the study data. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Table S1: Sensitivity analysis of factors associated with operative time after propensity score matching. Table S2: Sensitivity analyses for Clavien–Dindo grade ≥ III postoperative complications. Table S3: Perioperative outcomes according to neoadjuvant treatment regimen (Regimen‐specific perioperative outcomes). Table S4: Perioperative outcomes according to neoadjuvant treatment regimen (Adjusted differences in operative time (NAC as reference)).