
==== Front
BJS Open
BJS Open
bjsopen
BJS Open
2474-9842
Oxford University Press UK

10.1093/bjsopen/zrae103
zrae103
Original Article
AcademicSubjects/MED00910
Bjs/4
Short-term outcomes after primary total mesorectal excision (TME) versus local excision followed by completion TME for early rectal cancer: population-based propensity-matched study
https://orcid.org/0000-0002-2105-1531
van Lieshout Annabel S Department of Surgery, Amsterdam UMC location Vrije Universiteit Amsterdam, Amsterdam, The Netherlands
Treatment and Quality of Life, Cancer Center Amsterdam, Amsterdam, The Netherlands

https://orcid.org/0000-0001-5889-7939
Smits Lisanne J H Department of Surgery, Amsterdam UMC location Vrije Universiteit Amsterdam, Amsterdam, The Netherlands
Treatment and Quality of Life, Cancer Center Amsterdam, Amsterdam, The Netherlands

https://orcid.org/0000-0002-4914-4537
Sijmons Julie M L Treatment and Quality of Life, Cancer Center Amsterdam, Amsterdam, The Netherlands
Dutch Institute for Clinical Auditing, Leiden, The Netherlands
Department of Surgery, Amsterdam UMC location University of Amsterdam, Amsterdam, The Netherlands

https://orcid.org/0000-0002-3084-0495
van Dieren Susan Treatment and Quality of Life, Cancer Center Amsterdam, Amsterdam, The Netherlands
Department of Surgery, Amsterdam UMC location University of Amsterdam, Amsterdam, The Netherlands

https://orcid.org/0000-0002-6969-7956
van Oostendorp Stefan E Department of Surgery, Rode Kruis Ziekenhuis, Beverwijk, The Netherlands

https://orcid.org/0000-0002-3146-3310
Tanis Pieter J Department of Surgery, Amsterdam UMC location University of Amsterdam, Amsterdam, The Netherlands
Department of Surgical Oncology and Gastrointestinal Surgery, Erasmus MC, Rotterdam, The Netherlands

https://orcid.org/0000-0001-5952-8000
Tuynman Jurriaan B Department of Surgery, Amsterdam UMC location Vrije Universiteit Amsterdam, Amsterdam, The Netherlands

Correspondence to: Annabel S. van Lieshout, Department of Surgery, Amsterdam UMC location Vrije Universiteit Amsterdam, De Boelelaan 1117, Amsterdam, 1081 HV, The Netherlands (e-mail: a.vanlieshout@amsterdamumc.nl)
10 2024
05 9 2024
05 9 2024
8 5 zrae10314 3 2024
27 6 2024
25 7 2024
© The Author(s) 2024. Published by Oxford University Press on behalf of BJS Foundation Ltd.
2024
https://creativecommons.org/licenses/by-nc/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com.

Abstract

Background

Colorectal cancer screening programmes have led to a shift towards early-stage colorectal cancer, which, in selected cases, can be treated using local excision. However, local excision followed by completion total mesorectal excision (two-stage approach) may be associated with less favourable outcomes than primary total mesorectal excision (one-stage approach). The aim of this population study was to determine the distribution of treatment strategies for early rectal cancer in the Netherlands and to compare the short-term outcomes of primary total mesorectal excision with those of local excision followed by completion total mesorectal excision.

Methods

Short-term data for patients with cT1–2 N0xM0 rectal cancer who underwent local excision only, primary total mesorectal excision, or local excision followed by completion total mesorectal excision between 2012 and 2020 in the Netherlands were collected from the Dutch Colorectal Audit. Patients were categorized according to treatment groups and logistic regressions were performed after multiple imputation and propensity score matching. The primary outcome was the end-ostomy rate.

Results

From 2015 to 2020, the proportion for the two-stage approach increased from 22.3% to 43.9%. After matching, 1062 patients were included. The end-ostomy rate was 16.8% for the primary total mesorectal excision group versus 29.6% for the local excision followed by completion total mesorectal excision group (P < 0.001). The primary total mesorectal excision group had a higher re-intervention rate than the local excision followed by completion total mesorectal excision group (16.7% versus 11.8%; P = 0.048). No differences were observed with regard to complications, conversion, diverting ostomies, radical resections, readmissions, and death.

Conclusion

This study shows that, over time, cT1–2 rectal cancer has increasingly been treated using the two-stage approach. However, local excision followed by completion total mesorectal excision seems to be associated with an elevated end-ostomy rate. It is important that clinicians and patients are aware of this risk during shared decision-making.

This population study in the Netherlands explores the treatment strategies for early rectal cancer, comparing the one-stage approach of primary total mesorectal excision with the two-stage approach of local excision followed by completion total mesorectal excision. The data from 2012 to 2020 reveal an increasing trend in the use of the two-stage approach, but local excision followed by completion total mesorectal excision is associated with a higher risk of end ostomy. Clinicians and patients should be mindful of this risk during shared decision-making for colorectal cancer treatment.
==== Body
pmcIntroduction

Since the implementation of the colorectal cancer screening programme in the Netherlands in 2014, an incidence shift towards early-stage cancer has been observed1. This shift has led to more interest in local therapy preserving the organ, avoiding radical surgery. For rectal cancer, the impact of radical surgery in the form of total mesorectal excision (TME) is relatively high, with associated morbidity, poor functional outcomes, and high ostomy rates that contribute to a decrease in quality of life2–4. Therefore, less invasive treatment strategies for early rectal cancer, such as endoscopic or surgical local excision (LE), have gained popularity over the last decade. For pT1 rectal tumours without histopathological risk factors, treatment with LE only is considered sufficient. Risk factors such as poor differentiation, a positive resection margin (R1), lymphovascular invasion, and tumour budding have been shown to be associated with an increased risk of recurrence5–8. Therefore, when these risk factors are documented for the resected specimen, early completion TME (cTME) is recommended by guidelines7,9.

Staging of early rectal cancer using endoscopy, ultrasonography, or MRI is not accurate enough to distinguish low-risk T1 tumours from high-risk T1 and early T2 tumours10–12. A possible approach could be performing an LE to obtain a pathological assessment. Subsequently, for those patients with high-risk lesions, a cTME is performed, the so-called two-stage approach. However, a national study showed that 71% of locally resected high-risk lesions did not undergo further radical treatment, probably due to the presumed complexity of cTME, as well as patient wishes13. Although limited literature is available and small studies have reported conflicting results, cTME may be associated with worse outcomes, including high end-colostomy rates, poor specimen quality, higher risks of rectal perforation, and increased re-intervention rates14–21. These outcomes could potentially cause surgeons to opt for a primary TME (that is the one-stage approach) instead of performing an LE first. Therefore, it is important to investigate whether a rectal-sparing LE is a safe first step in treatment for early rectal cancer, avoiding unnecessary undertreatment and overtreatment2–4. The aim of this population-based study was to determine the distribution of the chosen treatment strategies (one-stage versus two-stage approach) for cT1–2 N0xM0 rectal cancer in the Netherlands between 2012 and 2020 and to compare the short-term outcomes of primary TME with those of LE followed by cTME.

Methods

This study was not pre-registered in an independent, institutional registry.

Patients and outcomes

Patient data for this observational retrospective nationwide cohort study were collected from the national Dutch Colorectal Audit (DCRA), coordinated by the Dutch Institute for Clinical Auditing (DICA). This database contains short-term (less than 90 days) follow-up data for all colorectal cancer patients in the Netherlands who received a surgical intervention. The study included all patients aged greater than or equal to 18 years with cT1–2 N0xM0 rectal carcinoma who underwent a primary TME, a surgical LE, or an LE (endoscopic or surgical) followed by cTME (LE + cTME) between 2012 and 2020. The only exclusion criterion was neoadjuvant treatment. For the first part of the study, outcomes included the distribution of surgical treatment strategies for cT1 and cT2 tumours and the proportion of one-stage and two-stage procedures per year. The primary outcome of the comparison between primary TME and LE + cTME was the end-ostomy rate. Secondary outcomes included postoperative complications within 90 days, conversion rate, diverting ostomy rate, radical (R0) resections (tumour tissue less than or equal to 1 mm from the resection margin), re-interventions, readmissions, and postoperative deaths within 90 days. The postoperative complication pelvic sepsis was defined as anastomotic leakage and/or pelvic abscess formation.

Statistical analysis

Patient characteristics, tumour characteristics, perioperative data, and short-term outcomes are described using descriptive statistics. Categorical data are presented as n (%) and continuous variables are presented as mean (95% c.i.). Univariable analyses were performed using logistic regressions. The results are presented as OR (95% c.i.). To reduce potential confounding, multiple imputation by chained equations of missing variables was performed by making five imputation sets using predictive mean matching. Furthermore, propensity score matching was performed using the Matchthem package22. Propensity score matching was done based on age, distance to the anorectal junction in mm, ASA grade, and cT category using a variable matching ratio, with a maximum of 1 : 323. P < 0.050 was considered statistically significant. All statistical analyses were performed using both SPSS® (IBM, Armonk, NY, USA; version 26) and R (version 4.2.1).

Results

Patient and tumour characteristics

Between 2012 and 2020, a total of 4154 patients received surgical treatment for cT1–2 N0xM0 rectal carcinoma in the Netherlands and were included in this study. Of these patients, 1103 underwent LE only, 2706 underwent primary TME, and 345 underwent LE + cTME. A flow chart of patient selection is presented in Fig. 1. Patient and tumour characteristics of these unmatched groups are presented in Table S1.

Fig. 1 Flow chart of patient selection

DICA, Dutch Institute for Clinical Auditing; TME, total mesorectal excision.

Distribution of chosen treatment per year

Figure 2 shows the number of surgical procedures for cT1–2 rectal cancer per year in the Netherlands. From 2012 until 2016, the total number of surgical procedures for early-stage rectal cancer increased from 193 to 663, with a steep incline in 2014. Subsequently, this number decreased from 663 in 2016 to 424 in 2020, which corresponds to the observed decline in the overall incidence of rectal carcinoma in the Netherlands during these years, shown in Fig. 31.

Fig. 2 Number of surgical procedures for cT1–2 rectal cancer per year in the Netherlands

TME, total mesorectal excision.

Fig. 3 Number of surgical procedures for cT1–2 rectal cancer per year in the Netherlands relative to the national incidence (all stages, stages 2–4, and stage 1)

TME, total mesorectal excision.

The proportions of each treatment group per year are presented in Fig. 4. For primary TME, percentages varied over the years: 64.2% in 2012 compared with 56.1% in 2020. The same applies for LE only: 31.6% in 2012 compared with 28.8% in 2020. Although the primary TME and LE percentages vary (in opposite directions) and with no clear trend over the years, the percentage of patients treated with LE + cTME seems to have increased from 2015 onwards; the percentages were 3.1% and 15.1% in 2015 and 2020 respectively.

Fig. 4 Distribution of different surgical procedures for cT1–2 rectal cancer over time: primary TME versus local excision followed by completion TME versus surgical local excision only

TME, total mesorectal excision.

In addition, when separating groups into one-stage (primary TME) and two-stage (surgical LE only and LE + cTME) approaches, a decreasing trend regarding the one-stage approach can be observed from 2015 onwards (77.7% in 2015 and 56.1% in 2020), as well as an increase in the two-stage approach (22.3% in 2015 and 43.9% in 2020) (Fig. 5). In particular, from 2018 to 2020, the percentages for both approaches are close to 50% (Fig. 5). Similar charts for cT1 and cT2 tumours are presented in Figure S1 and Figure S2 respectively. From 2015 to 2020, an increase in the two-stage approach can be seen for cT1 and cT2 tumours: 54.7% in 2015 and 87.2% in 2020 for cT1 tumours and 8.3% in 2015 and 24.1% in 2020 for cT2 tumours.

Fig. 5 Distribution of the one-stage approach and the two-stage approach for cT1–2 rectal cancer over time

TME, total mesorectal excision.

Primary total mesorectal excision versus completion total mesorectal excision

Table 1 shows the patient and tumour characteristics of the primary TME group and LE + cTME group before and after propensity score matching with corresponding standardized mean differences (SMDs).

Table 1 Patient and tumour characteristics before and after propensity score matching

	Unmatched cohort	Matched cohort	
Primary TME (n = 2706)	Local excision followed by completion TME (n = 345)	P	SMD	Primary TME (n = 740)	Local excision followed by completion TME (n = 322)	P	SMD	
Female	1082 (40.0)	148 (42.9)	0.302	0.059	322 (43.5)	138 (42.8)	0.817	0.016	
Age (years), mean (95% c.i.)	68.0 (67.7,68.4)	66.4 (65.4,67.3)	0.002*	0.181	66.8 (65.9,67.8)	66.4 (65.4,67.4)	0.544	0.046	
BMI (kg/m2), mean (95% c.i.)	26.5 (26.3,26.7)	26.3 (25.9,26.7)	0.449	0.044	26.4 (26.0,26.8)	26.4 (25.9,26.8)	0.874	0.012	
ASA grade			0.683	0.024			0.415	0.055	
 I–II	2187 (80.8)	282 (81.7)			152 (20.6)	56 (17.3)			
 III–IV	519 (19.2)	63 (18.3)			465 (62.9)	206 (64.1)			
Distance from ARJ (mm), mean (95% c.i.)	90.6 (89.0,92.3)	59.5 (54.6,64.3)	<0.001*	0.686	71.2 (67.0,75.3)	64.5 (59.7,69.3)	0.040*	0.156	
cT			<0.001*	1.013			<0.001*	0.252	
 cT1	387 (14.3)	199 (57.7)			312 (42.2)	176 (54.7)			
 cT2	2319 (85.7)	146 (42.3)			427 (57.8)	146 (45.3)			
cN			0.004*	0.141			0.104	0.132	
 cN0	2665 (98.5)	332 (96.2)			728 (98.4)	310 (96.3)			
 cNx	41 (1.5)	13 (3.8)			12 (1.6)	12 (3.7)			
Values are n (%) unless otherwise indicated. *Statistically significant. TME, total mesorectal excision; SMD, standardized mean difference; ARJ, anorectal junction.

Before matching, the primary TME group included older patients (68 versus 66.4 years) and more cT2 tumours (85.7% versus 42.3%). For the primary TME group, the tumours were located more proximally, with a mean distance from the anorectal junction of 90.6 mm versus 59.5 mm for the LE + cTME group (P < 0.001). After propensity matching for age, distance to the anorectal junction, ASA grade, and cT category, the primary TME group comprised 740 patients and the LE + cTME group comprised 322 patients. After matching, differences in cT category and distance from the anorectal junction remained significant. Univariable logistic regression analyses with corresponding ORs after propensity score matching of the primary and secondary outcomes are shown in Table 2.

Table 2 Univariable logistic regression analyses of primary and secondary outcomes after propensity score matching

	Primary TME (n = 740)	Local excision followed by completion TME (n = 322)	Univariable OR (95% c.i.)	P	
Conversion*	33 (4.7)	16 (5.0)	1.04 (0.52,2.08)	0.911	
Anastomosis	603 (81.5)	215 (66.6)	0.45 (0.33,0.62)	<0.001†	
Diverting ostomy	168 (22.7)	62 (19.2)	0.81 (0.56,1.17)	0.266	
End-ostomy	125 (16.8)	95 (29.6)	2.21 (1.59,3.05)	<0.001†	
R0 (>1 mm)	730 (98.7)	318 (98.7)	1.10 (0.201,5.72)	0.904	
Surgical complication	178 (24.0)	69 (21.4)	0.86 (0.62,1.20)	0.365	
Anastomotic leakage‡	79 (13.0)	23 (10.6)	0.79 (0.47,1.34)	0.381	
Readmission <90 days	95 (12.8)	42 (13.2)	1.03 (0.67,1.60)	0.877	
Re-intervention	124 (16.7)	38 (11.8)	0.67 (0.45,1.00)	0.047†	
Deceased <90 days	8 (1.1)	2 (0.6)	0.55 (0.11,2.72)	0.463	
Complication <90 days	241 (32.6)	97 (30.0)	0.89 (0.66,1.21)	0.448	
Values are n (%). *Primarily open procedures excluded. †Statistically significant. ‡Procedures without anastomosis excluded. TME, total mesorectal excision.

Perioperative outcomes

Perioperative outcomes are presented in Table 3. After matching, the end-ostomy rate was significantly higher for patients who underwent LE + cTME compared with the patients who underwent primary TME (29.6% versus 16.8%; OR 2.21 (95% c.i. 1.59 to 3.05); P < 0.001). In addition, the mean time interval between diagnosis and first surgery was longer for the LE + cTME group compared with the primary TME group (LE + cTME 48.0 days versus primary TME 42.6 days; P = 0.025). Conversion rates were comparable for the two groups: 4.7% and 5.0% for the primary TME group and the LE + cTME group respectively (OR 1.04 (95% c.i. 0.52 to 2.08); P = 0.911). The diverting ostomy rate was similar for the two groups (22.7% for the primary TME group versus 19.2% for the LE + cTME group; OR 0.815 (95% c.i. 0.56 to 1.18); P = 0.266).

Table 3 Perioperative outcomes after propensity score matching

	Primary TME (n = 740)	Local excision followed by completion TME (n = 322)	P	
Time between diagnosis and first surgery (days), mean (95% c.i.)	42.6 (40.3,45.0)	48.0 (43.6,52.5)	0.025*	
Conversion†	32.8 (4.7)	15.6 (5.0)	0.911	
Diverting ostomy	168 (22.7)	62 (19.2)	0.266	
End-ostomy	125 (16.8)	95 (29.6)	<0.001*	
Anastomosis	603 (81.5)	215 (66.6)	<0.001*	
Values are n (%) unless otherwise indicated. *Statistically significant. †Primarily open procedures excluded. TME, total mesorectal excision.

Histopathological outcomes

Histopathological outcomes are presented in Table 4. After matching, the primary TME group included slightly more pT1 tumours (39.9% versus 35.4%; P = 0.001) and pT3 tumours (19.0% versus 16.4%; P < 0.001), whereas the LE + cTME group included slightly more pT2 tumours (38.1% versus 37.6%; P = 0.003) and pN1 tumours (20.1% versus 17.9%; P = 0.046). Also, the LE + cTME group included more tumours with a poor differentiation grade (6.9% versus 1.7%; P < 0.001) and more mucinous carcinomas (5.4% versus 2.6%; P = 0.038). All other histopathological outcomes, including the rate of R0 resections, were comparable for the two groups (Table 4).

Table 4 Histopathological outcomes after propensity score matching

	Primary TME (n = 740)	Local excision followed by completion TME (n = 322)	P	
pT				
 pT0	12 (1.6)	20 (6.1)	Reference	
 pT1	295 (39.9)	114 (35.4)	0.001*	
 pT2	277 (37.6)	123 (38.1)	0.003*	
 pT3	145 (19.0)	53 (16.4)	<0.001*	
 pT4	6 (0.8)	1 (0.4)	0.079	
 pTis	1 (0.1)	1 (0.3)	>0.999	
 pTx	7 (0.9)	3 (0.9)	0.126	
 Missing	1 (0.1)	7 (2.3)	–	
pN				
 pN0	563 (76.1)	237 (73.6)	Reference	
 pN1	132 (17.9)	65 (20.1)	0.046*	
 pN2	31 (4.2)	9 (2.9)	0.059	
 pNx	8 (1.1)	2 (0.5)	0.041*	
 Missing	5 (0.7)	9 (2.9)	–	
Venous invasion				
 No	309 (41.8)	178 (55.4)	Reference	
 Yes	72 (9.8)	47 (14.5)	0.599	
 Missing	358 (48.4)	97 (30.1)	–	
CRM				
 Not involved	650 (87.9)	221 (68.5)	Reference	
 Involved	3 (0.4)	1 (0.4)	0.784	
 Missing	87 (11.7)	100 (31.1)	–	
R classification				
 >1 mm	724 (97.9)	314 (97.5)	Reference	
 0–1 mm	1 (0.2)	1 (0.6)	>0.999	
 Tumour in resection plane	8 (1.1)	2 (0.7)	0.696	
 Missing	6 (0.9)	4 (1.2)	–	
Differentiation grade				
 Well to moderate	692 (93.5)	263 (81.7)	Reference	
 Poor	13 (1.7)	22 (6.9)	<0.001*	
 Missing	35 (4.8)	37 (11.2)	–	
Histological type				
 Adenocarcinoma	684 (92.4)	294 (91.2)	Reference	
 Mucinous carcinoma	20 (2.6)	17 (5.4)	0.038*	
 Signet cell carcinoma	4 (0.6)	0 (0.0)	0.999	
 Other	7 (0.9)	4 (1.2)	0.652	
 Missing	25 (3.4)	7 (2.2)	–	
Values are n (%). *Statistically significant. TME, total mesorectal excision; CRM, circumferential resection margin.

Postoperative outcomes

Table 5 shows the postoperative outcomes of both groups. Complications within 90 days occurred in 32.6% of primary TME patients and in 30.0% of LE + cTME patients (OR 0.89 (95% c.i. 0.66 to 1.21); P = 0.448). More detailed analyses revealed that pulmonary events occurred more frequently in primary TME patients (4.6% versus 1.6%; P = 0.023), whereas more LE + cTME patients developed an ileus (5.4% versus 2.4%; P = 0.019). Primary TME patients required more re-interventions (16.7% versus 11.8%; OR 0.67 (95% c.i. 0.45 to 1.00); P = 0.047). No differences between primary TME patients and LE + cTME patients were observed with regard to anastomotic leakage (13.0% versus 10.6% respectively; P = 0.381), pelvic sepsis (13.8% versus 11.5% respectively; P = 0.418), surgical complications (24.0% versus 21.4% respectively; P = 0.365), readmissions (12.8% versus 13.2% respectively; P = 0.877), and deaths within 90 days (1.1% versus 0.6% respectively; P = 0.463).

Table 5 Postoperative outcomes after propensity score matching

	Primary TME (n = 740)	Local excision followed by completion TME (n = 322)	P	
Complication <90 days	241 (32.6)	97 (30.0)	0.448	
Surgical complication	178 (24.0)	69 (21.4)	0.365	
Complication specified				
 Anastomotic leakage*	79 (13.0)	23 (10.6)	0.381	
 Abscess*	7 (1.2)	3 (1.3)	0.940	
 Pelvic sepsis*	83 (13.8)	25 (11.5)	0.418	
 Bleeding	5 (0.6)	5 (1.5)	0.228	
 Ileus	18 (2.4)	17 (5.4)	0.019†	
 Fascial dehiscence	3 (0.4)	0 (0.0)	0.994	
 Bowel perforation	2 (0.2)	3 (0.9)	0.178	
 Urinary tract damage	0 (0.0)	2 (0.6)	0.991	
 Surgical-site infection	7 (0.9)	5 (1.5)	0.411	
 Pulmonary event	34 (4.6)	5 (1.6)	0.023†	
 Cardiac event	14 (0.2)	7 (2.0)	0.896	
 Thromboembolism	9 (1.2)	3 (0.9)	0.754	
 Other infectious event	33 (4.5)	15 (4.7)	0.883	
 Neurological event	12 (3.2)	3 (0.9)	0.367	
 Other	63 (8.5)	20 (6.2)	0.200	
Re-intervention	124 (16.7)	38 (11.8)	0.047†	
Re-intervention type‡				
 Laparoscopy	4 (3.2)	18 (47.4)	Reference	
 Radiological	2 (1.6)	1 (2.6)	0.994	
 Endoscopic	2 (1.6)	1 (2.6)	0.825	
 Laparotomic	35 (28.2)	6 (15.8)	<0.001†	
 Other	15 (12.1)	5 (13.2)	0.405	
 Re-intervention unknown	66 (53.2)	7 (18.4)	0.082	
Readmission <90 days	95 (12.8)	42 (13.2)	0.877	
ICU admission	91 (12.3)	28 (8.8)	0.116	
Deceased <90 days	8 (1.1)	2 (0.6)	0.463	
Values are n (%). *Procedures without anastomosis excluded. †Statistically significant. ‡Values are n (%) of re-interventions. TME, total mesorectal excision.

Discussion

This nationwide population study evaluated treatment strategies in early rectal cancer between 2012 and 2020 and compared the short-term outcomes of primary TME with those of LE + cTME. An increased incidence of two-stage procedures for both cT1 and cT2 rectal cancer was noted since 2015. Matched comparison revealed a significantly higher end-ostomy rate of 29.6% for the LE + cTME group versus 16.8% for the primary TME group (OR 2.21 (95% c.i. 1.59 to 3.05)). Although more re-interventions were observed for patients who underwent primary TME compared with patients who underwent LE + cTME, complication rates within 90 days were similar for both groups of patients.

Since the introduction of the bowel cancer screening programme in the Netherlands in 2014, the number of surgical procedures for both rectal carcinomas in general (stages 1–4) and early-stage carcinomas has increased temporarily. From 2015, there has been a steep decline in the total number of surgical interventions for all stages; however, this is not seen for early-stage rectal tumours, for which the number of procedures has decreased minimally. A previous study comparing the effect of colorectal screening programmes in European countries showed a similar peak and subsequent decline in Denmark, Belgium, and Slovenia24. These countries showed a rapid screening uptake and a high screening coverage. In contrast, colorectal cancer incidence either grew or remained steady in nations without extensive screening programmes, including Bulgaria, Estonia, and Norway24. Besides the relative increase in treatment of early-stage rectal cancer from 2015 to 2020, it seems that the treatment approach has changed, reflecting the demand for organ preservation and the expanding possibilities in piecemeal and en-bloc resection techniques. Whereas, in 2012, about two-thirds of patients with early-stage tumours were treated using a one-stage approach (primary TME), the two-stage approach (LE only or LE + cTME) has gained popularity over recent years and accounted for nearly half of all procedures in 2020. Nevertheless, the safety of the two-stage approach is still under debate.

This study showed a significantly higher end-ostomy rate within 90 days after surgery for LE + cTME patients compared with primary TME patients after propensity score matching (29.6% versus 16.8%). A potential explanation might be impaired healing of the anastomosis due to the fibrotic scar or local inflammation caused by the prior LE, which may contribute to the complexity of cTME15. A nationwide study performed in Norway also showed an increased end-colostomy rate after LE + cTME for early-stage rectal cancer. However, the reported ostomy rates were much higher in both groups (53% for the LE + cTME group and 32% for the primary TME group) and the reported difference was no longer significant after propensity score matching25. Another study showed a significantly higher ostomy rate for LE + cTME patients compared with primary TME patients (50.8% versus 45.9%; P = 0.006)26. The reported differences in ostomy rates could possibly be explained by the inclusion of abdominoperineal resections and Hartmann’s procedures, the prolonged follow-up of 5 years compared with 90 days, and the inclusion of advanced stages of rectal cancer. It is important to acknowledge that the increased ostomy rate may be due to multiple factors, including the tumour’s location, the patient’s initial wish for organ preservation, surgeons’ preferences, and the selection of tumours with unfavourable prognostic factors for the two-stage approach. Unfortunately, information on the shared decision process was not available in the DICA registry.

The comparable overall complication rates for the two groups is in line with previous studies showing no significant differences between primary TME and LE + cTME15,17–21,27. However, the higher rate of re-interventions for primary TME patients compared with LE + cTME patients (16.7% versus 11.8%) is in contrast with the latest meta-analysis, which showed a higher rate for LE + cTME compared with primary TME (11.6% versus 6.4%; OR 4.28 (95% c.i. 1.10 to 16.76); P < 0.04)19. This difference may be caused by the anastomosis rate for the primary TME group compared with the LE + cTME group and the small numbers of studies and patients included in the meta-analysis. Another possible explanation is the difference in type of LE before cTME. The meta-analysis focused on surgical full-thickness excisions only, whereas the present study included both full-thickness excisions and excisions that used more superficial and less invasive LE techniques, such as submucosal and intermuscular dissections, which may explain the lower rate of re-interventions for the LE + cTME group, due to less scar tissue and fewer anastomotic problems. Besides one study showing more surgical complications for the LE + cTME group (57.1% versus 20%; P = 0.048)20, other studies correspond to the present data, showing no differences between the two groups15,18. Although percentages of diverting ileostomies differ widely among studies (ranging between 15.3% and 100%), no differences were found between primary TME and LE + cTME14,15,20,21,26. In addition, the comparable rates of conversion, R0 resections, anastomotic leakage, readmissions, and death between primary TME and LE + cTME are consistent with previous studies14,15,17–21,25,27.

Current histological risk stratification manages to recognize a substantial proportion of T1 tumours as low risk. As a result of the two-stage approach, these patients are treated sufficiently using LE only and benefit from lesser morbidity and functional complaints compared with primary TME treatment8,28,29. However, for the remaining proportion of patients, for whom histopathological assessment reveals a high-risk pT1 or pT2 tumour, completion surgery is recommended and appears to be associated with an increased risk of end colostomy compared with primary TME. This is controversial and concerning, as end colostomies are known to be associated with an impaired quality of life, whereas the two-stage approach aims to result in improved quality of life30,31. Moreover, another study showed that the risk of an ostomy is one of the main reasons for patients with rectal cancer not to opt for surgery32. Despite the fact that accurate preoperative staging of early rectal cancer is challenging, it is important to be aware that LE before TME surgery increases the risk of an end colostomy and to inform patients on this matter before the start of treatment29. For patients with very low rectal cancers requiring abdominal perineal resection as primary surgery, LE may be particularly suitable, as concerns about the increased risk of end-ostomy are not applicable. Although not included in this study, it is crucial to emphasize the significance of long-term oncological outcomes, alongside morbidity and ostomy rates, for informed shared decision-making. This study focuses solely on short-term morbidity and ostomy rates, as previously published studies showed compromised short-term outcomes after two-stage procedures. Limited data are available regarding long-term outcomes of cTME; however, a recent study showed that cTME does not compromise oncological outcomes compared with primary resection33. Valuable insights are expected from the ongoing TESAR trial, a large randomized national study that will assess oncological outcomes, including local recurrence rate and overall survival, after cTME surgery in comparison with those after adjuvant chemotherapy after LE and the wait-and-see approach, significantly contributing to the decision-making process34.

Given that all Dutch hospitals are mandated to register their surgical data with the DICA, this study provides a high-quality and realistic reflection of the current surgical treatment of early rectal cancer in the Netherlands. Propensity matching for four relevant factors improved the comparability of primary TME and LE + cTME for a large group of patients. Although there were minor remaining differences in cT category and distance from the anorectal junction in the matched cohort, SMDs dropped significantly, thereby minimizing selection bias.

This study has several limitations. The design of the study based on the national DICA registry and the accompanying lack of data, including the type of LE, tumour diameter, location of the scar, histopathological risk factors of the LE specimen, timing of completion surgery, and information on the decision-making process (surgeon preferences in choice of treatment and patient wishes), causes selection bias and is therefore a limitation of this study. As the DCRA database only contains data for patients who received a surgical intervention, it was not possible to include patients who underwent endoscopic LE only. Therefore, it was not possible to determine the percentage of patients who benefitted from the two-stage approach and for whom radical surgery was prevented. For patients treated with LE + cTME, data on both endoscopic and surgical LE were available and both techniques were included. In addition, the DCRA database included missing variables due to alterations in registration protocols during the inclusion interval. Multiple imputation was used to account for the missing data instead of excluding many patients35. Moreover, hospital volume and surgeons’ experience may influence short-term outcomes, such as end-colostomy rates. Unfortunately, it was not possible to compare outcomes for low- and high-volume centres. Lastly, no distinction could be made between TME approaches (including laparoscopic TME, transanal TME, and robotic TME) due to high percentages of missing variables and relatively small groups. Nevertheless, these population-based data are highly valuable to detect real practice and publication bias present in cohort studies of specialized centres is avoided. As transanal TME may provide a better surgical overview of the tumour or fibrotic scar, future research should focus on whether the TME approach influences the end-ostomy rate.

This nationwide study on early rectal cancer showed an increase in the popularity of the two-stage approach (LE potentially followed by completion surgery) since 2015. Nevertheless, LE followed by completion surgery was associated with a higher end-ostomy rate compared with primary TME. Despite the rate of re-interventions, which was higher for primary TME, other outcomes, including complications, conversion rate, diverting ostomy rate, radical (R0) resection, readmissions, and death rate, were similar for both groups. It is important that clinicians and patients with early rectal cancer are aware of the higher end-ostomy risk associated with the two-stage approach when making a balanced and shared decision with regard to treatment.

Supplementary Material

zrae103_Supplementary_Data

Funding

The authors have no funding to declare.

Author contributions

Annabel S. van Lieshout (Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Visualization, Writing—original draft, Writing—review & editing), Lisanne J. H. Smits (Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Visualization, Writing—original draft, Writing—review & editing), Julie M. L. Sijmons (Data curation, Methodology, Resources, Writing—review & editing), Susan van Dieren (Conceptualization, Formal analysis, Methodology, Writing—original draft, Writing—review & editing), Stefan E. van Oostendorp (Conceptualization, Data curation, Methodology, Supervision, Writing—original draft, Writing—review & editing), Pieter J. Tanis (Conceptualization, Formal analysis, Investigation, Methodology, Resources, Supervision, Writing—original draft, Writing—review & editing), and Jurriaan B. Tuynman (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Supervision, Visualization, Writing—original draft, Writing—review & editing)

Disclosure

The authors declare no conflict of interest.

Supplementary material

Supplementary material is available at BJS Open online.

Data availability

All data supporting the findings of this study are available from the corresponding author upon reasonable request.
==== Refs
References

1 Integraal Kankercentrum Nederland (IKNL) . Cijfers Darmkanker. 2022. https://iknl.nl/kankersoorten/darmkanker/registratie (accessed 19 January 2023)
2 Lezoche  E, Paganini  AM, Fabiani  B, Balla  A, Vestri  A, Pescatori  L  et al  Quality-of-life impairment after endoluminal locoregional resection and laparoscopic total mesorectal excision. Surg Endosc  2014;28 :227–234 24002918
3 Traa MJ, Orsini RG, Den Oudsten BL, De Vries J, Roukema JA, Bosman SJ  et al  Measuring the health-related quality of life and sexual functioning of patients with rectal cancer: does type of treatment matter?  Int J Cancer  2014;134 :979–987 23934989
4 van der Heijden  JAG, Koeter  T, Smits  LJH, Sietses  C, Tuynman  JB, Maaskant-Braat  AJG  et al  Functional complaints and quality of life after transanal total mesorectal excision: a meta-analysis. Br J Surg  2020;107 :489–498 32154594
5 Borstlap  WAA, van Oostendorp  SE, Klaver  CEL, Hahnloser  D, Cunningham  C, Rullier  E  et al  Organ preservation in rectal cancer: a synopsis of current guidelines. Colorectal Dis  2018;20 :201–210
6 Backes  Y, Elias  SG, Groen  JN, Schwartz  MP, Wolfhagen  FHJ, Geesing  JMJ  et al  Histologic factors associated with need for surgery in patients with pedunculated T1 colorectal carcinomas. Gastroenterology  2018;154 :1647–1659 29366842
7 Hashiguchi  Y, Muro  K, Saito  Y, Ito  Y, Ajioka  Y, Hamaguchi  T  et al  Japanese Society for Cancer of the Colon and Rectum (JSCCR) guidelines 2019 for the treatment of colorectal cancer. Int J Clin Oncol  2020;25 :1–42 31203527
8 van Oostendorp  SE, Smits  LJH, Vroom  Y, Detering  R, Heymans  MW, Moons  LMG  et al  Local recurrence after local excision of early rectal cancer: a meta-analysis of completion TME, adjuvant (chemo)radiation, or no additional treatment. Br J Surg  2020;107 :1719–1730 32936943
9 Glynne-Jones  R, Wyrwicz  L, Tiret  E, Brown  G, Rödel  C, Cervantes  A  et al  Rectal cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol  2017;28 :iv22–iv40 28881920
10 Detering  R, Oostendorp  SE, Meyer  VM, Dieren  S, Bos  ACRK, Dekker  JWT  et al  MRI cT1–2 rectal cancer staging accuracy: a population-based study. Br J Surg  2020;107 :1372–1382 32297326
11 Marusch  F, Ptok  H, Sahm  M, Schmidt  U, Ridwelski  K, Gastinger  I  et al  Endorectal ultrasound in rectal carcinoma—do the literature results really correspond to the realities of routine clinical care?  Endoscopy  2011;43 :425–431 21234855
12 Ashraf  S, Hompes  R, Slater  A, Lindsey  I, Bach  S, Mortensen  NJ  et al  A critical appraisal of endorectal ultrasound and transanal endoscopic microsurgery and decision-making in early rectal cancer. Colorectal Dis  2012;14 :821–826 21920011
13 van Groningen  JT, van Hagen  P, Tollenaar  R, Tuynman  JB, de Mheen  PJM, Doornebosch  PG  et al  Evaluation of a completion total mesorectal excision in patients after local excision of rectal cancer: a word of caution. J Natl Compr Canc Netw  2018;16 :822–828 30006424
14 Hompes  R, McDonald  R, Buskens  C, Lindsey  I, Armitage  N, Hill  J  et al  Completion surgery following transanal endoscopic microsurgery: assessment of quality and short- and long-term outcome. Colorectal Dis  2013;15 :e576–e581 24635913
15 Levic  K, Bulut  O, Hesselfeldt  P, Bulow  S. The outcome of rectal cancer after early salvage TME following TEM compared with primary TME: a case-matched study. Tech Coloproctol  2013;17 :397–403 23192705
16 Serra-Aracil  X, Galvez Saldana  A, Mora-Lopez  LL, Montes  N, Pallisera-Lloveras  A, Serra-Pla  S  et al  Completion surgery in unfavorable rectal cancer after transanal endoscopic microsurgery: does it achieve satisfactory sphincter preservation, quality of total mesorectal excision specimen, and long-term oncological outcomes?  Dis Colon Rectum  2021;64 :200–208 33315715
17 Levic Souzani  K, Bulut  O, Kuhlmann  TP, Gögenur  I, Bisgaard  T. Completion total mesorectal excision following transanal endoscopic microsurgery does not compromise outcomes in patients with rectal cancer. Surg Endosc  2022;36 :1181–1190 33629183
18 Coton  C, Lefevre  JH, Debove  C, Creavin  B, Chafai  N, Tiret  E  et al  Does transanal local resection increase morbidity for subsequent total mesorectal excision for early rectal cancer?  Colorectal Dis  2019;21 :15–22 30300969
19 Eid  Y, Alves  A, Lubrano  J, Menahem  B. Does previous transanal excision for early rectal cancer impair surgical outcomes and pathologic findings of completion total mesorectal excision? Results of a systematic review of the literature. J Visc Surg  2018;155 :445–452 29657063
20 Piessen  G, Cabral  C, Benoist  S, Penna  C, Nordlinger  B. Previous transanal full-thickness excision increases the morbidity of radical resection for rectal cancer. Colorectal Dis  2012;14 :445–452 21689342
21 Morino  M, Allaix  ME, Arolfo  S, Arezzo  A. Previous transanal endoscopic microsurgery for rectal cancer represents a risk factor for an increased abdominoperineal resection rate. Surg Endosc  2013;27 :3315–3321 23479257
22 Pishgar  F, Greifer  N, Leyrat  C, Stuart  E.  MatchThem:: matching and weighting after multiple imputation.  R J  2021;13 :228
23 Ming  K, Rosenbaum  PR. Substantial gains in bias reduction from matching with a variable number of controls. Biometrics  2000;56 :118–124 10783785
24 Cardoso  R, Guo  F, Heisser  T, Hackl  M, Ihle  P, De Schutter  H  et al  Colorectal cancer incidence, mortality, and stage distribution in European countries in the colorectal cancer screening era: an international population-based study. Lancet Oncol  2021;22 :1002–1013 34048685
25 Lossius  WJ, Stornes  T, Myklebust  TA, Endreseth  BH, Wibe  A. Completion surgery vs. primary TME for early rectal cancer: a national study. Int J Colorectal Dis  2022;37 :429–435 34914000
26 van Gijn  W, Brehm  V, de Graaf  E, Neijenhuis  PA, Stassen  LP, Leijtens  JW  et al  Unexpected rectal cancer after TEM: outcome of completion surgery compared with primary TME. Eur J Surg Oncol  2013;39 :1225–1229 23972571
27 Overwater  A, Kessels  K, Elias  SG, Backes  Y, Spanier  BWM, Seerden  TCJ  et al  Endoscopic resection of high-risk T1 colorectal carcinoma prior to surgical resection has no adverse effect on long-term outcomes. Gut  2018;67 :284–290 27811313
28 Dekkers  N, Dang  H, van der Kraan  J, le Cessie  S, Oldenburg  PP, Schoones  JW  et al  Risk of recurrence after local resection of T1 rectal cancer: a meta-analysis with meta-regression. Surg Endosc  2022;36 :9156–9168 35773606
29 Smits  LJH, van Lieshout  AS, Grüter  AAJ, Horsthuis  K, Tuynman  JB. Multidisciplinary management of early rectal cancer—the role of surgical local excision in current and future clinical practice. Surg Oncol  2022;40 :101687 34875460
30 Näsvall  P, Dahlstrand  U, Löwenmark  T, Rutegård  J, Gunnarsson  U, Strigård  K. Quality of life in patients with a permanent stoma after rectal cancer surgery. Qual Life Res  2017;26 :55–64 27444778
31 Ross  L, Abild-Nielsen  AG, Thomsen  BL, Karlsen  RV, Boesen  EH, Johansen  C. Quality of life of Danish colorectal cancer patients with and without a stoma. Support Care Cancer  2007;15 :505–513 17103196
32 van der Valk  MJM, van der Sande  ME, Toebes  RE, Breukink  SO, Bröker  MEE, Doornebosch  PG  et al  Importance of patient reported and clinical outcomes for patients with locally advanced rectal cancer and their treating physicians. Do clinicians know what patients want?  Eur J Surg Oncol  2020;46 :1634–1641 32336626
33 Burghgraef  TA, Rutgers  ML, Leijtens  JWA, Tuyman  JB, Consten  ECJ, Hompes  R. Completion total mesorectal excision: a case-matched comparison with primary resection. Ann Surg Open  2023;4 :e327 37746593
34 Borstlap  WA, Tanis  PJ, Koedam  TW, Marijnen  CA, Cunningham  C, Dekker  E  et al  A multi-centred randomised trial of radical surgery versus adjuvant chemoradiotherapy after local excision for early rectal cancer. BMC Cancer  2016;16 :513 27439975
35 Pedersen  AB, Mikkelsen  EM, Cronin-Fenton  D, Kristensen  NR, Pham  TM, Pedersen  L  et al  Missing data and multiple imputation in clinical epidemiological research. Clin Epidemiol  2017;9 :157–166 28352203
