Executive Summary of the American Radium Society Appropriate Use Criteria for Local Excision in Rectal Cancer

Abstract

The goal of treatment for early stage rectal cancer is to optimize oncologic outcome while minimizing effect of treatment on quality of life. The standard of care treatment for most early rectal cancers is radical surgery alone. Given the morbidity associated with radical surgery, local excision for early rectal cancers has been explored as an alternative approach associated with lower rates of morbidity. The American Radium Society Appropriate Use Criteria presented in this manuscript are evidence-based guidelines for the use of local excision in early stage rectal cancer that include an extensive analysis of current medical literature from peer-reviewed journals and the application of a well-established consensus methodology (modified Delphi) used by a multidisciplinary expert panel to rate the appropriateness of imaging and treatment procedures. In those instances where evidence is lacking or not definitive, expert opinion may be used to recommend imaging or treatment. These guidelines are intended for the use of all practitioners and patients who desire information regarding the use of local excision in rectal cancer.

Introduction

Thirty-nine percent of patients diagnosed with rectal cancer present with American Joint Commission on Cancer stage I disease.¹ Historically these patients have been treated with low anterior resection or abdominoperineal resection with excellent local control and survival rates^2-4; however, these invasive procedures are associated with mortality and morbidity. Local excision (LE) has been presented as an option to patients whose other comorbid conditions would not allow them to tolerate more extensive surgery. Advantages of LE include minimal morbidity and mortality, with a more rapid recovery period.^5,6 Limitations of LE include lack of pathologic staging of lymph node micrometastases, which are less likely to be identified by staging imaging in early rectal cancer.⁷ In recent years there has been additional evidence supporting the use of LE.^8-13 Local transanal excision (TAE) has been typically recommended only for stage cT1N0 as defined by endorectal ultrasound (EUS) or magnetic resonance imaging (MRI) and conditional on specific patient selection criteria.¹⁴ There has been growing interest in the use of neoadjuvant radiation therapy (RT) or chemoradiation (CRT) to extend the indications for less radical surgery to selected patients with early-stage cancers at increased risk for locoregional recurrence (LR) or patients with severe comorbidities, and T3N0 cancers that have a complete or near-complete response to preoperative therapy. A few prospective multi-institutional trials have investigated the efficacy of combined LE and RT or CRT for these patients.^15-18

Methods and Materials

An updated analysis of current medical literature from peer-reviewed journals was conducted from 1/1/2008 to 10/15/2018 using the Preferred Reporting Items for Systematic Reviews and Meta-analyses¹⁹ guidelines to search the Ovid MEDLINE(R) without Revisions database to retrieve a comprehensive set of relevant articles. We developed strategies using subject and combinations of keywords search terms (Table 1). We reviewed the bibliographies of full articles for a comprehensive survey, and relevant studies were included. The literature was reviewed for quality of study design, cohort size, selection bias, variability of evaluation of participants in regard to time from exposure, and methods of assessments.

Table 1.

Literature search strategy for ARS Appropriate Use Criteria for Local Excision in Rectal Cancer^*

1	Exp rectal cancer or rectal adenocarcinoma or adenocarcinoma or the rectum/	(24,491)
2	Exp local excision or transanal excision or transanal endoscopic microsurgery or Transanal local excision or TEMS/	(11,310)
3	1 and 2/	(2046)
4	Exp radiotherapy or radiation or chemoradiotherapy or chemoradiation or Radiochemotherapy/	(381,919)
5	3 and 4/	(875)
6	3 or 5/	(2046)
7	Limit 6 to English language/	(1815)

Literature Search Summary

Of the 69 citations in the original bibliography, 39 citations were retained in the final document. Articles were removed from the original bibliography if they were more than 10 years old and did not contribute to the evidence or they were no longer cited in the revised narrative text.

A new literature search was conducted in October 2018 to identify additional evidence published since the Appropriateness Criteria Local Excision in Rectal Cancer topic was finalized. Using the search strategy described, 1815 articles were found. Some articles were not added to the bibliography because of poor study design; not being relevant or generalizable to the topic; giving unclear, misinterpreted, or biased results; or because the articles were already cited in the original bibliography.

The authors added 3 citations from bibliographies, websites, or books not found in the new literature search.

^*Literature search was performed on October 15, 2018. The beginning date was January 1, 2008, and the end date was October 15, 2018. Database: Ovid MEDLINE(R) without Revisions <1996 to October Week 2 2018>.

A well-established consensus methodology (modified Delphi)²⁰ was for the expert panel to rate the appropriateness of imaging and treatment procedures. The expert panel is a multidisciplinary panel of radiation, medical, and surgical oncologists with expertise in local excision in rectal cancer.

Results and Discussion

Workup

All patients should receive a full colonoscopy with biopsy, pathology review, proctoscopy, carcinoembryonic antigen (CEA), and computerized tomography (CT) of the chest, abdomen, and pelvis. Because depth of tumor invasion has been shown to be an independent predictor for lymph node metastases in rectal cancer,²¹ patients being considered for LE should have additional local staging to evaluate depth of penetration. EUS is 62% to 92% accurate for T staging and 64% to 88% accurate for N staging but is highly operator dependent.^22-27 MRI is more commonly included in the staging workup for patients with rectal cancer and provides more detailed anatomic information for locoregional staging.²⁸ A 2004 meta-analysis demonstrated that EUS and MRI have similar sensitivities and specificities for evaluations of pelvic lymph nodes (67% and 78%; 66% and 76%, respectively).²⁹ Currently pelvic MRI is the preferred modality for locoregional rectal cancer staging.³⁰

Surgical techniques

There are 3 operative approaches for LE of a distal rectal lesion: (1) TAE, (2) posterior transsphincteric (York-Mason procedure), or (3) posterior proctotomy (Kraske procedure). It is important to note that none of these procedures include lymph node evaluation. TAE is the most commonly used approach and is associated with low complication rates,³¹ with good patient reported quality of life (QOL) measures and anorectal function,³² and it has been found to be safe after CRT.^33-37 TAE involves full-thickness excision performed perpendicularly through the rectal wall into the perirectal fat avoiding tumor fragmentation. Negative deep and mucosal margins of at least 3 mm are recommended.

Patient selection

Historically, the best candidates for LE include small cT1N0 (<3 cm), low-lying tumors confined to the muscularis propria as defined by EUS or MRI and conditional on specific criteria (Table 2, variant 1). These inclusion criteria based on the work by Nash et al¹⁰ specify that the T1 lesion should: not invade the deepest one-third of the submucosa, be limited to ≤30% of the bowel circumference, be ≤3 cm in size, have clear margins (>3 mm), be without lymphovascular invasion (LVI) or perineural invasion (PNI), be mobile, and be within 8 cm of the anal verge. There should be no evidence of lymphadenopathy on pretreatment imaging.

Table 2.

Clinical condition: Local excision in rectal cancer^*

Treatment	Rating category†	Group median rating	SOE‡	SOR§
Local excision, pT1N0, and 4 mm negative margins
Observation	A	8	S	↑
Adjuvant RT alone	U	3	S	↑
Adjuvant CRT	U	3	S	↑
Local excision, pT1N0, and positive margins
Observation	U	1.5	S	↑
LAR or APR	A	8	S	↑
Adjuvant RT alone	M	4	S	↑
Adjuvant CRT	M	6	S	↑

Abbreviations: APR = abdominoperineal resection; CRT = chemoradiation; LAR = low anterior resection; RT = radiation therapy; SOE = strength of evidence; SOR = strength of recommendation.

5-FU based CRT in combination with LE may be appropriate for patients who are medically inoperable or refuse to undergo LAR or APR.

^*Variant 1: 57-year-old man with preoperative MRI staged cT1N0 freely mobile, moderately differentiated adenocarcinoma. Tumor is 2 cm in diameter, involves <25% of circumference, and is located 6 cm from anal verge. No lymphovascular space or perineural invasion is noted. CEA is normal.

^†A = usually appropriate; M = may be appropriate; U = usually not appropriate.

^‡S = strong; M = moderate; L = limited; EC = expert consensus; EO = expert opinion.

^§↑ = strong recommendation; ↓ = weak recommendation. Additional considerations do not strengthen or weaken the panel’s recommendation.

Patients with subclinical nodal metastases undergoing LE alone are at risk for LR. Advanced age,³⁸ pathologic features (large tumor size,³⁸ high tumor grade,³⁸ LVI, or PNI³⁹), and deep invasion⁴⁰ have been shown to be independent predictors for lymph node metastases and may be useful in identifying patients who would benefit from adjuvant therapy in addition to LE or potentially completion total mesorectal excision (TME).

Patients with positive margins after LE or piecemeal resections are at very high risk of LR and should be offered immediate radical surgery, that is, TME with low anterior resection or abdominoperineal resection performed via either a traditional open or laparoscopic approach. In general, patients with cT2 tumors have a sufficiently high risk of lymph node involvement to warrant consideration of neoadjuvant or adjuvant therapy if radical surgery is not performed (Table 3, variant 2). Patients with tumors invading the muscularis propria (T3) are at very high-risk (>30%) for LR after LE and should not be treated with LE alone but may be considered for neoadjuvant therapy followed by restaging and consideration of LE for nonsurgical candidates with adequate tumor response.

Table 3.

Clinical condition: Local excision in rectal cancer^*

Treatment	Rating category†	Group median rating	SOE‡	SOR§
Treatment options
LAR or APR	A	8	S	↑
LE alone	U	2	M	↑
LE followed by CRT∥	M	5	M	↑
Neoadjuvant CRT followed by LE¶	A	7	M	↑
LE followed by RT alone	U	3	M	↑
If LE with RT or CRT: RT dose
45 Gy/1.8 Gy	M	4	M	↑
50.4 Gy/1.8 Gy	M	5	M	↑
45-50.4 Gy/1.8 Gy to the pelvis followed by a 5.4-9 Gy/1.8 Gy boost	A	8	M	-
Simulation
Patient prone on belly board	A	8	S	↑
Supine	M	6	S	↑
Small-bowel contrast	A	7	M	↑
Patient immobilized	A	8	M	↑
Anal marker	A	8	EO	↑
Bladder full	A	8	S	↑
If LE with CRT: RT volume
Pelvis CTV and tumor GTV determined using CT/MRI; treatment to 2-3 cm proximal/distal to tumor	A	9	S	↑
Inclusive of internal iliac nodes	A	9	S	↑
Inclusive of external iliac nodes#	M	4	S	↑
Inclusive of inguinal nodes#	U	3	S	↑
Tumor bed alone with 2-3 proximal/distal margin	U	3	S	↑
Radiation technique
IMRT**	M	6	S	↑
3 field with photons	A	8	S	↑
4 field with photons	M	5	S	↑
AP/PA	U	3	S	↓

Abbreviations: AP/PA = anterior-posterior and posterior-anterior fields; APR = abdominoperineal resection; CRT = chemoradiation; CTV = clinical target volume; GTV = gross tumor volume; IMRT = intensity modulated radiation therapy; LAR = low anterior resection; LE = local excision; RT = radiation therapy; SOE = strength of evidence; SOR = strength of recommendation.

^*Variant 2: 65-year-old otherwise healthy woman with preoperative MRI staged cT2N0 moderately differentiated adenocarcinoma. Tumor is 3 cm in diameter, freely mobile, and is located 3 cm from anal verge without sphincter invasion. No lymphovascular space invasion is noted.

^†A = usually appropriate; M = may be appropriate; U = usually not appropriate.

^‡S = strong; M = moderate; L = limited; EC = expert consensus; EO = expert opinion.

^§↑ = strong recommendation; ↓ = weak recommendation. Additional considerations do not strengthen or weaken the panel’s recommendation.

^∥5-FU based CRT after LE may be appropriate for patients who are refusing to undergo LAR or APR.

^¶Neoadjuvant CRT + LE may be appropriate for carefully selected patients with excellent clinical response assessed by physical examination, endoscopy, and imaging.

^#Treatment of inguinal nodes and external iliac nodes may be considered for tumors with sphincter invasion.

^**IMRT may be appropriate in patients using supine position, especially those who are unable to lie prone.

Local excision with or without adjuvant (chemo)RT

Studies have reported LR rates of 7% to 40% and 25% to 62% for LE alone in T1 and T2 tumors, respectively.^10-12,41-49 There are increasing data to suggest the role of prognostic factors to select patients who are at risk for recurrence and may benefit from adjuvant treatment.⁵⁰ Tumor diameter,⁵¹ pathologic T stage and extent of submucosal invasion, high tumor grade, positive surgical margin, and PNI or LVI have been identified as independent predictors of recurrence after LE.^52-54

Although there are no randomized data comparing postoperative RT to observation, retrospective studies show that adjuvant RT may lower LR rates after LE for early stage rectal cancer by 10% to 20%.^{11,12,17,42,55,56} There are also studies evaluating the role of adjuvant CRT in patients with high-risk early rectal cancer after LE.^57,58 Non-randomized prospective studies that included patients receiving adjuvant CRT include the initial phase II study by the Radiation Therapy Oncology Group (RTOG 89-02), which assigned patients to observation (low-grade T1 tumors with negative margins) or CRT (54-65 Gy with 5-FU 1000 mg/m2 IV d1-3, d29-31) based on postexcision pathology. LR rates were 7%, 8%, and 23% for T1, T2, and T3 tumors, respectively.¹⁷ Another prospective study, Cancer and Leukemia Group B study (CALGB 8984) evaluated the role of LE with or without chemotherapy and RT in 177 patients with T1 and T2 adenocarcinomas of the rectum. T1 patients underwent LE followed by observation. T2 patients underwent LE followed by RT (54 Gy/30 fractions) and chemotherapy (5-FU 500 mg/m2 IV days [d] 1-3, d29-31). At a median follow-up of 48 months, the 6 year overall survival (OS) rate was 85%, and the disease-free survival (DFS) rate was 78% for all patients. Three of the 59 eligible T1 patients and 7 of the 51 eligible T2 patients had experienced LR, corresponding to 10-year local failure rates of 8% and 18%, respectively. It is important to note, however, that these were highly selected patients and that one-third of the patients were excluded after surgery because of large tumor size or questionable margin status.¹⁶

There are review data that attempt to further evaluate the relative benefit of CRT added to LE. A systematic review of 22 studies including 804 patients with T1, T2, and T3 tumors (35.1%, 58.0%, and 6.9%, respectively) who underwent LE (77% TAE) and adjuvant CRT or RT demonstrated LR rates of 5.8% for pT1, 13.8% for pT2, and 33.7% for pT3 tumors with an overall median DFS of 88%.⁵⁹ A 2004 to 2014 National Cancer Database (NCDB) study including 4822 patients with T2N0 rectal cancer included 4367 who underwent radical surgery, 242 who received CRT followed by LE, and 213 who received LE followed by CRT as the primary treatment. Five year OS was similar between the groups (76.1%, 79.7%, and 77.4%, respectively).⁶⁰ A meta-analysis including 14 studies including 405 patients with pT1/pT2 rectal cancer removed by TAE and followed by adjuvant CRT and 7 studies including 130 patients who underwent completion TME demonstrated a LR rate of 14% versus 7%, respectively. In this study, recurrence rates after adjuvant CRT versus completion TME were 10% and 6% for T1 tumors, and 15% versus 10% for T2 tumors.⁶¹ These studies support the use of adjuvant therapy in addition to LE of high-risk pT1 rectal adenocarcinomas with poor prognostic features and pT2 tumors.

A summary of selected publications that discuss local excision with or without adjuvant (chemo)RT is provided in Table 4.

Table 4.

Summary of selected publications for local excision with or without adjuvant (chemo)RT

Reference	Number of patients/stage(s) of patients	Study type	Study results
Nash et al10	284 patients/145 radical resections 137 TAE	Prospective database LE, No adjuvant therapy	LR -13.2% TAE vs 2.7% radical resection (P = .001); 5-y DSS - 87% TAE vs 96% radical resection (P = .03)
Paty et al11	125 patients/T1-T2	Retrospective study LE, No adjuvant therapy	10-y LR - 17% T1 and 26% T2; 10-y OS - 74% T1 and 72% T2;
Wentworth et al12	285 patients/localized rectal cancer	Retrospective study LE +/− adjuvant therapy	5-y OS - 76%, 10y OS - 42%; 5-y DFS - 69%, 10y - DFS -58%; LF rate - 16%; Adjuvant therapy did not affect OS or LF rates
Nam et al13	420 patients/T1	Prospective study LE or radical resection, No adjuvant therapy	LF rate - 3.7% patients with high-risk group pathologic features (resection type and vascular invasion); No LD in patients without high-risk group pathologic features; 5-y OS rate was not associated with resection type (radical 98.9%, LE 95.2%)
Greenberg et al16	110 patients/59 T1 51 T2	Prospective study LE, No adjuvant therapy	10-y OS -84% T1 and 66% T2; DFS - 75% T1 and 64% T2; LR rates - 8% T1 and 18% T2 lesions
Russell et al17	65 patients/localized rectal cancer	Prospective phase II study LE, No adjuvant therapy	LR - 16% overall; 5-y OS 88% for patients with limited cancer involving the middle and lower rectum; Local and distant failure is associated with increasing depth of tumor invasion
Rackley et al41	93 patients/T1-T3	Retrospective study LE, No adjuvant therapy	5-y OS - 78%; 5-y local control - 86% (92.5% for T1 lesions); 5-y PFS - 83%
Chakravarti et al42	99 patients/localized rectal cancer	Retrospective study LE +/− adjuvant therapy	5-y local control - 76% LE alone and 90% LE + adjuvant pelvic irradiation; 5-y RFS - 66% LE alone and 74% LE + adjuvant pelvic irradiation
Folkesson et al44	10,181 patients/localized rectal cancer 643 had a LE	Retrospective study LE or radical resection, +/−preoperative RT	5-y CSS - 95.3% for 256 patients with stage I who had LE; 5-y LR rate -7.2%; After adjustment for age, sex, tumor stage and preoperative RT, the relative risk of death from cancer was the same for LE as radical resection.
Peng et al51	350 patients/early stage	Retrospective study LE or radical resection, No adjuvant therapy	5-y LR rate - 14.1% LE vs 3.3% in standard resection (P = .0004); OS rate not significantly different; LE was an independent risk factor for 5 and 10-y OS; Tumor grade was independent risk factor for 5y LR; T stage was independent risk factor for 10y OS; Size ≥2.5 cm was independent risk factor for LR in LE group (5-y LR rate for ≥2.5 cm was 40%, compared with 4.3% for tumors <2.5 cm [P = .001])
Han et al55	83 patients/localized rectal cancer	Retrospective study LE, with or without pre- or postoperative RT	DFS rate - 90.0% T1 and 72.3% T2; 76% underwent adjuvant postoperative RT after LE and DFS rates between radiation and nonradiation group were significantly different for T2 (81.6% vs 33.3%; P < .05), but not for T1 tumors (90.9% vs 87.5%, P > .05); LR rate - 13.1% (median time to relapse 15 months, range: 10-53) Risk factors for LR - size ≥3 cm, poorly differentiated adenocarcinoma and T2 tumor
Jeong et al57	83 patients/high-risk pT1 or T2 (≤3 cm, and/or resection margin ≤3 mm, and/or LVI, and/or nonfull thickness excision, or unknown records regarding those features, or pT2 cancer)	Prospective study LE followed by CRT (RT [50.4 Gy] and concurrent 5-FU and leucovorin)	18.1% pT2; 26.5% with radial margin ≥3 mm; 25.3% size ≥3 cm; 15.7% with LVI; 5-y OS, LRFS, and DFS rates - 94.9, 91.0, and 89.8%, respectively; Multivariate analysis did not identify any significant factors for OS or LRFS, but the only significant factor affecting DFS was the pT stage (P = .027)
Cutting et al59	22 studies/early stage 804 patients	Meta-analysis LE with or without adjuvant therapy	Indications for LE included favorable histology, patient choice and comorbidities; Included 35.1% T1, 58% T2 and 6.9% T3; TAE was used in 77.7%; Adjuvant therapy included long-course CRT or RT; Pooled LF - 5.8% pT1, 13.8% pT2 and 33.7% pT3; Overall median DFS - 88%
Lee et al60	4822 patients/T2N0 242 neoadjuvant CRT + LE 213LE + adjuvant CRT 4,367 radical surgery	Meta-analysis Radical surgery or LE with or without neoadjuvant or adjuvant CRT	No differences in patient characteristics, but more high-risk features in the LE + Adj-CRT group 5-y OS - 77.4% radical surgery vs 76.1% neoCRT + LE vs LE + Adj-CRT 79.7%; Older age, male sex, and higher Charlson score independently risk for worse OS
Borstlap et al61	585 patients/pT1-2 405 patients treated with adjuvant CRT (14 studies) 130 patients treated with completion TME (7 studies)	Meta-analysis LE + adjuvant CRT or LE + completion TME	LR rate - 14% LE + adjuvant CRT vs 7% LE + completion TME; Weighted averages for LF rate for T1 tumors- 10% for LE + adjuvant CRT and 6% for LE + completion TME; Weighted averages for LF rate for T2 tumors - 15% for LE + adjuvant CRT and 10% for LE + completion TME

Abbreviations: CI = confidence interval; CRT = chemoradiation; CSS = cancer specific survival; DFS = disease-free survival; DSS = disease-specific survival; LE = local excision; LF = local failure; LR = locoregional recurrence; LRFS = locoregional relapse-free survival; OS = overall survival; PFS = progression-free survival; RFS = relapse-free survival; RS = radical surgery; RT = radiation therapy; TAE = transanal excision.

Local excision after neoadjuvant radiation with or without chemotherapy

Because LE alone is associated with a higher risk of LR and inferior OS for patients with >T1 rectal cancer, studies have aimed to evaluate the relative benefit of neoadjuvant RT with or without chemotherapy followed by LE. Data from retrospective studies^{18,36,55,62-66} and prospective studies^15,34,35 have demonstrated safety and LR rates ranging from 2.0% to 13.2% using this approach. The Polish Colorectal Cancer Study Group performed a phase III study for patients with cT1-2N0M0 or borderline cT2/T3N0M0 rectal cancer < 4 cm who were randomized to receive either 5 × 5 Gy to the whole pelvis plus 1 × 4 Gy tumor boost 1 week later or CRT (50.4 Gy in 28 fractions plus 3 × 1.8 Gy boost and 5-FU with leucovorin). LE was performed 6 to 8 weeks later. Patients with ypT0-1R0 disease were observed. Completion TME was recommended for poor responders (46% with ypT1R1/ypT2-3). Of 61 randomized patients, 51 were appropriate for analysis; 29 in the short-course group and 22 in the CRT group. Complete pathologic response (pCR) was not statistically different and was observed in 66% of patients in the short-course group and in 86% in the CRT group. The median follow-up was 8.7 years. LR and OS rates at 10 years were worse for the short-course group (35% and 47%, respectively) compared with the CRT group (5% and 85%, respectively). In total, 22% (n = 11) of patients experienced LR including 7% (1 of 15) with cT1 disease, 22% (6 of 22) with cT2, and 44% (4 of 9) with cT2/T3. Seventy-three percent of LR occurred within 3 years of follow-up and 91% within 5 years. Of the 11 patients with local failure, all were intraluminal, and salvage surgery was performed in 9 patients (82%); 7 patients underwent TME and 2 had a second LE because they refused TME or were deemed unfit. In the 2 patients remaining, salvage resection was not undertaken because of unresectable tumor or comorbidity.⁶⁷

The American College of Surgeons Oncology Group (ACOSOG) Z6041 investigated the oncological and functional outcomes of neoadjuvant CRT and LE for patients with pretreatment stage T2N0 rectal cancer. In this singlearm phase 2 study, 79 patients with clinical T2N0 rectal adenocarcinoma staged by EUS or endorectal coil MRI, measuring less than 4 cm in greatest diameter, involving less than 40% of the circumference of the rectum, located within 8 cm of the anal verge, received neoadjuvant CRT (twice daily capecitabine 725 mg/m² d1-14 and 22-35, with oxaliplatin 50 mg/m² on weeks 1, 2, 4, and 5), and RT (1.8 Gy to a dose of 45 Gy, followed by a boost of 5.4 Gy) followed by LE. Forty-four percent of patients achieved a pCR, and 64% of tumors were downstaged to ypT0-1. Approximately 5% of patients were found to have ypT3 tumors at the time of LE. All but one patient had negative margins. Two patients had no surgery, 1 underwent TME, and 4 additional patients who completed protocol treatment had TME (1 with positive margin and 3 ypT3 tumors). The therapy was associated with 39% of patients developing grade ≥3 treatment-related complications (29% had grade 3 gastrointestinal toxicity). Because of a higher than expected toxicity, capecitabine dose was reduced to 725 mg/m² twice a day, 5 days/wk, for 5 weeks, and the total dose of radiation was reduced to 50.4 Gy. Oxaliplatin dose was unmodified. With a median follow-up of 56 months, the estimated 3 year DFS for the intention-to-treat group was 88.2% and for the per-protocol group was 86.9%. The authors of this study concluded that the data suggest that neoadjuvant CRT followed by LE might be considered as an organ-preserving alternative in carefully selected patients with clinically staged T2N0 tumors who refuse, or are not candidates for, transabdominal resection.¹⁵ Numerous additional studies support these findings.^56,64,68-70 Anorectal function and QOL were assessed at enrollment (71 patients) and 1 year postoperatively (66 patients) for ACOSOG Z6041 using the Fecal Incontinence Severity Index, Fecal Incontinence Quality of Life Scale, and Functional Assessment of Cancer Therapy-Colorectal Questionnaire. CRT followed by LE had minimal effect on anorectal function 1 year after surgery and was associated with stable overall QOL, with mixed effects on different subscales. Fecal Incontinence Quality of Life results were significantly worse in the lifestyle, coping/behavior, and embarrassment categories; there were no differences in the Functional Assessment of Cancer Therapy overall score, but the physical well-being subscale was significantly worse and emotional well-being was improved after surgery.⁷¹

Neoadjuvant CRT was also evaluated in the prospective, randomized, multicenter French Research Group of Rectal Cancer Surgery-2 (GRECCAR-2) phase III trial, which included 186 patients with pretreatment T2-3 rectal cancers <8 cm from the anal verge, ≤4 cm, and without evidence of metastatic disease (N + allowed) who received CRT (45-55 Gy and concurrent fluoropyrimidine). After CRT 145 good clinical responders (residual tumor ≤ 2 cm) were randomly assigned to LE (n = 74) or TME group (n = 71). In the LE group, a completion TME was required in 26 patients (35%) because of tumor stage ypT2-3. Unfortunately, this study failed to show superiority of LE over TME in a composite outcome involving death, recurrence, morbidity, and side effects at 2 years after surgery because many patients in the LE group received a completion TME, thus increasing their side effects and morbidity. Of the 89 patients who underwent TME the nodal positivity rate was 0% in the T0/T1 tumors and 8% in the T2 and T3 tumors, suggesting that many patients could have avoided TME. However, better techniques may help identify patients who may be able to avoid TME (Table 5, variant 3).⁷²

Table 5.

Clinical condition: Local excision in rectal cancer^*

Treatment	Rating category†	Group median rating	SOE‡	SOR§
APR	M	6	S	↓
Neoadjuvant CRT followed by LAR or APR∥	A	8	S	↑
Neoadjuvant CRT followed by LE¶	M	4	M	↓
LE alone	U	1	L	↓
LE followed by CRT	U	2	M	↓
Neoadjuvant CRT followed by active surveillance#	M	4	M	↓

Abbreviations: LAR = low anterior resection; APR = abdominoperineal resection; LE = local excision; CRT = chemoradiation; SOE = strength of evidence; SOR = strength of recommendation.

^*Variant 3: 60-year old woman with MRI staged cT3N0 adenocarcinoma located 4 cm from anal verge, 3.5 cm in maximum dimension with 40% circumferential involvement. Surgeon recommended APR.

^†A = usually appropriate; M = may be appropriate; U = usually not appropriate.

^‡S = strong; M = moderate; L = limited; EC = expert consensus; EO = expert opinion.

^§↑ = strong recommendation; ↓ = weak recommendation. Additional considerations do not strengthen or weaken the panel’s recommendation.

^∥LAR or APR should be used in patients with poor response or stable disease after neoadjuvant therapy as assessed by physical examination, repeat endoscopy, and imaging studies.

^¶LE may be considered in patients with good response or cCR after neoadjuvant therapy as assessed by physical examination, repeat endoscopy, and imaging studies, preferably in the setting of clinical trial.

^#Active surveillance may be considered in patients with cCR after neoadjuvant therapy as assessed by physical examination, repeat endoscopy, and imaging studies, preferably in the setting of clinical trial.

A summary of selected publications that discuss local excision after neoadjuvant RT with or without chemotherapy is provided in Table 6.

Table 6.

Summary of selected publications for local excision after neoadjuvant radiation with or without chemotherapy

Reference	Number of patients/stage(s) of patients	Study type	Study results
Garcia-Aguilar et al15	72 patients/T2N0	Prospective phase II trial	pCR (44%); Downstaged to ypT0-1(64%); ypT3 (5%) 1 with positive node and 1 with positive margins
Perez, et al18	23 patients/localized rectal cancer 13 patients/benign tumors	Retrospective study	Median hospital stay - 2 d; 30-d complication rate - 56% for grade II/III complications for patients undergoing neoadjuvant CRT (vs 23% for no neoadjuvant; P = .05); 30-d readmission rate 43% for patients undergoing neoadjuvant CRT (vs 7% for no neoadjuvant; P = .02); Wound dehiscence −70% for patients undergoing neoadjuvant CRT (vs 23% no neoadjuvant; P = .03)
Lezoche et al34	100 patients/T2N0 50 LE 50 laparoscopic TME	Prospective randomized phase III trial	Downstaging 51% for neoadjuvant CRT vs 26% laparoscopic; R0 resection in all patients; LR 8% LE vs 6% TME; No statistically significant difference in DFS (P = .686).
Lezoche et al35	100 patients/uT2-3	Retrospective study	9 pT1, 54 pT2 and 19 pT3 tumors; pCR 3%;15% microscopic residual tumor; LF rate 5%; CSS 89% at 90 mo; OS rate 72%. Salvage APR 3%.
Guerrieri M, et al62	196 patients/T1N0, T2N0, T3N0	Retrospective study	CSS rate 100% for pT1, 90% for pT2, and 77% for pT3 patients.
Guerrieri et al63	120 patients/T1N0, 185 patients/T2N0 120 patients/T3N0	Retrospective study	Patients with T1-N0 lesions and favorable histology underwent TEM immediately; Patients with preoperative stage T2-T3-N0 underwent preoperative CRT; Patients with T2-T3-N0 lesions were restaged 30 d after RT and were then operated on 40-50 d after neoadjuvant therapy; No perioperative mortality nor intraoperative complications. Conversion to other surgical procedures was never required; Major complications (urethral lesions, perianal or retroperitoneal phlegmon and rectovaginal fistula) - 1.4%; Minor complications (partial suture line dehiscence, stool incontinence and rectal hemorrhage) - 9.9%; Pathology 18.8% pT0, 36% pT1, 35.5% pT2, and 9.6% pT3 lesions; CSS rates - 100% for pT1, 93% for pT2, and 89% for pT3 patients.
Kundel et al64	320 patients/localized rectal cancer	Retrospective study	After CRT, 93% patients had radical surgery, 6% had LE, and 3% did not have surgery; Median follow-up of 48 mo, 4 patients who had radical surgery and no patients with LE had a recurrence; DFS, pelvic RFS, and OS rates were similar in both groups.
Shin et al66	34 patients/T2N0	Retrospective study	All patients had TAE or transanal minimally invasive surgery; 55.9% pCR; no patients with pCR had LR 3-y LRFS and DFS - 100.0% and 97.1%, respectively
Wawok et al67	51 patients/cT1-2N0M0 or cT2/T3N0M0 < 4 cm 29 short course RT 22 long course CRT	Prospective randomized phase III trial	ypT0-1R0 - 66% of short-course group and 86% in the CRT group, P = .11; Completion TME - 46% of patients with ypT1R1/ypT2-3; 10-y LF - 35% short course vs 5% CRT, P = .036; LR at 10 y - 79% for ypT1R1/T2-3 without TME; 10y OS 47%short course vs 86% CRT, P = .009.
Rullier et al72	186 patients/T2 or T3 ≤ 4 cm	Prospective, randomized phase III trial	Patients with good clinical response to neoadjuvant CRT (residual tumor ≤2 cm) were randomly assigned by the surgeon to LE or TME; In the LE group, a completion TME was required if tumor stage was ypT2-3 (26 patients); 145 good clinical re were randomly assigned, 74 to LE and 71 to TME; No difference between the groups - one or more events from the composite primary outcome occurred in 41 (56%) of 73 patients in the LE group and 33 (48%) of 69 in the TME group (OR 1.33, 95% CI 0.62-2.86; P = .43)
Hallam et al73	1068 patients/T2 or T3 22 studies: 14 cohort 5 comparative cohort 1 randomized controlled	Meta-analysis	Pretreatment T2 and T3 tumors accounted for 46.4% and 30.7% of cases; Long-course CRT in all of the studies, except one cohort of 64 patients who received short-course RT; Pooled complete clinical response - 45.8% (95% CI, 31.4%-60.5%); Pooled pCR - 44.2% (95% CI, 36.4%-52.0%); Median follow-up - 54 months (range, 12-81 months); ypT0 tumors pooled LR rate - 4.0% (95% CI, 1.9%-6.9%); Median DFS rate - 5.0% (95% CI, 87.4%-100%); Pooled LF and median DFS rates for ypT1 tumors or higher (21.9% [95% CI, 15.9%-28.5%] and 68.0% [58.3%-69.0%])

Abbreviations: CI = confidence interval; CRT = chemoradiation; CSS = cancer specific survival; CT = computed tomography; DFS = disease-free survival; EUS = endoscopic ultrasound; LE = local excision; LF = local failure; LR = locoregional recurrence; LRFS = locoregional relapse-free survival; MRI = magnetic resonance imaging; OR = odds ratio; OS = overall survival; pCR = pathologic complete response; RFS = relapse-free survival; RT = radiation therapy; TAE = transanal excision; TEM = transanal endoscopic microsurgery; TME = total mesorectal excision.

The degree of tumor response to neoadjuvant therapy is variable. In many cases patients without clinical evidence of persistent tumor after neoadjuvant treatment (complete clinical response [cCR]) are found to have a pCR. The management of patients with a near-cCR to neoadjuvant therapy and the potential prognostic value of pCR to neoadjuvant therapy followed by LE for LR is a topic of current investigation. A systematic review was conducted to determine the oncological outcomes and morbidity of LE after neoadjuvant therapy incorporating 20 studies (14 cohort, 5 comparative cohort, and 1 randomized controlled trial), consisting of 1068 patients with pretreatment clinical stage T2 and T3 tumors accounting for 46.4% and 30.7% of cases, respectively. Long-course neoadjuvant RT (with or without chemotherapy) followed by LE was delivered in all the studies, except to a cohort of 64 patients who received short-course RT without chemotherapy. Pooled cCR rate was 45.8% and pooled pCR rate was 44.2%. At a median follow-up of 54 months, ypT0 tumors had a pooled LR rate of 4.0% and a median DFS rate of 95.0%. For ypT1 tumors and higher pooled LR, median DFS rates were 21.9% and 68.0%. Pooled incidence of complications was 23.2%, with suture-line dehiscence observed in 9.9%. The results of this pooled analysis are limited by selection bias, limited sample sizes, and study quality/design.⁷³ The multi-institutional Transanal Endoscopic Microsurgery (TEM) After Radiochemotherapy for Rectal Cancer prospectively evaluated the number of patients with minimal residual disease (ypT0-1) after neoadjuvant CRT and TEM for early stage rectal cancer. This study included 10 patients with pretreatment clinical stage T1N0, 29 patients with T2N0, and 16 patients with T3N0 in rectal cancer who received neoadjuvant CRT with planned TEM. Among 47 patients who had TEM, ypT0-1 disease was found in 30, ypT0 N1 in 1, ypT2 in 15, and ypT3 in 1. After median follow-up of 17 months, 4 LRs were observed, including 3 of 9 patients with ypT2 tumors who declined further surgery and including one patient with ypT1 disease.⁷⁴

Restaging of patients being considered for LE after neoadjuvant therapy can be even more challenging using standard staging techniques. It is important to evaluate not only response in the primary tumor after neoadjuvant CRT when considering LE; draining lymph nodes should also be carefully re-examined. The importance of this concept is demonstrated in a retrospective study of 725 patients, of which 51% had node-positive disease at diagnosis based on CT, MRI, or EUS, for whom the incidence of lymph node metastases was 9.7% for ypT0 and 17.6% for ypT1 after neoadjuvant CRT and radical surgery.⁷⁵ One prospective multicenter study demonstrated that restaging MRI using lymph node-specific contrast (ultrasmall superparamagnetic iron oxide) interpreted by an experienced radiologist can select rectal cancer with low risk of undetected nodal metastases (negative predictive value = 0.9) after neoadjuvant CRT and may be useful in identifying candidates for LE.⁷⁶ A subsequent meta-analysis assessing this study and 13 other articles found nodal restaging accuracy ranged from 60% to 88%, with a mean accuracy of 72%.⁷⁷ However, this study included articles that assessed nodes on the basis of size, morphologic criteria, or both, and because up to 15% of nodes <3 mm will be malignant, it is not recommended that nodes are evaluated on the basis of size.⁷⁸ Other investigators have demonstrated that MRI can detect reductions in tumor volume after neoadjuvant therapy and that a >75% tumor volume reduction ratio is significantly associated with a high pCR rate, which may identify patients who are candidates for LE after neoadjuvant CRT.⁷⁹ In a series of 36 patients who had an unrecognized complete response at the time of restaging after CRT, it was noted that overstaging was mainly because of residual mucosal abnormalities at endoscopy, mixed signal intensity of irregular fibrosis on T2-MRI, diffusion restriction on DWI, and suspicious lymph nodes, which may not be associated with residual tumor.⁸⁰

Comparisons between local excision with or without (chemo)radiation and standard resection

An analysis of data from >154,000 resected rectal cancer patients diagnosed from 1998 to 2010 from the NCDB study found that T1 and T2 rectal cancer excised with proctectomy was associated with higher rates of tumor-free surgical margins compared with LE (95% vs 76% respectively). There was also a small but significant decrease in OS for the T1N0 group.⁸¹ Other systematic reviews note patient heterogeneity included in the studies resulted in difficulty drawing conclusions when comparing the effectiveness of TEM and radical resection in the treatment of T1 and T2 rectal cancer. The 2016 systematic review published by Sajid et al⁸² included 10 trials (942 patients) that were significantly diverse in stage and grade of rectal cancer and the use of neoadjuvant CRT. Results demonstrated a trend toward a higher risk of LR (odds ratio, 2.78; P < .003) and overall recurrence (P < .01) after TEM compared with radical resection. The risk of distant recurrence, OS, and mortality was similar. TEM was associated with a shorter operation time and hospital stay and a reduced risk of postoperative complications (P < .0001). Although TEM appeared to have clinically measurable advantages, firm conclusions could not be drawn.⁸² Another meta-analysis published by Lu et al⁸³ comparing the efficacy of TEM compared with TME for treatment of T1 rectal cancer demonstrated that the distant metastasis, OS, and DFS rates did not differ between TEM and TME, although the LR rate after TEM was higher compared with TME. This study included 1 randomized control trial and 6 nonrandomized controlled trials (860 patients total with 557 treated with TME and 303 treated with TEM). LR rates were significantly different between TEM and TME (odds ratio, 4.62; P = .0003).

Comparisons between local excision and standard resection for rectal cancer

Three small randomized controlled trials including 25 to 50 patients per treatment arm compared LE to radical resection for stage I rectal cancer and demonstrated similar oncologic outcomes.^34,84,85 In addition, 3 meta-analyses have also compared LE to radical resection for this subset of patients.^86-88 Shaikh et al⁸⁷ published a meta-analysis that compared LE and radical resection following neoadjuvant CRT which included patients with all stages of disease. No differences in LR, 10-year OS, or DFS were noted between LE and radical resection in the pooled analyses. Subgroup analyses were possible for LR and DFS for T3 tumors, also showing no worse outcomes, with LE leading the authors to note this as an option for patients at high risk for radical surgery.⁸⁷ Another meta-analysis that did not include studies incorporating CRT reported unadjusted risk ratios for 5-year OS from 12 observational studies ranging from 0.11 to 2.87. This meta-analysis included 7 studies that compared TAE to radical resection and 5 studies that compared TEM to radical resection (risk ratios for 5 year survival, 0.11-1.53). Although significantly worse OS was noted for those patients receiving LE, the TEM subgroup did not have worse OS compared with radical resection. Local resection was found to be associated with lower perioperative mortality, postoperative complications, and need for a permanent stoma. The authors stated that results were not influenced by a higher proportion of tumors located in the lower third of the rectum because metaregression in case of similar ratio of lower-third cancers was not significant. Five year DFS risk ratios, reported in this study from 10 observational studies comparing local resection to radical resection, ranged from 0.31 to 8.31.⁸⁶ An additional systematic review and meta-analysis evaluating the combination of the Shaikh et al⁸⁷ and Kidane et al⁸⁶ studies showed no difference in OS or DFS between LE versus radical resection for stage I rectal cancer (T1-T2, N0). In this study LR more frequently occurred after LE (relative risk, 1.90; 95% CI, 0.57-6.32), but significance was not reached because of the low event rate (P = .30). Secondary outcomes, including blood loss, operative time, hospital duration, number of permanent stomas, and perioperative mortality, favored LE.⁸⁸ An additional meta-analysis compared outcomes for 121 patients with T2 rectal cancer who received TEM alone (n = 59) or after neoadjuvant therapy (n = 62) with 174 patients who were treated with TME. Although there were no significant differences in LR, overall recurrence, or OS rates between the TME and TEM + neoadjuvant treatment groups, TEM without neoadjuvant therapy was associated with increased LR, overall recurrence, and shorter OS for patients with T2 low lying rectal cancer.⁸⁹ Similar results were observed in the NCDB analysis evaluating outcomes for 4822 patients with T2N0 rectal cancer (4367 underwent radical surgery, 242 received CRT followed by LE, and 213 received LE followed by CRT). With a mean follow-up period of 48.6 months, there were no differences in 90-day mortality or 5-year OS.⁶⁰

Comparison between LE and standard resection continues to be evaluated in a noninferiority, multicenter, prospective, randomized controlled study (https://clinicaltrials.gov/ct2/show/NCT01308190). This trial will included 173 patients with T2-T3(superficial), N0, M0 rectal adenocarcinoma less than 10 cm from the anal verge and up to 4 cm in size, randomized to either preoperative CRT followed by TEM or TME alone. Patients will be followed for local control and systemic relapse.⁹⁰

Nonoperative management: The watch-and-wait strategy after CRT

Although there is a high rate of LR among incomplete responders in this population,⁶⁵ future studies are focusing on predicting patients who may potentially achieve a sustained cCR after neoadjuvant therapy, thereby avoiding surgery. The main challenge is selection of patients who can be considered for this approach. The evidence available comes mainly from retrospective data, which include patients with variable tumor characteristics and pretreatment clinical stage, many who have been evaluated with inaccurate and insufficient staging modalities. Habr-Gama et al⁹¹ published a 26.8% cCR rate in rectal cancer patients who received CRT, which included 69% pretreatment clinical stage T3 tumors. A retrospective study also published by Habr-Gama et al⁹¹ assessed the outcomes of a watch-and-wait strategy for patients with cT2N0 rectal cancer located <7 cm from the anal verge who received CRT. Patients were treated with 54 Gy and 6 cycles of 5-FU-based chemotherapy or 50.4 Gy with 2 cycles of 5-FU-based chemotherapy. Those treated with higher doses of RT were more likely to achieve a cCR (85.7 vs 56.6%, respectively; P < .001).⁹² These data support consideration of a treatment strategy of surveillance in patients with early-stage low cancers who achieve cCR. Other studies have demonstrated that nonoperative management of patients with cCR after CRT results in better anorectal function in comparison with patients with near-complete response managed by TEM.^93,94 A recent publication examined the potential of organ preservation with LE or active surveillance after CRT for 362 patients with non-metastatic cT3 or any stage N + locally advanced rectal cancers treated with neoadjuvant CRT evaluated for clinical response. Active watch-and-wait surveillance was offered to 10 patients who were found to have a cCR, and TAE was performed in 50 patients who were found to have an objective clinical response with residual ulcer measuring <3 cm. Of the 60 patients offered LE or active surveillance an 8.9% LR rate was observed. There was no significant difference in OS or DFS, however, when the outcomes of radical surgery were compared with LE.⁹⁵

A major challenge in selecting appropriate patients for the watch-and-wait strategy is the appropriate selection of patients suitable for this approach because not all patients who achieve cCR are found to have pCR. There is no consensus regarding the methods employed for reassessment because all are associated with limitations. Despite the selection of surveillance tools, an active surveillance protocol is necessary in all patients who are considered for the watch-and-wait approach, which usually includes clinical examination, monitoring of CEA level, periodic flexible sigmoidoscopy, or complete colonoscopy and imaging examinations. No clear recommendation as to the best surveillance program has been defined, but most agree that patients should be evaluated every 1 to 3 months for at least the first 2 years, during which time most tumor recurrences will occur.

Controversy also exists over the optimal interval between treatment completion and response assessment, which should ideally assess the greatest tumor regression, while considering the effect of time interval on salvage surgery, if necessary. Several retrospective studies have suggested a higher pCR rate when delaying surgery after neoadjuvant CRT.^96,97 Some studies performed reassessment at fixed time points,^95,98-101 while others assessed response at longer intervals (8-12 weeks).^102-104 Two NCDB studies and a meta-analysis designed to answer the question on optimal timing of surgery after neoadjuvant CRT in patients with pretreatment clinical state II and III rectal cancer suggested that waiting at least 8 weeks after completion of neoadjuvant CRT provided the best rate of pCR and downstaging without increasing morbidity.^105-107 One prospective study directly evaluated the effect of increasing the interval between the end of RCT and surgery on the pCR rate. The Effect of Interval (7 or 11 weeks) Between Neoadjuvant Radiochemotherapy and Surgery on Complete Pathologic Response In Rectal Cancer: A Multicenter, Randomized Controlled Trial (GRECCAR-6) study randomized 265 patients with cT3/T4 or Tany N + tumors of the mid or lower rectum who had received CRT (45-50 Gy with 5-FU or capecitabine) at 7 versus 11 week intervals before surgery. The primary endpoint of pCR rate (defined as a ypT0N0 specimen) was not statistically different between the 2 groups (7 weeks: 20 of 133, 15.0% vs 11weeks: 23 of 132, 17.4%; P = .5983). Morbidity was significantly increased in the 11 week group (44.5% vs 32%; P = .0404), which was also associated with worse quality of mesorectal resection (complete mesorectum [78.7% vs 90%; P = .0156]). This study concluded that waiting 11 weeks after CRT did not increase the rate of pCR and may be associated with higher morbidity and more difficult surgical resection.¹⁰⁸ Although these studies do not directly address the optimal time to assess clinical response after neoadjuvant treatment to determine whether salvage surgery is necessary, these data provide information regarding the effect of time interval on surgical complications and outcomes, which are important when considering the watch-and-wait approach.

Currently, there is no level I evidence to support a watch-and-wait approach in patients achieving cCR after CRT for rectal adenocarcinoma. The Rectal Sparing Approach after Preoperative Radio- and/or CHemotherapy (RESARCH) is a multicenter observational study to evaluate the effectiveness of rectum-sparing approaches at 2 years after the completion of neoadjuvant treatment (https://www.clinicaltrials.gov/ct2/show/NCT02710812). In this study, patients with rectal cancer eligible to receive neoadjuvant CRT will be prospectively enrolled and restaged 7 to 8 weeks after the completion. Those patients with major clinical response or cCR (defined as absence of mass, small mucosal irregularity ≤2 cm in diameter at endoscopy, and no metastatic nodes at MRI) will undergo LE, while patients with cCR will either undergo LE or a watch-and-wait policy. The primary endpoint is to determine rectum preservation rate at 2 years.¹⁰⁹ The Deferral of Surgery trial (https://clinicaltrials.gov/ct2/show/NCT01047969) is a prospective study of the watch-and-wait approach in rectal cancer patients using a controlled surveillance program. This study is designed to estimate the percentage of patients who can safely omit surgery, which is defined as the percentage of patients who achieved cCR at 2 years after completion of CRT who have not had surgery and who achieved cCR. This study uses digital rectal examination, CEA, MRI, and fluorodeoxyglucose-positron emission tomography (FDG-PET) at 8 weeks after completion of CRT to demonstrate absence of visible tumor, which defines cCR. Flexible sigmoidoscopy and colonoscopy are performed periodically starting at 6 months, and CR is performed annually. Patients are reevaluated with CEA, digital rectal examination, and MRI more frequently in the first 16 weeks after completion (every 4 weeks), every 3 months up to 2 years, every 6 months from 2 to 5 years, then annually thereafter. Of note, biopsies are only indicated if regrowth is suspected and MRI and Positron emission tomography—computed tomography (PET-CT) are performed to reduce the rate of false-positive findings on imaging.

Simulation, treatment technique, and radiation dose

Patients treated with 3-dimentional (3-D) conformal RT should be physically positioned at the time of simulation to displace the small bowel to minimize treatment toxicity, and small-bowel contrast can be used to assist in identification of small bowel for treatment planning purposes. The use of a belly board with the patient in prone position with a full bladder has been shown to reduce the volume of irradiated small bowel by approximately 70%.¹¹⁰ However, this position may be difficult for some patients to tolerate. Retrospective comparison of treatment in the prone versus supine position, with or without daily image guidance, demonstrates that prone positioning leads to a greater systematic error, whereas the supine position was associated with increased random error. The increased use of image guided radiation therapy was noted to decrease the setup error associated with supine positioning.¹¹¹ Another prospective study comparing treatment in the prone versus supine position demonstrated a higher volume of small bowel receiving lower doses (5 and 10 Gy) for the supine position, but there was no appreciable difference between supine and prone positioning in the volume of small bowel receiving higher doses (≥20 Gy).¹¹² For consistency regarding bladder filling, it is best to give the patient clear instructions (eg, first empty bladder and then drink 16 oz of water 1 hour before simulation and each subsequent treatment). Use of CT-based image guidance should be considered to ensure adequate bladder filling in addition to target coverage.

A 3-field or 4-field 3-D conformal treatment technique with prone setup using a belly board with or without full bladder to displace bowel from radiation field is an acceptable method of treatment. Likewise, 3-field or 4-field 3-D conformal radiation using a supine technique with careful attention to bowel in the field may be also acceptable in patients who cannot tolerate the prone position. Although a 3-field approach using posterior-anterior and lateral fields is preferred to best avoid small bowel, for larger patients a more lightly weighted fourth anterior field may be appropriate to improve RT plan dose homogeneity.

A total radiation dose of 50 to 56 Gy should be used for patients with T2 tumors or T1 tumors with high-risk features and margins of at least 3 mm. For patients with high-risk features and close margins, higher radiation doses may be considered. This recommendation is derived from RTOG 89 to 02, which included patients with rectal tumors ≤4 cm diameter and occupying 40% or less of the rectal circumference who underwent local excision. Patients with the most favorable risk profiles remained under observation. Patients with high-risk features but at least 3 mm surgical margins received adjuvant RT to a total dose of 50 to 56 Gy plus 2 cycles of 5-FU (1000 mg/m² over 96 hours). Patients with high-risk features and close or positive margins underwent similar treatment, but to a total RT dose of 59.4 to 65 Gy.¹⁷ CALGB 8984 included 110 patients with T1 or T2 rectal tumors who underwent local excision with pathologically negative margins. Patients with T1 tumors underwent observation while those with T2 tumors received adjuvant RT to 54 Gy with 2 cycles of 5-FU (500 mg/m² over 72 hours) resulting in similar failure-free survival.¹⁶

Conclusions

• The panel recommends that LE alone may be an acceptable treatment strategy for T1N0 rectal cancers without high-risk features associated with increased risk of recurrence.

• The panel strongly recommends adjuvant RT or CRT for patients who undergo LE for T1N0 rectal cancers and have known clinical or pathologic adverse risk factors.

• The panel recommends strongly that adjuvant or neoadjuvant therapy should be considered in patients treated with LE for T2N0 rectal cancers, which are associated with a higher risk of lymph node metastases.

• The panel recommends with reservation the use of MRI to assess tumor response in patients with early stage rectal cancers to define candidates appropriate for LE after neoadjuvant therapy.

• The panel does not recommend neoadjuvant therapy followed by restaging and LE for T3N0 rectal cancers with or near complete tumor response outside of a clinical trial. These patients are at very high risk for nodal involvement and LR, and therefore the panel strongly recommends that TME after neoadjuvant CRT is the standard of care for curative intent treatment for those able to undergo this procedure.

• The panel does not recommend neoadjuvant therapy followed by a watch-and-wait approach in patients achieving cCR after CRT outside of a clinical trial setting.

The ARS Appropriate Use Criteria and its expert panels have developed criteria for determining appropriate radiologic procedures for diagnosis and treatment of specified medical condition(s). Generally, the complexity and severity of a patient’s clinical condition should dictate the selection of appropriate imaging procedures or treatments. The availability of equipment or personnel may influence the selection of appropriate imaging procedures or treatments. Imaging techniques classified as investigational by the FDA have not been considered in developing these criteria; however, study of new equipment and applications should be encouraged. The ultimate decision regarding the appropriateness of any specific radiologic examination or treatment must be made by the referring physician and treating radiation oncologist in light of all the circumstances presented in an individual examination.