Comparison of guidelines for the management of rectal cancer

A comparison between NCCN, ESMO and JSCCR Guidelines is presented, concerning the treatment of rectal cancer, with an analysis and discussion of their discrepancies. Differences indicate areas for research


Introduction
The 5-year survival rate for rectal cancer approaches 60 per cent 1 . Although around a quarter of all patients present with metastases at initial diagnosis and half of all patients develop metastases 2 , the median overall survival for patients with metastatic disease is around 30 months, more than double that of 20 years ago. Factors that may have contributed to this improvement include closer follow-up and earlier detection of recurrent disease, efficacy of systemic therapies, resection of metastases and implementation of a 'continuum of care' 3 .
To standardize care, several organizations have created guidelines and protocols regarding the management of rectal cancer. Despite this, there remains a certain degree of variation among guidelines.
The aim of this study was to compare the guidelines for the management of rectal cancer from the European Society of Medical Oncology (ESMO, Europe) 1 -3 , the National Comprehensive Cancer Network (NCCN, USA) 4 and the Japanese Society for Cancer of the Cancer of Colon and Rectum (JSCCR, Japan) 5 , highlighting and analysing both agreements and discrepancies, and considering further developments. These guidelines were chosen because of the high incidence of rectal cancer in those countries and the lack of previous collaborative guidelines among those societies.

Methods
NCCN, ESMO and JSCCR guidelines were included, with the relative position papers: NCCN Guidelines for Rectal Cancer version 3.2017 4 ; ESMO Clinical Practice Guidelines for Diagnosis, Treatment and Follow-up of rectal cancer (2013 edition) 1 and metastatic colorectal cancer (2014 edition) 2 ; ESMO consensus guidelines for the management of patients with metastatic colorectal cancer (2016 edition) 3 ; and JSCCR guidelines 2016 for the treatment of colorectal cancer 5 . Guidelines were accessed at organizational websites or collected from available publications. Access was sought for guidelines that were not open-access.
Outcome measures were agreements and discrepancies among NCCN, ESMO and JSCCR for each specific topic of rectal cancer management, and results are presented in sections.

Results
Analyses included four journal articles 1 -3,5 and one guideline document 4 . Both NCCN and ESMO guidelines discussed rectal cancer as a specific entity, with the exception of metastatic disease, which was considered in conjunction with colonic cancer. JSCCR guidelines presented colonic and rectal cancer without distinction.
There was some concordance between Western and Asian guidelines for some topics, such as the definition of total mesorectal excision (TME) as surgical standard, administration of adjuvant therapy for stage III disease, surgical resection for metastases and/or recurrent disease, and cytoreductive surgery followed by intraperitoneal chemotherapy for peritoneal carcinosis.

Pedunculated or sessile polyp (adenoma) with invasive cancer
Guidelines from all three societies indicated the need for pathology review. NCCN advises simple observation after adequate removal of pedunculated polyps; this approach is also considered for sessile polyps removed completely. Transabdominal resection or transanal excision (standing on specific criteria) is recommended in the case of fragmented specimens, unassessable margins and/or unfavourable histological features.
According to ESMO, the choice between local procedure or radical standard surgery with or without adjuvant treatment is based on Haggitt's level or Kikuchi's system 6,7 , grading and vascular invasion.
JSCCR and ESMO recommend endoscopic management of pedunculated and sessile polyps (endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD) according to size), preferably located distally to the peritoneal reflection; or TME, considering depth of invasion, histology and budding ( Table 1).

Rectal cancer (appropriate for resection)
Both NCCN and ESMO state a formal pretreatment work-up, whereas JSCCR does not. NCCN recommends a complete blood count, chemistry profile, carcinoembryonic antigen (CEA) measurement and colonoscopy, whereas ESMO advises sigmoidoscopy (either rigid or flexible), endoscopic ultrasonography and circumferential resection margin evaluation ( Table 1). Both NCCN and ESMO advise the use of MRI in patient work-up. NCCN considers transanal excision only for T1 N0 non-fixed tumours (stage I) that are less than 3 cm in size, occupy less than 30 per cent of the circumference of the bowel and lie within 8 cm of the anal verge; advises transabdominal surgery without neoadjuvant therapy for T1-2 N0 tumours (stage I) that do not meet the previous criteria; and recommends neoadjuvant treatment, followed by surgery after 5-12 weeks, then adjuvant therapy, for all resectable disease from stage II to IV. In contrast, ESMO suggests surgery alone even for cT3a N1 high tumours, leaving postoperative treatment for those with poor prognostic features.
In consideration of local invasion (T3 mesorectal fascia (mrf) +/−, T4a-b, extramural vascular invasion (EMVI) +), both NCCN and ESMO agree on neoadjuvant treatment followed by surgery and then postoperative therapy. ESMO also suggests a 'deferred surgery' policy for patients at high surgical risk. NCCN advises palliative surgery in the setting of symptomatic patients with unresectable metastatic disease, whereas neither ESMO nor JSCCR describes this event.
The same criteria for TME are shared between NCCN and ESMO (mesorectal margin of 5 cm from the tumour distal edge), whereas JSCCR focuses more on the extent of lymph node dissection, which is based on the perceived spread of lymph node metastases and depth of tumour invasion (from D0-1 for pTis to D3 for cT3-4a/b or cN+). JSCCR also recommends lateral lymph node dissection for tumours where the lower border lies distal to the peritoneal reflection or there is invasion beyond the muscularis propria.
Conversely, NCCN does not recommend extended lymph node dissection unless suspicious nodes are present, whereas ESMO highlights the importance of radical resection of mesorectal fat, including all lymph nodes ( Table 1). cT2 very low, cT3 mrf − (unless cT3a (b) and mid or high rectum), N1-2, EMVI+, limited cT4a N0: preoperative RT or CRT followed by TME (wait-and-see in high-risk patients for surgery if CRT and clinical complete remission ‡ ‡) cT3mrf+, cT4a,b, lateral node+: preoperative CRT followed by surgery (TME + more extended surgery if needed) (RT with surgery delay in elderly or in patients with severe co-morbidity) TME or tumour-specific mesorectal excision (TSME) Lateral lymph node dissection is indicated when the lower border of the tumour is located distal to the peritoneal reflection and the tumour has invaded beyond the muscularis propria Additional comments **Surgery should be 5-12 weeks after full-dose 5⋅5-week neoadjuvant chemoradiotherapy † TME should extend 4-5 cm below distal edge of tumours; in distal rectal cancers (< 5 cm from anal verge) negative distal bowel margin of 1-2 cm may be acceptable † Extended lymph node resection is not indicated in the absence of clinically suspected nodes † † †For tumours situated in the upper third, partial mesorectal excision can be carried out with a mesorectal margin of ≥ 5 cm distal to the tumour TME implies that all mesorectal fat, including all lymph nodes, should be excised A good TME without damaging the rectal fascia surrounding the mesorectal fat and rectum is prognostically relevant If an abdominoperineal excision is planned, the dissection from above must be stopped at the tip of the coccyx and be continued from below ‡ ‡If no tumour can be detected and/or no viable tumour cells are found (i.e. a clinical or a pathological complete response is achieved), no further therapy is provided (organ preservation) and the patient is monitored closely for at least 5 years

Postoperative (adjuvant) treatment after surgery
There is substantial agreement between guidelines on adjuvant chemotherapy administration for stage III disease. Before chemotherapy, all three guidelines agree on gene testing of RAS status. Both NCCN and ESMO also recommend assessment for BRAF V600E mutation. ESMO alone advises microsatellite instability testing in the setting of metastatic disease, and JSCCR is the only one to advise UDP-glucuronosyltransferase1A1 (UGT1A1) phenotyping. Mismatch repair and dihydropyrimidine dehydrogenase deficiency testing are of secondary importance for NCCN and ESMO respectively.
Only NCCN recommends chemoradiotherapy (as an alternative to transabdominal resection) for T1 NX tumours with high-risk features and for T2 NX found after transanal excision.
For stage II disease, both NCCN and ESMO recommend postoperative chemotherapy and/or radiotherapy. Although specific regimens have not been stated, ESMO follows precise principles of administration in these patients and has recently questioned its routine use for pT3 N0 tumours. According to JSCCR, the usefulness of adjuvant chemotherapy has not been proven for stage II rectal cancer. Furthermore, NCCN recommends systemic treatment when there are contraindications to surgery, along with possible observation.
For stage III disease, all three guidelines recommend adjuvant treatment. NCCN and JSCCR agree on most of the standard regimens and on the length of the administration period (6 months preferred). NCCN suggests that there is no benefit for patients older than 70 years from oxaliplatin, whereas JSCCR advises adjuvant chemotherapy to patients aged 70 years and above with good performance status, adequate organ function and no postoperative complications ( Table 2).

Postoperative surveillance (follow-up)
Both NCCN and JSCCR recommend a precise, systematic follow-up, whereas ESMO advises a more flexible, patient-tailored one. NCCN divides follow-up between early and more advanced stages (II-III and IV). Complete colonoscopy at 1 year is suggested for all stages, to be repeated at 3 years and then every 5 years, unless advanced adenoma (villous  polyp, size greater than 1 cm and/or high-grade dysplasia) is found. For patients treated by transanal excision, proctoscopy with endoscopic ultrasonography or MRI with contrast is advised every 3-6 months for the first 2 years, then every 6 months for up to 5 years. Both ESMO and JSCCR divide recommendations between stage I-III and stage IV disease. Although the plan for stage I-III disease is not well defined, it broadly consists of clinical assessment every 6 months for 2 years and history and colonoscopy every 5 years. ESMO advises intensive follow-up for patients with stage IV disease, with evaluation (history, physical examination, CEA, CT and/or MRI) every 2-3 months, especially if chemotherapy is undertaken.
The JSCCR-recommended follow-up schedule is more frequent than that proposed by NCCN, at least for the first years of surveillance, with annual colonoscopy and examination with tumour markers every 3 months for the first 3 years, along with CT every 6 months for the first 3 years (or 5 years in stage III disease). JSCCR alone suggests digital rectal examination every 6 months for 3 years, and establishes the duration of the postoperative surveillance as 5 years, on the basis that 95 per cent of recurrences are detected within 5 years of surgery.
None of the three guidelines gives indications for PET ( Table 3).

Synchronous metastases/locally invasive disease
All three guidelines recommend curative resection of the primary tumour and distant metastases whenever possible ( Table 4). NCCN distributes patients into three categories: those with resectable metastases, unresectable metastases or local invasion. For patients with resectable metastases, there are two pathways, both characterized by neoadjuvant treatment, followed by surgery and then adjuvant therapy. Surgery is intended as staged or synchronous resection and/or local therapy for metastases and resection of the primary. The choice for preoperative and postoperative therapy is 5-fluorouracil or capecitabine in conjunction with radiotherapy or combination chemotherapy (FOLFOX (5-fluorouracil-leucovorin-oxaliplatin), CAPEOX (capecitabine-oxaliplatin)) or systemic therapy     In selected patients, complete cytoreductive surgery and HIPEC may provide prolonged survival when carried out in experienced high-volume centres The efficacy depends on the extent of peritoneal dissemination and is scored using the PCI, which is the main prognostic factor Involvement of the lower ileum is a negative prognostic factor Cytoreductive surgery is particularly effective in patients with low-volume peritoneal disease (PCI < 12) and no evidence of systemic disease If the resection is not significantly invasive, the peritoneal dissemination should be resected at the same time as the primary tumour  (combination chemotherapy plus biological targeted agents). For patients with unresectable metastases, the presence of symptoms is an indication for palliative surgery or stenting, whereas systemic therapy represents the backbone of treatment. In consideration of local invasion, there are again two pathways, distinguished by the timing of combination chemotherapy ( Table 4).
ESMO divides patients into two groups (historical and revised) for treatment stratification. In the former, for group 0 patients (R0 resection technically achievable), radical surgery and perioperative chemotherapy are suggested. Surgery is advised in the same way for group 1 patients (potentially resectable disease with curative intention or oligometastatic disease), after neoadjuvant chemoradiotherapy followed by adjuvant chemotherapy. Conversely, for group 2 patients (disseminated disease, unlikely to be resectable) active first-line treatment is recommended to induce regression, followed by assessment every 2-3 months. Chemotherapy with the intention of prolonging life and preventing tumour progression is indicated for group 3 patients (never resectable disease). Considering the revised groups, there are fit patients in group 1 for whom radical treatment is the goal, fit patients in group 2 eligible for disease control and prolonged survival, and unfit patients who might be referred for palliative and/or best supportive care. ESMO is the only guideline to state a scheduled continuum of care and to list technical and oncological contraindications to hepatic resection.
JSCCR discourages resection of distant metastases if the primary tumour is unresectable, whereas it suggests resection of the primary in patients with unresectable metastases, based on clinical symptoms and prognostic impact. Furthermore, it establishes criteria for both hepatectomy  and pneumonectomy, as well as stereotactic body radiotherapy. JSCCR also provides general indications for systemic chemotherapy ( Table 4).

Metachronous metastases/local recurrence
All three guidelines agree that surgical resection of metachronous metastases or local recurrence should be performed whenever possible. NCCN guidelines describe three different scenarios: resectable metachronous metastases, unresectable metachronous metastases and pelvic/anastomotic recurrence (local recurrence) ( Table 4).
There are two possible pathways for resectable metachronous metastases based on previous administration of chemotherapy. If chemotherapy has already been given, resection or local therapy followed by adjuvant combination treatment is suggested. Alternatively, neoadjuvant combination therapy followed by resection or local treatment can be offered. In case of metastatic growth during neoadjuvant treatment, a switch to systemic or biological therapy is advised. Where chemotherapy has been used previously, recommendations are similar, although systemic/biological therapy and observation are given more prominence.
The treatment strategy for unresectable metachronous metastases depends on previous adjuvant therapy. If FOL-FOX or CAPEOX has been given during the past year, treatment with FOLFIRI with or without a biological agent is advised, with the aim of conversion to resectable disease (re-evaluation every 2 months) for radical surgery. If that goal is not achieved, systemic therapy is recommended. If no previous chemotherapy has been administered, systemic therapy with the aim of conversion to resectable disease is advised in the same way.
Surgery is recommended for pelvic/anastomotic recurrence if the local recurrence is amenable to resection, possibly associated with intraoperative radiotherapy (IORT), and preceded or followed by chemoradiotherapy.
ESMO guidelines do not state specific treatment for metachronous disease ( Table 4). Despite that, radiotherapy is strongly suggested for local recurrence, and IORT and brachytherapy are also advised. Radical surgery is recommended 6-10 weeks after radiotherapy.
Similarly, JSCCR recommendations are brief: resection should be considered for local recurrence and distant metachronous metastases whenever R0 resection is considered obtainable ( Table 4).

Peritoneal disease
All three guidelines agree on the principle that complete cytoreductive surgery and intraperitoneal chemotherapy should be considered in experienced centres whenever R0 resection can be achieved. ESMO underlines the importance of the peritoneal cancer index (PCI) for prognosis, and regards involvement of the distal ileum as a negative prognostic factor ( Table 4).

Minimally invasive surgery
Laparoscopic surgical approaches are formally stated only by NCCN and JSCCR, with agreement on the importance of experience and technical skills. NCCN cites locally advanced disease, with a threatened or high-risk circumferential margin based on staging, and emergency setting, such as acute bowel obstruction or perforation, as contraindications. JSCCR underlines the importance of tumour location, obesity and previous open abdominal surgery.
None of the guidelines mentions robotic surgery ( Table 5).

Restaging
NCCN and ESMO agree on re-evaluation every 2 months in patients receiving chemotherapy with the aim of conversion to resectability. ESMO holds that maximum response is expected to be achieved after 12-16 weeks of therapy. ESMO alone envisages a complete radiological response on MRI and/or PET-CT, although the relevance of this remains unclear. JSCCR does not formally state the restaging. None of the guidelines mentions tumour regression grading (TRG) classification ( Table 5).

Watch-and-wait policy
ESMO considers a watch-and-wait policy for complete response after chemoradiotherapy and suggests strict monitoring for a minimum 5 years, without providing supporting evidence. NCCN does not support a watch-and-wait policy in the routine management of localized disease, and JSCCR does not describe this approach ( Table 5).

Surgery beyond total mesorectal excision
All guidelines agree that TME should be viewed as the standard for surgery, but none describes more advanced procedures such as extralevator abdominoperineal excision or total pelvic exenteration ( Table 5).

Discussion
There were areas of general agreement among American, European and Japanese guidelines in the management of rectal cancer. Due to the qualitative rather than quantitative nature of this study, no attempt was made to score concordance between the analysed guidelines. The aim was to highlight the main points of each guideline, in order to promote cooperations that might explain differences and to see whether these might be resolved.
Guidelines are designed to provide up-to-date recommendations to enable surgeons to deliver best practice. Guidelines should, consequently, analyse key points 8 . There were some key issues where there were clear differences. Regarding polyps with invasive cancer, JSCCR promoted the role of advanced endoscopic techniques (ESD and EMR) that were less considered by Western guidelines. In particular, ESD, although considered a safe and effective procedure 9 -11 , has yet to achieve widespread adoption in Western countries mainly due to its long learning curve and poorly defined training processes 12 . In terms of pathology review of polyps, JSCCR alone considered tumour budding among indications for radical surgery, on the basis of evidence indicating its value as a prognostic factor 13 -15 . Regarding radical surgery, Japanese guidelines advocated lateral lymph node (or pelvic side wall) dissection whenever the lower tumour border was located distal to the peritoneal reflection 16 . Despite evidence, this remains controversial, with concerns about complications and lack of additional benefit 17,18 . The wider use of chemoradiotherapy in the West probably explains the guideline differences 19,20 .
Regarding postoperative chemo/radiotherapy, NCCN supported this indication for T1-2 tumours after transanal excision, mainly as an alternative to radical surgery 21 -23 . In the same way, only NCCN guidelines advised adjuvant treatment for stage II cancer. Although results in terms of overall and disease-free survival may be impressive 24 , the therapeutic decision should be taken carefully by a multidisciplinary team, balancing possible benefits and drawbacks 25 . Interestingly, neoadjuvant chemotherapy was not described by JSCCR guidelines, even though its positive outcomes have been widely demonstrated 26 -28 . For surveillance protocols, it was noteworthy that only NCCN suggested just follow-up colonoscopy for stage I patients with full surgical staging, in accordance with other American guidelines 29,30 .
With respect to the management of metastatic disease, it was interesting to note that European guidelines did not specifically assess the scenario of distant recurrent metastases, possibly reflecting the similar outcome of synchronous and metachronous disease 31,32 . All three guidelines agreed that for metastatic disease, local recurrence or locally advanced disease, surgery should be attempted whenever possible.
Despite TME being regarded as the standard procedure, none of the guidelines mentioned more 'aggressive' surgery, as suggested by the Beyond TME Collaborative 33 , nor robotic approaches, in spite of relatively widespread adoption 34,35 .
The issue of restaging in the context of patients thought initially to have unresectable metastatic disease, who have apparent responses to induction therapy, was covered only in the Western guidelines. The concepts of complete radiological response and TRG were not taken into consideration by the guidelines, despite available evidence 36,37 .
Although well described, non-operative management or deferred surgery 38,39 , also referred to as a watch-and-wait approach, was described as a possible therapeutic option only in the ESMO guidelines; American and Japanese documents failed to mention it.
Differences between the guidelines existed, potentially reflecting the frequency with which different clinical patterns of disease present in different parts of the world. Political, economic and social contexts are also likely to be influential. Despite these considerations, specific discrepancies relating to preoperative work-up, management of early disease, extended lymph node dissection, adjuvant treatment for early stages, and neoadjuvant therapy make these logical topics where future research could be directed profitably.