Is Non-operative Management of Rectal Cancer Feasible?

Introduction

Historically, the standard of care for locally advanced rectal cancer (LARC) includes neoadjuvant long-course chemoradiotherapy (LCRT), total mesorectal excision (TME) and postoperative adjuvant chemotherapy¹. However, the treatment algorithm for LARC has become more complex trying to balance better rates of survival and reducing the late sequelae on the patient’s quality of life (QoL) due to the toxicities associated with the treatment components².

It is well established that a group of LARC patients that undergo neoadjuvant therapy (NAT) experienced the complete absence of residual tumor cells at either the primary tumor site and the mesorectal nodes, a phenomenon known as pathological complete response (pCR)³. Approximately 20% of LARC patients treated with LCRT alone may develop a pCR, but this can be as high as 40% in those treated with concomitant consolidative chemotherapy given after LCRT⁴. Patients with a pCR have improved survival rates, with lower than 5% of distant metastasis and 1% of locoregional failure⁵.

Given higher rates of tumor response with some NAT modalities, surgeons rise the question about the need of TME in patients who achieve a pCR, especially when surgery is associated with late bowel, sexual and genitourinary sequelae that significantly impairs the QoL of patients^6,7. Nonoperative management (NOM) for rectal cancer may be a potential treatment option for selected LARC patients avoiding radical surgery and its late dysfunctions.

The aim of the present review is to discuss the principles about NOM for rectal cancer patients and the main considerations used for selection of potential candidates.

Overview of Neoadjuvant Therapy for Rectal Cancer and Treatment Response

In the classic paradigm of LARC treatment, two neoadjuvant modalities have been used to improve local control: long course chemoradiation and short course radiotherapy (SCRT). A LCRT strategy was described as radiotherapy (45 to 56 Gy) administered during a 5-to-6-week period with concomitant sensitizing chemotherapeutic agents and a 6-to-10-week period of rest before TME for tumor regression. This strategy offers advantages such as tumor-free surgical margins and higher rates of sphincter preservation surgery in low rectal tumors⁸. Additionally, multiple studies support the use of LCRT based on a reduction in the local recurrence rate^8–10 and the possibility to identify good responders^11,12. Short-course radiotherapy (SCRT) consisting of 25Gy in 5 doses (5×5 Gy) in addition to TME in the following 7 days, has shown a significative reduction in local relapse in several phase III trials^13–15.

There is no differences between both strategies in terms of overall survival, local recurrence and surgical complications¹⁶. However, due to the shorter time interval between RT and surgery, SCRT has been associated with lower tumor response rates. Despite all of this, none of these strategies have demonstrated better overall survival rates. Systemic recurrence remains the main issue facing LARC patients with 25% developing distant metastasis during follow up^17–19. In consideration of these findings, the addition of systemic chemotherapy as a part of the NAT strategy has been proposed to diminish the risk of distant metastasis.

The concept of Total Neoadjuvant Therapy (TNT) implies the use of either SCRT or LCRT and moving the adjuvant dose of systemic chemotherapy into the realm of NA treatment^20–22. TNT may diminish the risk of distant metastasis and enhance the rate of pCR²¹, allowing the potential for organ preservation in selected patients.

One of the most important predictors of long-term oncological outcomes in rectal cancer is the degree of tumor response after NAT²³. Pathological staging of the surgical specimen post-NAT is one of the most reliable predictor of oncological outcome, even more than the baseline clinical staging²³. In patients with persistence of macroscopic tumor post-NAT, also known as incomplete response (IR), TME is the only treatment of choice.

In the recent years, focus has been put on clinical complete response (cCR), defined as the absence of detectable macroscopic tumor by clinical means. However, cCR does not correlate completely with histological pCR, with higher rates of patients achieving a cCR after NAT²⁴.

Watch and wait (WW) and NOM are used interchangeably to describe an organ preservation strategy for selected patients that experienced a cCR after NAT. Despite the benefits of a WW strategy, some surgeons are reluctant to widely adopt it²⁵ due the lack of standardization in response assessment criteria and the absence of level I evidence. While stricter criteria may increase the accuracy of individual patients selected for the WW strategy, more liberal criteria may risk worsening oncological outcomes. Additionally, there is an intermediate group of patients with near cCR who demonstrate a significant tumor regression, but who fall short of achieving a cCR (the tumor never truly evolves to a cCR)²⁶.

Standardized definitions of cCR and pCR exist and have been utilized in both case series and trials, but the current challenge is to correctly select patients with a sustained cCR, based on clinical assessment, who would be found to have a pCR if they were to undergo resection. The critical point to make NOM feasible is for clinicians to correctly identify those would-be candidates for a WW strategy without compromising oncologic safety.

Considerations for Non-operative Management Approach and Patient Selection

Organ preservation is currently a valid treatment strategy for patients, who are willing to accept potentially worse oncological outcomes to achieve this objetive. Gani et al²⁷ reported that 83% of patients would consider a WW strategy if they achieved a cCR in spite of the potentially high rate of local regrowth during the first 2 years of follow-up. More interestingly, up to 30% of patients are willing to sacrifice rates of long-term oncological cure in comparison with clinicians²⁸.

To correctly assess candidates, the initial patient evaluation should be the standard rectal cancer work-up based on the guidelines of the National Comprehensive Cancer Network¹. It is important to have an endoscopic baseline evaluation to correctly assess the subsequent response to NAT.

Once NAT is finished, patients with a cCR or a near cCR may be offered to enter a WW protocol (Figure 1) if the patient, surgeon, and the disease management team agree upon this approach. Our strong preference is that a WW approach is completed in the context of a clinical trial if possible. Typical WW candidates usually are patients with distal rectal adenocarcinomas considered for abdominoperineal resections preoperatively or very low stapled/handsewn colorectal/coloanal anastomoses which may significantly alter their QoL²⁹.

Figure 1:

Memorial Sloan Kettering three-tiered response/regression schema

Pre-treatment characteristics of the tumor have demonstrated high-risk of local recurrence (i.e. <1mm circumferential margin, extramural venous invasion and extensive mesorectal/pelvic lymph nodes involved) and are associated with lower rates of cCR^30,31. However, patients with node-positive LARC at baseline that develop a cCR after NAT are not at increased risk for local tumor regrowth or development of more advanced disease at the time of recurrence³². The presence of tumor ulceration and complete circumferential involvement may be a relative contraindication due the fact that scarring may potentially develop and distort the lumen creating a rectal stenosis and not allowing an adequate endoscopic follow up – however this is not an absolute contraindication to a WW approach in our experience as a number of patients have been watched with stable scar for years. Metanalysis have found that older age, smaller tumor, shorter distance from the anal verge, and negative lymph node status are associated with higher rates of pCR³³. From a genetic perspective, mutations in p53 and KRAS, found in about 70% and 40% of rectal tumors, respectively^34,35, were associated with poor response to NAT; however, recent work has shown that only KRAS alterations seem to be associated with faster rates of relapse but that no somatic alteration was associated with rate of response to NAT³⁶. Contrarily, mismatch repair deficiency tumors are associated with better rates of response to NAT specifically to immunotherapy in place of cytotoxic chemotherapy³⁷.

The current challenge in WW is the correct selection of cCR patients. Ideally, patients with a pCR should be identified as cCR before undergoing TME to offer WW. However, the correlation between cCR and pCR is not perfect. Up to 25–30% of patients with a cCR experienced a regrowth of the primary tumor³⁸, most of them during the first 2 years of follow up. Conversely, up to 15% of patients with an cCR end up having a pCR^39,40 in the pathology report – in the recently completed OPRA trial the rates were 8% and 9% for those in the induction TNT group versus those in the consolidation TNT group respectively. Three modalities are considered to assess response after NAT: digital rectal exam, magnetic resonance imaging (MRI), and endoscopy. All 3 modalities combined report an accuracy of 98% to predict absence of tumor⁴⁰. Digital rectal exam of a cCR should be normal, but some minor mucosal abnormalities or soft scar can be palpated. Classic endoscopic features of cCR include a flat white scar, telangiectasia, and absence of both ulcer and nodularity⁴¹. MRI features of cCR include a scar not thicker than the rectal wall, only dark T2 signal, no visible lymph nodes, no restricted diffusion, and lack of or low signal on apparent diffusion coefficient (ADC) map.

One of the main problems in the current implementation of WW is the lack of uniform and reproducible criteria for tumor response and patient selection. To improve the response assessment, the Memorial Sloan Kettering three-tiered response/regression schema has been devised and was tested prospectively in the OPRA trial with excellent results⁴² (Figure 1).

In addition, other tools have been tested in the search for better ways of assessing the response to NAT. A systematic review showed that dynamic contrast enhanced (DCE) MRI can identify tumors that exhibit a high pre-NA therapy K_trans (representing the rate at which the contrast agent transfers from the blood to the interstitium) along with a subsequent decrease in K_trans. Both findings appear to be predictors of a favorable response to NAT.⁴³ A recent report using 2,252 features from patient-based imaging collected both pre- and post-NAT, showed good discrimination of pCR when used in combination with tumor length (AUC 0.9756 95% CI, 0.9185–0.97)⁴⁴.

The use of ctDNA is not well-explored in this setting and there are no prospective data using ctDNA to make treatment decisions about WW, but exploratory analyses from OPRA and the next generation of WW trials could provide some insight into ctDNA clearance and sustained cCR.

Is it possible to Optimize Tumor Response?

Improving rates of response to NAT may be associated with higher rates of patients that may undergo WW. Potential strategies that can be used alone or in combination, includes (1) dose escalation of preoperative radiation therapy, (2) adding systemic chemotherapy to radiotherapy in the neoadjuvant setting, (3) increasing the interval period after NAT and surgery (4) administering consolidation chemotherapy after LCRT or SCRT rather than induction chemotherapy followed by LCRT.

Dose escalation strategies have utilized using different approaches such as external beam radiotherapy, brachytherapy, contact radiotherapy and proton/iron beam radiotherapy. A report of nearly 3300 patients showed that the dose of NAT was a predictor of pCR⁴⁵. Additionally, neoadjuvant brachytherapy or external radiation therapy boost did not add a significant benefit in survival or local recurrence when given after standard NA treatment and TME^46,47. Currently, it is unknown whether a dose escalation approach will improve the rate of organ preservation.

A single-institution phase II trial with 32 patients⁴⁸ demonstrated the potential selective elimination of preoperative LCRT in patients with LARC using only 6 cycles of FOLFOX plus bevacizumab. Additionally, the Timing of Rectal Cancer Response to Chemoradiation Consortium trial showed a 38% rate of pCR if 6 cycles of FOLFOX were given after LCRT in an extended interval period compared with an 18% pCR rate in the LCRT alone group⁴⁹. Using a chemo-intensification approach (triplet vs. doublet chemotherapy), the PRODIGE 23 trial⁵⁰ randomized patients with LARC to either six cycles of neoadjuvant mFOLFIRINOX followed by LCRT and surgery and 3 months of adjuvant chemotherapy or standard of care (LCRT + surgery + adjuvant FOLFOX). Patients in the mFOLFIRINOX arm had a higher rate of pathological complete response (27.5% vs 11.7%, p-value<0.001) and better 3-year rates of disease-free survival (75.7% vs 68–5%, p-value=0.034). Currently, the JANUS trial (NCT05610163) is enrolling stage II-III rectal cancer patients in order to compare the use of consolidation FOLFIRINOX after LCRT as a strategy to enhanced clinical complete response⁵¹.

Higher rates of pCR are associated with longer delay interval between NAT and surgery. The Stockholm III trial showed that waiting 4 to 8 weeks after SCRT was associated with similar tumor regression rates compared to LCRT⁵². An analysis of a National Cancer Database that included all stage II and III rectal cancer patients undergoing LCRT suggested that any surgery interval longer than 8 weeks had higher odds of pCR (odd ratio 1.12, 95% CI 1.01 to 1.25)⁵³.

Finally, it seems that using a consolidation chemotherapy based-TNT approach is associated with higher rates of response and organ preservation than using induction chemotherapy. The phase II German trial CAO/ARO/AIO-12 compared 4 cycles of FOLFOX before (induction) or after (consolidation) LCRT. They reported 17% (induction) and 25% (consolidation) pathological complete response rates, respectively⁵⁴. Long term follow-up of this cohort showed similar results in favor of LCRT followed by consolidation chemotherapy as the preferred TNT sequence, especially if organ preservation is a priority⁵⁵. Although no WW was offered, patients in the experimental arm the RAPIDO Trial (SCRT followed by consolidation systemic chemotherapy) had higher rate of pCR (27.7% vs 13.8%, p-value<0,001)⁵⁶. The OPRA trial, in which patients were randomized to induction or consolidation TNT and then proceeded to surgery or WW depending on response, showed higher rates of organ preservation in the consolidation arm (58% vs 43%; P 0.01), with no difference in disease-free survival or distant-metastasis-free survival⁵⁷. These data suggest that higher rates of patients managed with WW may be achieved if consolidation chemotherapy is used over induction chemotherapy followed by LCRT. There is scarce evidence about the use of SCRT and consolidation chemotherapy as a WW strategy. A non-randomized single arm trial demonstrated that SCRT followed by 8 cycles of FOLFOX was associated with 68% a cCR at 1 year follow up⁵⁸. Building off OPRA and RAPIDO, Rödel et al. have an ongoing trial (ACO/ARO/AIO-18.1) which randomizes SCRT versus LCRT each followed by consolidative chemotherapy and is using cCR and organ preservation rates as an endpoint⁵⁹.

A summary of the most relevant WW studies is presented in Table 1.

Table 1:

Summary of important WW studies for locally advanced rectal cancer patients treated with neoadjuvant therapy

Study	n	NAT Strategy	Regrowth, n(%)	Salvage therapy, n(%)	Overall Survival, %
Habr-Gama el al, 2004	71	LCRT	2 (3%)	2 (100%)	OS: 100%, DFS: 92%
Smith et al, 2012	32	LCRT	6 (18.8%)	6 (100%)	OS: 96% DFS: 88%
Habr-Gama el al1, 2014	90	LCRT	28 (31%)	26 (92.8%)	OS: 91% DFS: 68%
Appelt et al, 2015	40	LCRT	10 (25.9%)	9 (90%)	OS: 100% DFS: 70%
Lai et al, 2016	18	LCRT	2(11%)	2 (100%)	OS: 100%
Martens et al, 2016	100	LRCT: 95% SCRT: 5%	15(15%)	13 (87%)	OS: 96.6% DFS: 80.6%
OnCore Project, 2016	129	45 Gy w/ 5-FU	44(34%)	36 (81.8%)	OS: 96% DFS:88%
IWWD Consortium, 2019	880	LCRT: 91%	222 (25.2%)	141 (69%)	OS: 85% DFS: 94%
Smith et al, 2019	113	LCRT: 31(27%) Induction: 47(42%) Consolidation: 33 (29%) Chemotherapy alone: 2(2%)	22 (19.5%)	22 (100%)	OS: 85% DFS: 94%
Jimenez-Rodriguez et al, 2021	33	Induction TNT (FOLFOX)	2 (6%)	2(100%)	OS:97%, DFS:94%
Garcia-Aguilar et al (OPRA Trial), 2022	225 Induction group: 105/146 Consolidation group: 120/158	TNT Induction Chemotherapy + LCRT: 146 LCRT + consolidation chemotherapy: 158	75 Induction: 42/105 Consolidation: 33/120	62/75	DFS: 78% (Induction) vs 77% (Consolidation)
Fokas et al (ACO/ARO/AIO 18.1), TBD	702 (Estimated Enrollment)	TNT (SCRT + Consolidation chemotherapy vs LCRT + Consolidation chemotherapy)	N/A	N/A	N/A
Smith et al (JANUS Trial) TBD	312 (Estimated Enrollment)	TNT (LCRT + consolidation mFOLFOX6 vs LCRT + consolidation mFOLFIRINOX)	N/A	N/A	N/A

SCRT: Short Course Radiotherapy; LCRT: Long Course Radiotherapy; OS: Overall Survival, DFS: Disease Free survival, IWWD: International Watch and Wait Database, TNT: Total Neoadjuvant Therapy

Non-Operative Management Controversies

One of the mayor uncertainties of WW is the long-term oncologic results²³. Systematic reviews showed a local regrowth rate between 15.7 to 30% and that surgical salvage was feasible in almost 93–95.4%% of the cases^60,61. The International Watch and Wait Database (IWWD)³⁸ reported a 2-year cumulative incidence of local regrowth of 25.2%, with all of the cases diagnosed within the first 2 years of surveillance.

An important concern in patients with apparent cCR, is a potential higher rate of distant metastasis after tumor regrowth. Data from retrospective evaluation of a 10-year experience at Memorial hospital suggest a higher rate of distant metastases in patients with local regrowth when compared to those without local regrowth⁶². Data from the IWWD suggests the same pattern between local regrowth and distant metastases³⁸. Recently, Jimenez-Rodriguez et al reported a lower rate of local regrowth (6%) in patients treated with TNT by a single surgeon that actively performed WW⁶³. Interestingly, only half of the patients that experienced local regrowth developed distant metastasis. One of the reasons that may explained this result is the strict surveillance commitment of patient and surgeon using a standardize method to evaluate response (MSK Regression Schema). Another publication from the IWWD showed that development of local regrowth at any time is a risk factor for distant metastases, with this risk remain higher for 5 years after development of local regrowth⁶⁴. Still, whether removing the primary tumor after completion of NAT would have mitigated this risk is unknown, but this finding urges us to develop a WW as part of a standardized clinical protocol.

Although WW has been frequently associated with older patients, another publication from the IWWD group observed no differences between younger versus older of 50 years old WW patients in terms disease-specific survival, local regrowth and cumulative risk of distant metastasis⁶⁵.

The presence of pelvic radiation may also significantly affect the QoL in WW patients. A case-matched study comparing 47 WW patients with 41 patients after NAT and TME, showed that QoL was better in the WW group⁶⁶. Notably, a third of the WW patients experienced major low anterior resection syndrome (LARS) as measured by the LARS Score. In the Memorial Sloan Kettering published experience, WW patients usually report better bowel function when measured by the Memorial Sloan Kettering Cancer Center Bowel Function Instrument⁶⁷.

Summary and Future Directions

NOM is a valid and feasible strategy for selected LARC patients, with the challenge still to identify true responders by clinical assessment. The OPRA trial was the first to integrate WW into a TNT strategy aimed at increasing response rates, but most of the current evidence for WW is retrospective, making prospectively collected data of great value. NOM or WW strategies should be part of the treatment discussion of LARC, considering patient interest, patient preference and acceptance of risk, along with the possible reduction in morbidity with avoidance of TME. Currently, the best way for patients to utilize WW strategies is in the context of a prospective trial, if possible, with a strict protocol and objective assessment standards.

Key Points.

• The aim of non-operative management for locally advanced rectal cancer is to correctly identify patients that may develop a pathological complete response after neoadjuvant treatment, using standard clinical methods, in order to offer active surveillance.

• Non-operative management is feasible as a standardized management protocol, due the fact that correlation between a clinical complete response and a pathological complete response is not completely accurate.

• Prospectively collected data from trials suggests safety and high rates of organ preservation (e.g., the OPRA trial) after optimal neoadjuvant therapy approaches in early readouts, but long-term follow-up and additional trials integrating non-operative management will be critical to validate these findings (e.g., the Janus trial).

Synopsis.

Over the past decade, the treatment of locally advanced rectal cancer (LARC) has become more complex. Total neoadjuvant treatment (TNT) has increased the rates of both clinical and pathological complete response, resulting in improved long-term oncological outcomes. Non-operative management (NOM) for rectal cancer has been included as a treatment alternative for selected LARC patients. The feasibility to implement NOM depends on solving current challenges such as how to correctly identify the best candidates for a NOM without compromising oncologic safety. NOM should be part of the treatment discussion of LARC, considering increasing rates of cCR, potential quality of life gains, avoidance of surgical morbidity, and patient preferences.

Abbreviation/Glossary list

LARC

Locally Advanced Rectal Cancer

TME

Total Mesorectal Excision

NAT

Neoadjuvant Therapy

TNT

Total Neoadjuvant Therapy

NOM

Non-operative management

WW

watch-and-wait

pCR

pathological complete response

cCR

clinical complete response

LCRT

Long course chemoradiation

SCRT

Short course radiotherapy