Colorectal cancer is one of the most common malignancies in the United States, with a third of cases located in the rectum (1). The management of rectal cancer has changed drastically in the last few decades due to improved surgical techniques and the development of multimodal treatment approaches. Until recently, the standard of care for locally advanced rectal cancer (LARC) included neoadjuvant chemoradiation and total mesorectal excision (TME) followed by adjuvant chemotherapy. With this approach, up to 18% of patients had complete eradication of tumor, or a pathologic complete response (pCR), on surgical specimen (2, 3). Studies of patients with a pCR demonstrated excellent long-term outcomes, with 5-year overall survival and disease-free survival rates approaching 90% and 87%, respectively (4–6).

This excellent prognosis raised the question of whether patients with a pCR gained any oncologic benefit from surgical resection. In 2004, Habr-Gama et al. reported outcomes from the first group of highly selected patients enrolled in a watch and wait (WW) surveillance program (7). The safety and efficacy of this approach relied on accurately identifying a complete response using a clinical assessment in place of histologic confirmation of a pCR. WW offered the promise of organ preservation and improved quality of life by eliminating the long-term functional deficits associated with TME. Since Habr-Gama et al’s landmark paper, an extensive volume of literature has demonstrated that WW is a viable treatment strategy that does not compromise long-term oncologic outcomes (8–11). However, challenges remain in accurately identifying patients with complete eradication of disease and in optimizing multimodal treatment regimens to maximize tumor response.

In this article, we review the history of multimodal treatment for LARC and the development of a WW approach. Next, we summarize the restaging process and assess the selection criteria for WW, surveillance strategies and long-term outcomes.

Before exploring the history of WW and its current applications, it is important to first define several common terms used in the literature. LARC includes stage II (T3/T4N0) or III (T1-4N1-2) rectal tumors. A pCR is the gold-standard for confirming eradication of tumor and corresponds to ypT0N0 on TME specimen. In patients considered for nonoperative management, a clinical complete response (cCR) is used as a surrogate marker for a pCR. Patients with a cCR have no evidence of residual disease on restaging flexible sigmoidoscopy, MRI or clinical exam after completing neoadjuvant treatment. Current assessment measures cannot predict a pCR with absolute accuracy, and some of the patients with a cCR subsequently develop local regrowth (12). Patients who continue to have no evidence of residual disease for two years after restaging are considered sustained cCRs (s-cCR). Organ preservation refers to the desired outcome whereby a patient with a s-cCR successfully preserves the rectum with a WW strategy.

A near complete response (nCR) is a relatively recent addition to the clinical response ranking system. These patients have minor abnormalities at restaging, which may resolve into a s-cCR with a longer period of observation. However, some patients with a nCR never achieve a cCR. The exact criteria and management guidelines for a nCR are poorly described, and subsequent sections will discuss this topic in more detail. An incomplete clinical response (iCR) refers to patients who have evidence of residual tumor at restaging. These patients are not candidates for WW and should proceed to TME. In addition to clinical response, other patient factors may influence recommendation for surgical resection, including baseline fecal incontinence and anticipated difficulty adhering to a strict WW surveillance schedule.

Local regrowth refers to disease reappearance at the site of the treated tumor in patients managed with a WW strategy ( Figure 1 ). Local regrowth can occur in the rectal wall, mesorectum or within the lateral internal iliac or obturator lymph nodes (13). The majority of patients with local regrowth undergo successful salvage TME with R0 margins (11, 14, 15). This term should be distinguished from local recurrence (LR), which refers to non-salvageable local regrowth or recurrent cancer in the pelvis following a curative resection.

Until the 1980s, the effectiveness of surgical resection for LARC remained limited with 15-40% of patients developing pelvic recurrence, significant rates of postoperative mortality, and 5-year overall survival rates of 50-70% (16–19). Two innovations- preoperative (chemo)radiation and TME– greatly improved locoregional outcomes. Heald et al. first reported their institution’s experience with TME in 1986, noting that the technique greatly reduced rates of LR and improved overall survival (20). Subsequent pathologic studies reinforced the importance of TME for local disease control by demonstrating that specimens with a violated mesorectal plane were more likely to have positive circumferential resection margins or an inadequate lymph node harvest (21, 22). During the same period, the Swedish Rectal Cancer Trial found that patients who received preoperative short course radiotherapy had significantly lower rates of LR compared to patients who proceeded directly to surgery (23–25). However, the study was criticized for the use of non-standardized surgical techniques, which may explain the higher-than-expected rates of LR in the surgery only group. The subsequent Dutch TME trial demonstrated that preoperative radiation followed by TME reduced LR by 50% without any improvement in overall survival compared to TME alone (26, 27). The CAO/ARO/AIO-94 trial addressed the question of whether radiation should be delivered before or after surgery, with results showing that the preoperative chemoradiation arm had better compliance, less toxicity, increased tumor downstaging and lower rates of LR (28).

Two types of preoperative radiation regimens are available. These include short-course radiotherapy (SCRT) delivered as 25 Gy in 5 fractions and long-course radiotherapy (LCRT) delivered as 45-54 Gy in 25-30 fractions with sensitizing fluorouracil or capecitabine. While LCRT is favored by most providers in the United States, SCRT offers several advantages such as more efficient resource utilization, lower cost to the healthcare system and shorter treatment duration with fewer hospital visits for patients (29). Multiple randomized control trials have compared LCRT to SCRT and found no differences in terms of toxicity, local tumor control, survival or quality of life (30–32).

Simultaneous with the attempts to improve local tumor control, trials explored the use of systemic chemotherapy to reduce the risk of distant metastases, which remained a major driver of mortality in patients with LARC (33, 34). The results of these studies are mixed, with some showing a survival benefit (35, 36), while others found no improvement in oncologic outcomes (37–39). Additionally, these trials demonstrated poor adherence to adjuvant chemotherapy, with nearly 50% of patients declining treatment or receiving less than the recommended dose (33, 40).

Poor tolerance of adjuvant chemotherapy, along with the realization that tumor response to neoadjuvant chemoradiation was time-dependent, generated interest in total neoadjuvant therapy (TNT) (3). TNT delivers chemotherapy preoperatively, either as induction chemotherapy followed by chemoradiation (INCT-CRT), or as chemoradiation followed by consolidation chemotherapy (CRT-CNCT). The most common systemic chemotherapy regimens involve either eight cycles of FOLFOX (leucovorin, fluorouracil, oxaliplatin) or five cycles of CapeOx (capecitabine, oxaliplatin).

This multimodal approach offers several potential advantages, including increased tumor downstaging, earlier introduction of systemic chemotherapy to combat micrometasases, better compliance and decreased time with a temporary ostomy postoperatively (40–42). While approximately 18% of patients receiving neoadjuvant chemoradiation have a pCR, studies evaluating TNT regimens have demonstrated rates closer to 40% (3, 41, 43). Additionally, randomized control trials have shown that when compared to the standard of care (neoadjuvant chemoradiation, TME and adjuvant chemotherapy), patients receiving TNT have better disease-free survival and lower rates of distant metastases (3, 43–45). As a result, the National Comprehensive Cancer Network (NCCN) now recommends TNT as first line treatment for LARC (46).

Janeway first described the successful treatment of rectal tumors with contact radiation and implantation of radioactive seeds in 1917 (53). While this approach remained a cornerstone of rectal cancer management for decades, it fell out of favor with advances in perioperative safety and improvements in surgical outcomes (54). In the modern era, Habr-Gama has pioneered the study of non-operative management for tumors with a complete response to neoadjuvant chemoradiation (7). Her landmark paper from 2004 compared outcomes between 71 WW patients and 22 patients with a pCR after surgical resection. Patients were evaluated 8 weeks after completion of chemoradiation using endoscopy, digital rectal exam (DRE) and CT imaging. Patients with a cCR followed a strict surveillance schedule, including monthly clinical and endoscopic exams, as well as pelvic CT scans every 6 months for the first year. The surveillance interval increased to 2 and 6 months during the second and third years, respectively. With a mean follow-up of 57.3 months, two (2.8%) WW patients developed salvageable local regrowths and three (4.2%) developed distant metastases. Five-year disease-free survival was 92% and did not differ significantly from patients with a pCR (7).

Habr-Gama et al’s findings suggested that a clinical restaging assessment could detect a complete response and that patients with a cCR could be safely monitored using a strict surveillance protocol. Subsequent studies validated these findings, repeatedly showing that patients who achieved a s-cCR had oncologic outcomes comparable to those with a pCR (9–11, 55–58). Nevertheless, WW remained confined to specialized academic centers, with many providers hesitant to adopt the approach in the absence of data from a randomized control trial.

The Organ Preservation in Rectal Adenocarcinoma (OPRA) Trial was the first randomized phase II trial to evaluate long-term oncologic outcomes in patients offered WW (8). Participants with LARC were randomized to receive either INCT-CRT or CRT-CNCT and were restaged 8 ± 4 weeks after completion of neoadjuvant treatment using flexible sigmoidoscopy, DRE and MRI. Long-term follow-up demonstrated that nearly half (46.7%) of the enrolled patients achieved organ preservation (59). Oncologic outcomes did not differ compared to historical controls that underwent TME (8, 59).

Patients with local regrowth harbor residual disease not clearly evident during the restaging assessment. Between 15% to 36% of patients who enter WW surveillance develop local regrowth, with most cases occurring within two years of restaging (10, 11, 59, 112–115). Data from the International Watch and Wait Database indicate that 97% of all regrowths occur within the bowel wall, highlighting the importance of careful endoscopic surveillance (9). Unlike local recurrence following TME, where less than a third of patients have resectable disease, the vast majority of regrowths are salvageable (8, 11, 113).

The effects of local regrowth on oncologic outcomes remain poorly described and are difficult to determine using retrospective data. Studies suggest that patients with local regrowth have a higher risk of distant metastases compared to those with a s-cCR (11, 116) but have similar survival outcomes to patients with an iCR taken to TME immediately after restaging (59). However, local regrowths are a biologically distinct set of tumors with an excellent (but incomplete) response to neoadjuvant therapy. Therefore, patients with complete eradication of disease (s-cCR) and those with a poor response to treatment (iCR) cannot act as appropriate control groups.

Additionally, it remains unclear whether the delay in definitive surgical resection for patients with local regrowth increases the risk of distant metastases. Our understanding of the temporal relationship between metastatic progression and local regrowth is limited. Distant metastases may be a consequence of inherently poor tumor biology already present at diagnosis or may develop after TNT from occult cancer cells left at the primary tumor site (11, 116). Unfortunately, it is difficult to provide a clear answer to this question, as evaluating the impact of delaying surgery would require prospective randomized control trials that are unlikely to accrue. Current focus lies in accurate detection of regrowths with surveillance exams and an expedient TME once residual disease is suspected.

Definitive proof of the safety and efficacy of nonoperative management requires a large phase III trial with a non-inferiority design. However, randomized control trials comparing selective WW and mandatory TME strategies are considered infeasible due to the risk of patient crossover into the WW arm. Until recently, most evidence supporting the safety of a WW strategy came from studies using multi-institutional, retrospective databases. However, these analyses could not account for differences in treatment regimens, clinical response criteria, timing of restaging assessment or surveillance protocols. The OPRA trial was the first prospective phase II trial to demonstrate that patients with an excellent response to TNT could achieve long-term organ preservation with oncologic outcomes equivalent to historical controls treated with the standard of care (chemoradiation, TME and adjuvant chemotherapy) (8). While the OPRA trial provides the best information to date on the safety and efficacy of nonoperative management, its findings are limited by constraints in study design. The trial did not randomize patients by treatment strategy (TME vs. WW) and instead relied upon comparisons to historical controls treated with a different neoadjuvant regimen (chemoradiation vs. TNT). Despite these limitations, the OPRA trial considerably improved our understanding of patient selection for WW and strengthened available evidence demonstrating that nonoperative management is a viable treatment option for patients with a complete response following TNT.

WW has become an accepted alternative to TME in LARC patients with a complete response to neoadjuvant treatment. The OPRA trial demonstrated that over 45% of patients achieve long-term organ preservation with oncologic outcomes comparable to historical controls. While approximately 1/3 of patients who entered WW experienced local regrowth, all were salvageable by TME and this subgroup exhibited similar outcomes to patients with an iCR after TNT. Several challenges to expanding the use of WW remain. First, high-quality prospective data on the feasibility and oncologic safety of this strategy is limited to a single, phase II randomized control trial. Second, the restaging exam’s diagnostic performance continues to suffer from sub-optimal accuracy, particularly in identifying the presence of viable tumor cells that later become local regrowth. Third, the advantages of various TNT regimens with regards to maximizing a complete response, decreasing risk of local regrowth and minimizing treatment-associated toxicity remain unknown.

Future directions for research involve tailoring treatment regimens to optimize patient outcomes. This includes identifying genetic and molecular markers responsive to immunotherapy or predictive of a complete response (60, 122, 123). The Janus Rectal Cancer Trial, which randomizes patients to receive CRT-CNCT with doublet (FOLFOX or CapeOx) or triplet (mFOLFIRINOX) chemotherapy will provide data on the potential efficacy of a triplet chemotherapy regimen in improving rates of disease-free survival and organ preservation (124). Machine learning and radiomics may prove useful in improving the accuracy of restaging endoscopy and MRI (125, 126). Finally, WW remains limited to stage II or III rectal cancers. The ongoing STAR-TREC trial aims to evaluate the suitability of this approach for stage I disease treated with neoadjuvant (chemo)radiation (127).