Long-term outcomes in a retrospective cohort of patients with rectal cancer with complete response after total neoadjuvant therapy: a propensity-score weighted analysis

Abstract

Background:

The watch-and-wait (W&W) strategy is a novel treatment option for patients with rectal cancer who have a strong desire for organ preservation. The study aimed to explore the long-term outcomes of the W&W strategy in a large cohort of rectal cancer patients who achieved a clinical complete response (cCR) after consolidation total neoadjuvant therapy (TNT), and to compare with patients who achieved a pathological complete response (pCR) after radical surgery.

Methods:

The W&W group comprised patients who were assessed as having a cCR after consolidation TNT and adopted the W&W strategy. Patients who underwent standard resection and achieved a pCR were compared as a reference. Inverse probability of treatment weighting (IPTW)-adjusted Kaplan–Meier analysis with log-rank test was used to compare survival outcomes.

Results:

We included 89 and 171 patients in the W&W and pCR groups, respectively. The median follow-up period was 45 and 58 months for the W&W and pCR groups, respectively. After IPTW adjustment, the 2-year local regrowth/recurrence rate for the W&W and pCR groups were 9.9% and 2.0%, respectively (p < 0.001). The W&W and pCR groups had similar 5-year outcomes, including overall survival, disease-free survival, and distant metastasis-free survival (all p > 0.05). No significant difference was observed in the rates of distant metastasis between patients in the W&W group with local regrowth and those without local regrowth (25% versus 6.2%, p = 0.119).

Conclusion:

Patients who were managed with a W&W strategy after consolidation TNT had favorable survival outcomes, which were similar to those of patients with a pCR. The rate of local regrowth in W&W patients was lower in our study than in other studies as a result of the implementation of consolidation TNT.

Introduction

Over the past 2 decades, the focus on neoadjuvant treatment for patients with locally advanced rectal cancer (LARC) has shifted toward a more tailored approach, with an emphasis on both longer survival and better quality of life (QoL). ¹ LARC is typically defined as cT3–4 and/or positive lymph nodes, and the standard treatment for LARC patients is neoadjuvant chemoradiotherapy (CRT), followed by total mesorectal excision (TME) and adjuvant chemotherapy, ² through which nearly 15–30% of patients achieve a pathological complete response (pCR) and have a favorable prognosis.^3,4 However, patients who undergo TME may experience bowel, urinary, and sexual dysfunction, or even permanent colostomy, leading to impaired QoL. ⁵ To preserve organ function, a watch-and-wait (W&W) strategy was initiated by Habr-Gama’s et al. ⁶ group. In this strategy, patients who had a clinical complete response (cCR) were treated with watchful waiting instead of radical surgery, making it a welcome alternative for patients who were highly likely to succeed in non-operative management.

The safety and efficacy of the W&W strategy have been investigated in both retrospective case series and prospective observational studies. Due to small sample sizes and the nature of retrospective investigations, results including 2-year local regrowth rates (2.8–50.0%), 5-year disease-free survival (DFS) rates (76.3–93.0%), and 5-year overall survival (OS) rates (85–100%) varied widely between studies.^6–11 Patients in most studies were heterogeneous in terms of baseline characteristics, CRT regimen, response assessment criteria, and surveillance protocols. Adjustment for these confounding factors is warranted to compare the outcomes in patients with a cCR or pCR.

Based on conventional preoperative CRT plus adjuvant chemotherapy, a new treatment approach, total neoadjuvant therapy (TNT), has been developed to improve the delivery of planned chemotherapy and eliminate micrometastases at an early stage.^4,12,13 In this novel modality, courses of both CRT and chemotherapy are administered as neoadjuvant therapy prior to radical surgery. More recently, the TNT approach has been supported by phase III trial data and endorsed by the National Comprehensive Cancer Network. ¹⁴ The potential benefits of this scheme are not only an improvement in survival but also a higher rate of complete response (CR),^15–17 thereby increasing the proportion of patients who have the opportunity to adopt the W&W strategy. ¹⁸ Although results from the Organ Preservation for Rectal Adenocarcinoma (OPRA) Trial showed that adoption of TNT contributed to half of the rectal cancer patients achieving organ preservation, ¹⁹ no real-world study has been conducted to compare the oncologic outcomes of W&W patients after TNT with those of pCR patients.

Therefore, we aimed to report the first real-world data on the clinical outcomes of the W&W strategy in a large cohort of patients who achieved a cCR after chemoradiation plus consolidation chemotherapy as TNT. We also provided unbiased evidence comparing patients with a cCR and patients who achieved a pCR after radical surgery.

Materials and methods

Study design and participants

After institutional review board approval, patients who were diagnosed with rectal cancer and received neoadjuvant treatment between 1 January 2015 and 1 September 2021 were reviewed for this study. Patients were included in the W&W group if they achieved cCR after neoadjuvant CRT followed by consolidation chemotherapy and did not undergo radical surgery. Patients who achieved pCR after neoadjuvant CRT and radical surgery (pCR group) were compared as a reference. Patients were excluded if they had distant metastases before or during neoadjuvant treatment. Patient characteristics, including age at diagnosis, sex, tumor distance from the anal verge, clinical tumor and nodal stage on rectal magnetic resonance imaging (MRI), neoadjuvant treatment regimens, baseline serum carcinoembryonic antigen (CEA) levels, status of mesorectal fascia (MRF), and extramural vascular invasion (EMVI), were collected from the electronic health record review for further analysis. All patients provided informed consent for their data to be used in retrospective studies. All patients were informed of the risks of the W&W strategy.

Treatment

Pelvic long-course radiotherapy was administered at a planned dose of 5000 cGy in 25 fractions over 5–6 weeks. Fluorouracil or capecitabine, alone or in combination with irinotecan, was administered concurrently. Patients who were good candidates for organ preservation received consolidation chemotherapy. The consolidation regimen included fluorouracil/capecitabine or fluorouracil/capecitabine-based therapy in combination with irinotecan and/or oxaliplatin, depending on the regimen they received concurrently with radiotherapy. For those who continued with the standard procedure, consolidation chemotherapy was administered in the interval before radical surgery and TME was mandatory. Adjuvant chemotherapy was prescribed at the discretion of the treating physicians.

Response assessment and follow-up

The assessment of cCR was based on a combination of digital rectal examination (DRE), rectal MRI, colonoscopy, and CEA level tests. Patients were evaluated as having cCR if they met the following criteria: (1) no induration or palpable tumor on DRE; (2) no residual tumor or suspicious lymph nodes on MRI; (3) no visible tumor, ulceration, or nodularity on endoscopy; and (4) CEA level <5 ng/mL. For patients undergoing radical resection, pCR was defined as no histopathological evidence of viable tumor cells in the surgical specimen at the site of the primary tumor and lymph nodes (ypT0N0).

All patients managed with the W&W protocol underwent monthly DRE for the first year after the completion of treatment. Serum CEA levels, colonoscopy, and rectal MRI were assessed every 3–4 months during the first 2 years after treatment completion, and then every 6 months for 3–5 years after treatment. Chest and/or abdominal computed tomography (CT) was performed every 6 months during the first 2 years after the completion of treatment and annually for 3–5 years after treatment. Patients who underwent TME were followed-up in accordance with the National Comprehensive Cancer Network guidelines.

Outcomes

Local regrowth was defined as the reappearance of a tumor in the rectal wall detected by DRE and colonoscopy or enlargement of regional lymph nodes on pelvic imaging. All patients found to have local regrowth were indicated for histologic confirmation by biopsy or salvage surgical specimens. Chest and abdominal CT were performed to determine the presence of distant metastases. Endpoints, including OS, non-regrowth DFS, and distant metastasis-free survival (DMFS), were evaluated and compared between patients in the W&W and pCR groups.

All time-to-events were calculated from the date CRT started to the time of their occurrence or were censored at the last follow-up. Local control was compared between the two groups by calculating the local regrowth rate for patients with a cCR from the date of the W&W decision and the local recurrence rate for patients with a pCR from the date of surgery.

Statistical analysis

Patient characteristics are described as the median and interquartile range (IQR) for continuous variables and frequency and proportion for categorical variables. The Wilcoxon test and Chi-square test (or Fisher’s exact test) were used to compare differences between groups. OS, DFS, and local control rates were estimated using the Kaplan–Meier method.

To account for possible baseline covariates between the W&W and pCR groups, a stabilized inverse probability of treatment weighting (IPTW)-adjusted analysis was performed. We determined the probability of achieving cCR by fitting a logit model of an indicator variable denoting cCR versus pCR based on age, tumor distance from the anal verge, clinical tumor, nodal stages, and MRF status. The IPTW method balanced the covariates of the two groups by weighing all patients in the database using the inverse of their propensity score (PS) [1/PS in the W&W group and 1/(1-PS) in the pCR group]. IPTW-adjusted Kaplan–Meier curves were used to calculate survival and time-dependent rates. Differences between groups were compared using the IPTW-adjusted log-rank test, and 95% confidence intervals (CIs) were calculated using the IPTW-adjusted Cox proportional hazards regression model. All statistical analyses were performed using R, version 4.2.0 (The R Foundation for Statistical Computing; Vienna, Austria). A two-sided p value < 0.05 was considered statistically significant.

Results

Between 1 January 2015 and 31 August 2021, 89 patients with rectal cancer who received the TNT regimen with consolidation chemotherapy achieved cCR and were subsequently managed with a W&W strategy (W&W group). During the same period, 171 patients who received neoadjuvant CRT followed by TME demonstrated pCR (pCR group). Patients who chose W&W were all good surgical candidates but had a strong desire for organ preservation (Figure 1).

Figure 1.

Schematic illustration of the study.

cCR, clinical complete response; pCR, pathologic complete response; TME, total mesorectal excision; W&W, watch-and-wait.

Patient and tumor characteristics of the unweighted population are listed in Table 1. The cohorts differed significantly in terms of age, tumor height from the anal verge, clinical tumor stage, nodal stage, and MRF status in the unweighted population (all p < 0.05). Patients in the W&W group were slightly older [median (IQR), 57.0 (52.0–65.0) versus 55.0 (45.5–63.0) years; p = 0.047] and presented with tumors located closer to the anal verge [median (IQR) height, 3.0 (2.0–4.0) versus 4.7 (3.0–6.0) cm; p < 0.001]. Regarding clinical tumor stages, most patients (80.1%) in the pCR group had T3 tumors, while nearly 35% of patients in the W&W group had T1–2 (p < 0.001). The nodal stages were evenly distributed in the W&W group; however, a higher proportion of patients (53.8%) in the pCR group had N2 disease (p = 0.001). Patients in the W&W group were more likely to have uninvolved MRF compared with the pCR group (84.3% versus 66.1%, p = 0.003). As all patients in the W&W group were administered a consolidation TNT regimen, the duration of consolidation chemotherapy was significantly longer in the W&W group than in the pCR group [median (IQR) chemotherapy interval: 5.0 (4.0–6.0) versus 2.0 (1.0–3.0) months; p < 0.001]. Other variables, including sex, baseline CEA level, EMVI status, and concurrent and consolidation chemotherapy regimens, were not significantly different between the two groups. The IPTW adjustment for statistically different covariates resulted in a distinctive balance of baseline characteristics between the two groups (Table 2).

Table 1.

Patients’ unmatched clinical characteristics.

Characteristics	W&W (N = 89)	pCR (N = 171)	p
Sex, No. (%)
Men	59 (66.3)	124 (72.5)	0.368 $
Women	30 (33.7)	47 (27.5)
Age, median (IQR), y	57.0 (52.0, 65.0)	55.0 (45.5, 63.0)	0.047*
Height from anal verge, median (IQR), cm	3.0 (2.0, 4.0)	4.7 (3.0, 6.0)	<0.001*
Clinical tumor (T) classification, No. (%) ‡
cT1	2 (2.2)	0 (0.0)	<0.001 $
cT2	29 (32.6)	12 (7.0)
cT3	53 (59.6)	137 (80.1)
cT4	5 (5.6)	22 (12.9)
Clinical nodal (N) classification, No. (%) ‡
cN0	26 (29.2)	17 (9.9)	0.001 $
cN1	29 (32.6)	61 (35.7)
cN2	34 (38.2)	92 (53.8)
cN3	0 (0.0)	1 (0.6)
EMVI status, No. (%)
+	73 (82.0)	120 (70.2)	0.054 $
−	16 (18.0)	51 (29.8)
MRF status, No. (%)
+	75 (84.3)	113 (66.1)	0.003 $
−	14 (15.7)	58 (33.9)
Baseline CEA level, median (IQR), ng/ml	2.73 (1.92, 5.18)	3.62 (1.78, 5.58)	0.548*
Concurrent chemotherapy regimen, No. (%)
Capecitabine	23 (25.8)	46 (26.9)	0.972 $
Iritnotecan	66 (74.2)	125 (73.1)
Consolidation chemotherapy, No. (%)
Yes	89 (100.0)	159 (93.0)	0.025 $
no	0 (0.0)	12 (7.0)
TNT regimen, No. (%) §
Capecitabine	4 (4.5)	11 (6.4)	0.979 $
Irinotecan	61 (68.5)	113 (66.1)
Oxaliplatin	15 (16.9)	29 (17.0)
Triple drug	8 (9.0)	16 (9.4)
Target therapy	1 (1.1)	2 (1.2)
Consolidation chemotherapy duration, median (IQR), m	5.0 (4.0, 6.0)	2.0 (1.0, 3.0)	<0.001*

^*The Kruskal–Wallis test.

^$The χ² test or Fisher exact test where applicable.

^‡According to the 7th edition of the American Joint Committee on Cancer Tumor-node-metastasis (TNM) staging system.

^§The consolidation chemotherapy regimen consists of long-course chemoradiation followed by chemotherapy.

CEA, carcinoembryonic antigen; EMVI, extramural vascular invasion; IQR, interquartile range; MRF, mesorectal fascia; pCR, pathologic complete response; TNT, total neoadjuvant therapy; W&W, watch-and-wait.

Table 2.

Patients’ matched clinical characteristics.

Characteristics	W&W	pCR	p
Age, median (IQR), y	56.0 (48.0–62.0)	56.0 (47.0, 64.0)	0.922*
Height from anal verge, median (IQR), cm	4.0 (3.0, 5.0)	4.0 (3.0, 5.0)	0.931*
Clinical tumor (T) classification, No. (%) ‡
cT1	0.7 (0.8)	0.0 (0.0)	0.514 $
cT2	13.5 (14.9)	21.7 (13.2)
cT3	62.3 (68.9)	124.8 (75.8)
cT4	14.0 (15.5)	18.2 (11.0)
Clinical nodal (N) classification, No. (%) ‡
cN0	14.4 (15.9)	24.6 (15.0)	0.890 $
cN1	28.9 (32.0)	57.5 (34.9)
cN2	47.1 (52.1)	81.8 (49.7)
cN3	0.0 (0.0)	0.7 (0.4)
MRF status (%)
+	66.5 (73.5)	116.7 (70.9)	0.732 $
−	24.0 (26.5)	47.9 (29.1)

^*The Kruskal–Wallis test.

^$The χ² test or Fisher exact test where applicable.

^‡According to the 7th edition of the American Joint Committee on Cancer TNM staging system.

IQR, interquartile range; MRF, mesorectal fascia; pCR, pathologic complete response; W&W, watch-and-wait.

Table 3.

Clinical outcomes of patients with cCR and pCR.

Outcome	W&W (N = 89)	pCR (N = 171)
Months of follow-up, median (IQR), m	45 (25–63)	58 (26–72)
Local regrowth, No. (%)
Total	8 (9.0)
With distant metastasis	2 (2.2)
Local recurrence, No. (%)
Total		3 (1.8)
With distant metastasis		3 (1.8)
Distant metastasis, No. (%)	7 (7.9)	9 (5.2)
Death, No. (%)	5 (5.6)	7 (4.1)

IQR, interquartile range; pCR, pathologic complete response; W&W, watch-and-wait.

The median follow-up period for the W&W group was 45 months (IQR, 25–63 months). Overall, eight of the 89 patients developed local regrowth, with all tumors located in the bowel wall. Before IPTW adjustment, the 24-month local regrowth rate was 7.5% (95% CI: 1.5–13.1%). The median time to local regrowth from the W&W decision for these eight patients was 11.0 months (range, 5.0–48.0 months) and 62.5% of the patients had local regrowth in the first 12 months. Six of the eight patients underwent salvage surgery, of whom two underwent low anterior resection, three underwent abdominoperineal resection, and one patient underwent transanal endoscopic microsurgery for definitive salvage therapy. The pCR group had a median follow-up of 58 months (IQR, 26–72 months), in which three patients developed local recurrence, corresponding to a 2-year local recurrence rate of 1.4% (95% CI: 0–3.4%) (Supplemental Figure S1).

After IPTW adjustment, the 2-year local regrowth/recurrence rate assessed from no evidence of disease (NED) for the W&W and pCR groups was 9.9% (95% CI: 1.6–17.5%) and 2.0% (95% CI: 0–4.9%), respectively, and it was significantly higher in the W&W group (p < 0.001) [Figure 2(a)].

Figure 2.

IPTW-adjusted survival outcomes in the W&W and pCR groups. (a) Local control, (b) DMFS, (c) OS, and (d) DFS.

cCR, clinical complete response; CRT, chemoradiotherapy; DFS, disease-free survival; DMFS, distant metastasis-free survival; IPTW, inverse probability of treatment weighting; OS, overall survival; pCR, pathologic complete response; W&W, watch-and-wait.

At the end of follow-up, seven patients (7.9%) in the W&W group developed distant metastasis, including three with liver metastasis, two with lung metastasis, and one with systemic relapse. Two patients (28.6%) experienced simultaneous local regrowth and systemic recurrence. Two of the eight patients (25%) with local regrowth in the W&W group developed distant metastasis (Table 2), whereas 5 of 81 patients (6.2%) without local regrowth developed distant metastasis. No significant difference was observed in the rates of distant metastasis between patients in the W&W group with local regrowth and those without local regrowth due to small sample size (25% versus 6.2%, p = 0.119). In addition, one in six patients (16.7%) who received salvage surgery developed distant metastases, while one in two patients (50%) who refused salvage surgery developed distant metastases. No significant difference was observed in distant metastasis rates between patients receiving salvage surgery and those not receiving salvage surgery after local regrowth (50% versus 16.7%, p = 0.464).

In the pCR group, nine patients (5.2%) developed metastasis, of whom one had liver metastasis, three had lung metastasis, and five had metastasis at other sites (e.g. peritoneal, brain, and bone). Three (33.3%) patients were also found to have local recurrence. The IPTW-adjusted 5-year DMFS rates were 92.8% (95% CI: 86.7–99.4%) and 93.6% (95% CI: 89.2–98.1%) for the W&W and pCR groups, respectively (p = 0.68) [Figure 2(b)].

Overall, five (5.6%) patients in the W&W group and seven (4.1%) patients died in the pCR group. IPTW-adjusted Kaplan–Meier curves showed that the 5-year OS rates were not significantly different between the W&W and pCR groups [92.1% (95% CI: 84.6–100%) versus 93.1% (95% CI: 87.9–98.6%); p = 0.57] [Figure 2(c)]. The IPTW-adjusted 5-year DFS rates were 93.3% (95% CI: 87.6–99.4%) and 92.0% (95% CI: 87.2–97.0%) for the W&W and pCR groups, respectively (p = 0.66) [Figure 2(d)]. Analysis of the unweighted cohort also showed no significant differences between the two groups (Supplemental Figure S2).

Discussion

The results of the present study suggested that the W&W strategy after TNT, using consolidation chemotherapy, has a good safety profile with a low rate of local regrowth. An unbiased comparison of long-term survival outcomes showed no significant difference between patients with a cCR and those with a pCR.

Since the pioneering work by Habr-Gama’s et al. ⁶ group demonstrated that non-operative management following CRT in cCR patients had non-inferior survival outcomes to those with a pCR after TME, evidence has accumulated that the management of the W&W strategy in cCR patients is oncologically safe and feasible, from single-center reports with small sample sizes to large-scale registry cohorts.

Notably, the IPTW-adjusted 5-year OS rate in our study was 92.1% in the W&W group, similar to that reported by Van Der Valk et al. ¹¹ (85%), Martens et al. ²⁰ (3-year: 96.6%), and Renehan et al. ⁹ (3-year: 96%). Furthermore, the IPTW-adjusted 5-year OS rate was similar to that of pCR patients (93.1%), who have a favorable prognosis,^3,4 indicating that the strict assessment criteria in our study successfully selected patients with a true CR for the W&W strategy.

However, the main concern with the W&W strategy is local regrowth. Previous studies reported that the 2-year local regrowth rates ranged from 2.8% to 34%. ¹⁰ Although the consensus is that most local regrowth in W&W is salvageable and not equivalent to local recurrence after radical surgery, ²¹ we still compared the local control in patients with cCR to those with pCR from the date of NED. Our results showed that the IPTW-adjusted 2-year local regrowth rate of cCR patients was 9.9%, which was significantly higher than that of pCR patients (2.0%), but noticeably lower than that reported in studies by Van Der Valk et al. ¹¹ (25.2%), Smith et al. ⁷ (5-year: 21%), Martens et al. ²⁰ (15.4%), and Renehan et al. ⁹ (38%).

Multiple reasons may account for the lower rate of local regrowth in our study, including the adoption of TNT and patient selection bias. First, all patients in the W&W cohort received TNT with consolidation chemotherapy. According to the results of the CAO/ARO/AIO-12 trial, ²² the rate of pCR was higher in patients with rectal cancer treated with CRT followed by consolidation chemotherapy than in those receiving induction chemotherapy. Similarly, in the OPRA trial, ¹⁹ the group receiving TNT with consolidation chemotherapy achieved a higher rate of organ preservation and a lower rate of local regrowth. Second, nearly 70% of the patients in our study were treated with an irinotecan-containing TNT regimen. Few studies have clearly established the value of irinotecan in TNT. The results from the CinClare trial showed that the concurrent addition of irinotecan during CRT significantly increased the pCR rate compared with capecitabine alone, ²³ and the updated results showed a tendency to improve oncologic outcomes in patients with rectal cancer, ²⁴ which reflected the potential efficacy of irinotecan in the neoadjuvant setting. We provided real-world data on the value of irinotecan in TNT regimens. Third, a median interval of 5 months from the end of CRT to the assessment of response was observed in the W&W group, as it was considered that the interval was a potential factor contributing to a higher rate of pCR. ²⁵ Lastly, approximately 20% of the patients in our study were diagnosed as clinical stage I by rectal MRI.

In our study, we also noted that patients who experienced local regrowth were more likely to develop distant metastasis, as reported by Smith et al. ⁷ However, no statistical difference was observed because of the limited sample size. As distant metastasis remains the main cause of treatment failure, this implied that a more aggressive approach for patients with local regrowth is needed.

Compared with most studies evaluating the long-term outcomes of patients with cCR and pCR, the strength of our study was the correction of potential confounding effects by the IPTW approach. Additionally, we included a large cohort of patients from a single center with uniform enrollment and response assessment criteria, which allowed us to compare patients with cCR and pCR with more consistency.

However, our study has some limitations, mainly due to the nature of retrospective studies. Although most patients in our study received the irinotecan-containing TNT regimens that showed promising results in the W&W strategy, prospective randomized trials are warranted. Future prospective studies are also needed to optimize the management of W&W strategies, including the exploration of more effective treatment modalities to increase the CR rate, such as immunotherapy in the neoadjuvant setting, and the establishment of more sensitive response assessment criteria to select patients with true CR for the W&W strategy.

Conclusion

To the best of our knowledge, our study was the first to report real-world data of the clinical outcomes of the W&W strategy in patients with rectal cancer after chemoradiation plus consolidation chemotherapy as TNT regimen. The IPTW-adjusted survival data showed that patients managed with the W&W strategy had favorable long-term outcomes, similar to those of patients with a pCR. Compared to other W&W studies, the local regrowth rate in our study was significantly reduced by adopting TNT with consolidation chemotherapy. The broad use of irinotecan in our study might also contribute to a lower local regrowth rate to some extent. However, the benefit of irinotecan needs to be further validated in prospective randomized trials. Our study may provide significant insights for future clinical practice on the W&W strategy, especially after TNT using consolidation therapy.