A watch-and-wait approach for locally advanced rectal cancer after a clinical complete response following neoadjuvant chemoradiation: a systematic review and meta-analysis

Summary

Background

A watch-and-wait approach for patients with clinical complete response to neoadjuvant chemoradiation could avoid the morbidity of conventional surgery for rectal cancer. However, the safety of this approach is unclear. We synthesised the evidence for watch-and-wait as a treatment for rectal cancer.

Methods

We systematically searched MEDLINE, Embase, and the grey literature (up to June 28, 2016) for studies of patients with rectal adenocarcinoma managed by watch-and-wait after complete clinical response to neoadjuvant chemoradiation. We determined the proportion of 2-year local regrowth after watch-and-wait. We assessed non-regrowth recurrence, cancer-specific mortality, disease-free survival, and overall survival from studies comparing patients who had watch-and-wait versus those who had radical surgery after detection of clinical complete response or versus patients with pathological complete response.

Findings

We identified 23 studies including 867 patients with median follow-up of 12–68 months. Pooled 2-year local regrowth was 15·7% (95% CI 11·8–20·1); 95·4% (95% CI 89·6–99·3) of patients with regrowth had salvage therapies. There was no significant difference between patients managed with watch-and-wait after a clinical complete response and patients with pathological complete response identified at resection with respect to non-regrowth recurrence (risk ratio [RR] 1·46, 95% CI 0·70–3·05) or cancer-specific mortality (RR 0·87, 95% CI 0·38–1·99). Although there was no significant difference in overall survival between groups (hazard ratio [HR] 0·73, 95% CI 0·35–1·51), disease-free survival was better in the surgery group (HR 0·47, 95% CI 0·28–0·78). We found no significant difference between patients managed with watch-and-wait and patients with clinical complete response treated with surgery in terms of non-regrowth recurrence (RR 0·58, 95% CI 0·18–1·90), cancer-specific mortality (RR 0·58, 95% CI 0·06–5·84), disease-free survival (HR 0·56, 95% CI 0·20–1·60), or overall survival (HR 3·91, 95% CI 0·57–26·72).

Interpretation

Most patients treated by watch-and-wait avoid radical surgery and of those who have regrowth almost all have salvage therapy. Although we detected no significant differences in non-regrowth cancer recurrence or overall survival in patients treated with watch-and-wait versus surgery, few patients have been studied and more prospective studies are needed to confirm long-term safety.

Funding

None.

Introduction

Recommendations for the treatment of locally advanced rectal cancer, defined as T3–T4 or node-positive non-metastatic disease, continue to evolve. Conventional surgical treatment with anterior or abdominoperineal resection and tumour-specific total mesorectal excision has substantial drawbacks, including a 2% risk of perioperative mortality, 11% risk of anastomotic leak, 5% risk of reoperation for complications, and risk of sexual and urinary dysfunction.1, 2, 3, 4 In patients who are candidates for sphincter preservation, bowel dysfunction or low anterior resection syndrome can substantially affect quality of life,⁵ whereas those who have an abdominoperineal resection will live with a permanent colostomy.

Neoadjuvant chemoradiation is recommended for the management of locally advanced rectal cancer with the goal of decreasing rates of local recurrence. Response to neoadjuvant chemoradiation is a predictor of survival.6, 7 After treatment, roughly 15% of patients have a pathological complete response on assessment of the surgical specimen (no detectable tumour cells).7, 8 Compared with patients who do not have a complete response, patients with a complete response have superior 5-year disease-free survival.7, 9 The case for radical surgery, with the associated potential for significant morbidity, in patients who have had a complete response has been questioned.¹⁰ Although pathological complete response can only be determined after surgical resection, clinical complete response (no clinical, endoscopic, or radiographic evidence of disease) has been used as a surrogate measure. However, pathological complete response and clinical complete response are not always concordant.11, 12, 13, 14

In 2004, Habr-Gama and colleagues¹⁵ reported the results of a retrospective cohort study of patients managed with a watch-and-wait approach after a complete clinical response to neoadjuvant chemoradiation. Patients managed by watch-and-wait had exceptional outcomes, including 100% 5-year overall survival and 92% 5-year disease-free survival. Subsequent studies16, 17, 18, 19, 20, 21 by this group accorded with these results and showed that surgical salvage was often successful in patients whose tumours regrew.

Research in context

Evidence before this study

Conventional surgery for rectal cancer is associated with significant morbidity. Although the success of watch-and-wait in selected patients has been reported since 2004, the approach is still met with substantial scepticism; concerns remain regarding the potential for tumour regrowth during follow-up, the effectiveness of salvage therapy after regrowth, and survival. However, most studies assessing outcomes of watch-and-wait for rectal cancer are small and retrospective.

Added value of the study

This study is, to the best of our knowledge, the first meta-analysis of outcomes for patients managed with a watch-and-wait strategy and the first systematic comparative analysis of watch-and-wait versus surgery. For patients managed by watch-and-wait, the pooled proportion of patients who had local regrowth was 15·7% (95% CI 11·8–20·1). Only three (1·9%) of 157 patients with data available could not have salvage therapy after local regrowth because of the extent of local or systemic disease. We did not detect a survival benefit of radical surgery for patients with clinical complete response.

Implications of all the available evidence

We provide stronger evidence for the safety of a watch-and-wait approach; however, few patients managed with watch-and-wait have been studied and not many studies have compared watch-and-wait with surgery in patients with clinical complete response. The establishment of the International Watch & Wait Database and ongoing prospective studies will be crucial in assessing the long-term safety of the watch-and-wait approach.

Although promising, these results have not been uniformly replicated and have not been tested in randomised trials; the true safety of this approach, therefore, remains unclear. We did a systematic review and meta-analysis to assess the evidence for watch-and-wait in patients with complete clinical response by assessing rates of local regrowth and salvage therapy, and comparing rates of recurrent disease and survival outcomes between surgically and non-surgically managed patients.