Is postoperative chemotherapy necessary for resected esophageal carcinoma after neoadjuvant chemoradiotherapy? A systematic review and meta-analysis CURRENT STATUS: UNDER REVIEW

Background Both neoadjuvant chemoradiotherapy (NCRT) and postoperative chemotherapy (PT) are crucial treatments for esophageal carcinoma (EC). However, it is not clear whether PT is required for EC treatment after NCRT. This systematic review and meta-analysis aimed at clarifying the necessity of PT for resected EC after NCRT. Methods We searched PubMed, Embase, and The Cochrane Library databases for relevant studies published up to March 2020, that have compared PT and non-PT for resected EC after NCRT (NCRT + PT vs. NCRT). The primary outcome of this study was overall survival (OS). Hazard risk ratio (HR) and 95% confidence interval (CI) were calculated. Subgroup and sensitivity analyses were employed to explore heterogeneity, and the random effect model was used to merge the meta-analysis data, regardless of whether the heterogeneity was large or small. Results This study included seven retrospective cohorts, with more than 10720 patients. Most of the patients had esophageal adenocarcinomas. The Meta-analysis showed that NCRT followed by PT increased the patient OS ( HR = 0.79, 95% CI 0.74–0.85, P < 0.001). However, further subgroup analysis showed that NCRT + PT might not improve the OS of resected EC patients with a negative lymph node status ( HR = 0.82, 95% CI 0.67–1.01, P = 0.124). Further, we showed that NCRT with PT improved the survival of EC patients with a positive lymph node status who underwent resection ( HR = 0.78, 95% CI 0.70–0.86, P < 0.001). Conclusion PT may improve the survival of lymph node-positive EC patients previously treated by NCRT. This conclusion may be more applicable to EAC patients treated with NCRT at the ypN + stage.


Introduction
Esophageal carcinoma (EC) is one of the most prevalent malignant tumors worldwide. 1 Compared to surgery alone, multimodality therapy results in a significant improvement in survival rates in locally advanced EC.
For EC patients who did not receive neoadjuvant therapy, the decision of whether to provide PT depends on the status of lymph nodes after surgery. Compared with surgery alone, PT does not result in significant survival benefits to EC patients with negative lymph nodes. However, PT has been proven to improve the survival of patients with locally advanced EC with positive lymph nodes. 2 A previous cross-trial showed that NCRT significantly prolongs the overall survival (OS) of locally advanced EC and the strategy is currently the best method for the treatment of the EC. 1, 3 However, whether PT is needed in patients with locally advanced EC after NCRT and surgery should be assessed. The NCCN clinical practice guidelines recommend that resected Esophageal adenocarcinoma (EAC) patients with either lymph node-positive or negative observation need PT, while resected Esophageal squamous cell carcinoma (ESCC) patients do not need PT.
Currently, patients with locally advanced EC receive NCRT combined with surgery, which increases medical costs and requires longer recovery time. Conducting PT further increases medical costs and prolongs patient recovery. Therefore, careful consideration should be made before conducting PT on EC patients after NCRT combined with surgery. At present, no randomized controlled study outcomes have been reported on whether EC patients could benefit from PT after NCRT and surgery.
However, the results from published cohort studies show that PT improves the OS of patients who received NCRT with EC, especially for patients who showed a positive lymph node status after NCRT.
This study aimed to collect substantial evidence from published cohort studies to assess the effect of conducting PT on the survival of patients with resected EC after NCRT.

Materials And Methods
This systematic review and meta-analysis confirmed with The Cochrane Collaboration Handbook of Interventions Systematic Reviews. Data reporting in this manuscript was done using the PRISMA Statement.

Criteria for considering studies for this review
The inclusion criteria were: (1) Studies that have compared the resected survival of EC patients after NCRT, with or without PT; (2) Both randomized controlled trials (RCTs) and non-randomized controlled studies (prospective and retrospective studies); (3) A follow-up time of more than one year; (4) Studies with results on hazard ratios (HRs) and analysis was done at a 95% confidence interval (CI), or the hazard ratio and 95% CI could be estimated from the data.

Search strategy
We conducted a systematic literature search in PubMed, Embase, and the Cochrane Library (from the inception to March 2020). Our search strategy included free words, Emtree terms and MeSH terms, such as "Esophageal carcinoma," "Neoadjuvant chemoradiotherapy," "Preoperative chemoradiotherapy," "Adjuvant chemotherapy," and "Postoperative chemotherapy."

Selection of studies
Literature search and selection was conducted by two researchers who independently read the abstracts and full articles to evaluate whether they qualify for inclusion. Disagreements between the two researchers were solved by consensus with a third reviewer.

Quality assessment
The quality of the literature in all the cohort studies was evaluated using the Newcastle-Ottawa quality scale (NOS). 4 NOS consists of three factors: patient selection, comparability of the study groups, and outcome assessment. A score of 0-9 was allocated to all studies. We defined the literature with a score ≥ 6 as of good quality and literature with a score 6 as of poor quality.

Outcome Measure and data extraction
We defined OS as the primary outcome. The basic information of each study was extracted, and HR and 95% CI were used as the indices for survival evaluation. We assessed the HR values for multiple follow-up time points and selected the results with the longest follow-up time. For studies that only provided survival curves (Kaplan-Meier curves) and other relevant data, HR and 95% CI were estimated using the Digitizer software. 5 For studies that provided results with multiple different correction methods, the results with the highest number of correction factors were included.

Data Analysis
The Stata software version 15.0 was used for all data analyses. Survival data were analyzed using HR at 95% CI. The random-effects model was used in all meta-analyses.
Subgroup analysis was performed to assess the survival of the two lymph node metastases status (ypN0 and ypN+) after NCRT combined with surgery.
The heterogeneity was considered to be low if I 2 50%, and high if I 2 ≥ 50%. We used the random effect model in all the analyses to ensure the reliability of the conclusion.
A sensitivity analysis was performed to identify the source of heterogeneity, and publication bias was evaluated using funnel plots and the Egger's test. A P value of less than 0.05 in two-tailed tests was considered statistically significant.

Trial Flow and Characteristics
Our meta-analysis included seven eligible studies that had more than 10720 patients. 6-12 A flow chart of the procedures followed in literature retrieval and selection is shown in Fig. 1. The basic characteristics of the literature are presented in Table 1.

A meta-analysis of the OS in ypN + patients received NCRT + PT vs. NCRT
Only four studies compared the OS of lymph node-positive patients (ypN+) after NCRT and surgery. 6, 7, 11, 12 There was low heterogeneity between included studies (I 2 = 0, PH = 0.450). The results showed that NCRT + PT had a better OS as relative to NCRT alone (HR = 0.78, 95% CI 0.70-0.86, P < 0.001). PT improves the OS of patients who receive NCRT and surgery. (Fig. 4)

Sensitivity analysis
Analysis of study sensitivity was performed by excluding one study at a time and incorporating the effect sizes of the remaining studies. The results of the meta-analysis for the OS in NCRT + PT vs.
NCRT revealed the stability of the combined total effect value, and exclusion of any of the studies did not affect the overall estimation.

Publication bias
No potential publication bias among the included studies was identified by both funnel plots and Egger's test (P = 0.278). (Fig. 6) 4. Discussion NCRT is the standard treatment strategy for advanced esophageal cancer, and PT is also an important treatment for esophageal cancer. Currently, it is not clear whether esophageal cancer (EC) patients who receive NCRT and esophagectomy need further postoperative chemoradiotherapy. Therefore, it is worth assessing whether conducting a PT after NCRT improves the OS of resected EC patients.
However, there are no published RCTs that address this question. [(ypN),: ycN + / ypN +, ycN + / ypN0, ycN0 / ypN +, and ycN0 / ypN0].The other studies have did not distinguish between the lymph node status (ypN0 or ypN+). 11 We further analyzed the survival of patients with esophageal cancer according to the different states of ypN0 and ypN+. The analysis results showed that PT might not significantly improve the OS of resected EC patients who received NCRT and had the ypN0 pathological staging. Although we showed that the NCRT + PT group have a similar or even higher ypN + patients' ratio with NCRT alone (see Table 1), the NCRT + PT group generally had a better OS. It also indirectly suggests that PT could improve the prognosis of EC patients with ypN +.
Several limitations of this meta-analysis need to be considered with caution. First, all the studies included were retrospective cohort researches. Second, the chemotherapy and radiotherapy regimens of NCRT in the various studies were not similar, and some of the studies have not detailed the PT regimens followed. Third, almost all the patients included in the meta-analysis had esophageal adenocarcinoma. Therefore, our conclusion may only apply to esophageal adenocarcinoma cases.
In conclusion, PT may improve the survival of NCRT patients with resected esophageal cancer.
However, subgroup analysis showed administration of PT to ypN0 patients did not improve the OS, while the OS of ypN + patients was improved after receiving PT. Considering that the study included an absolute advantage for adenocarcinoma patients. This conclusion may be more applicable to EAC patients treated with NCRT at the ypN + stage. Combined with the current NCCN guidelines, it could be necessary to procure PT for lymph node-positive esophageal adenocarcinoma patients who undergo NCRT combined with surgery. Considering that the researches we included are all retrospective studies, this conclusion should be confirmed by further high-level prospective studies.

Declarations
Not applicable.