Advantageous factors of R0 curative conversion esophagectomy and the optimal extent of lymphadenectomy after induction therapy for cT4b thoracic esophageal cancer

Abstract Aim This study aimed to clarify the prognostic factors, the advantageous factors of R0 curative resection, and optimal extents of lymph node dissection for conversion esophagectomy after induction therapy. Methods Among 1903 patients with esophageal cancer at Toranomon Hospital between January 2006 to May 2020, 151 patients with locally advanced T4b thoracic esophageal cancer were divided into two groups according to treatment: conversion surgery group (n = 54) and non‐surgical treatment group (n = 97) for comparison. Results The patients who underwent R0 curative resection showed preferable survival comparable to the survival rate of patients with cCR in the non‐surgical treatment group (1‐, 3‐ and 5‐year survival: 96.9%, 82.1% and 76.7% vs 94.1%, 86.3%, and 86.3%; P = 0.770). Multivariate analysis revealed that the T4b tumor invasion by primary site (odds ratio (OR) = 6.100; 95% CI, 1.439‐25.865: P = 0.014) and time to conversion surgery from start of induction therapy within four months (OR = 5.229; 95% CI, 1.296‐21.102: P = 0.020) were all independent advantageous factors of R0 curative resection. Actuarial 1‐, 3‐ and 5‐year survival rates in patients who underwent conversion surgery with D2‐3 lymphadenectomy were 90.9%, 48.6%, and 40.8%, respectively. Conclusions R0 resection led to improved prognosis in conversion esophagectomy for cT4b esophageal cancer. The T4b tumor invasion by primary site and time to conversion surgery from start of induction therapy within 4 months were independent advantageous factors of R0 curative resection. In addition, standard radical esophagectomy including prophylactic D2‐/3‐ lymphadenectomy should be performed if it is possible, while taking adequate care regarding the increased risk after induction therapy.


| INTRODUC TI ON
The standard treatment for stage II and III esophageal cancer is neoadjuvant chemotherapy (NACT) or neoadjuvant chemoradiotherapy (NACRT) + radical surgery. However, definitive chemoradiotherapy (DCRT) with organ preservation is an option for esophageal cancer and has become one of the most common nonsurgical treatments for locally advanced esophageal cancer particularly when difficulty in R0 resection is suspected. In our hospital, we tend to opt for DCRT for patients with unresectable cT4b cancer. However, the complete response (CR) rate of patients with cT4 cancer who received DCRT was about 17%-39%, which is considerably lower than that of T3 cases; 64%-69.2%. [1][2][3][4] Therefore, if the tumor invasion is relieved by induction chemotherapy or chemoradiotherapy, conversion esophagectomy emerges as one of the radical treatment options.
However, a suitable treatment strategy for cT4b cases has not been established and remains unclear. The aim of the present study was to clarify the effectiveness of conversion esophagectomy after induction therapy, the advantageous factors of R0 curative conversion esophagectomy and the treatment strategies such as extent of lymphadenectomy especially for locally advanced T4b esophageal cancer.

| Study population
In this single-center retrospective study, a total of 1903 consecutive patients with esophageal cancer were identified from a database that was prospectively constructed at Toranomon hospital between January 2006 and May 2020. Among these, 151 patients who had thoracic esophageal cancer invading adjacent vital structures (aorta and/or trachea/bronchus) without distant organ metastases were selected for this study. Patients with cervical lymph node metastases were included. We indicate by "bulky T3," a T3 tumor which has the potential of direct invasion to adjacent organs but safe resection is highly probable regardless of microscopic radial margin diagnosis. We think it is very difficult to clearly differentiate resectable bulky T3 and real T4 tumors. In this study, we selected only T4b tumors which were clearly diagnosed before treatment. We did not include the patients with bulky T3 tumors. We present an example of a patient with apparent T4b selected in this study in Figure 1. These 151 patients were divided into two groups according to treatment: conversion esophagectomy group (n = 54) and non-surgical treatment group (n = 97) for comparison. We evaluated the prognostic factors, surgical indications, optimal extent of lymph node dissection and the effectiveness of conversion esophagectomy compared with non-surgical treatment. Furthermore, we analyzed the advantageous factors of R0 curative resection by comparing the R0 resection group and the R1/2 resection group using univariate and multivariate analysis. We judged whether or not R0 resection had occurred by pathological findings after operation in the conversion esophagectomy group. We investigated the association between these significant factors and the tumor differentiation for the interpretation of these advantageous factors. Assessment of invasion of adjacent structures was performed using enhanced computed tomography (CT) scan, magnetic resonance imaging (MRI), endoscopy, endoscopic ultrasonography (EUS) and bronchoscopy before treatment and before surgery. EUS was an excellent method for evaluating tumor invasion of aorta and trachea/bronchus, but sometimes the endoscopy could not pass the tumor because of stenosis caused by the esophageal tumor. In contrast, enhanced CT scan was applied to all the patients even when the esophageal lumen was stenotic. The overall circumference of contact between the tumor F I G U R E 1 Patient who underwent R0 curative resection after induction therapy (5-fluorouracil+cisplatin:FP + 40 Gy). A, Primary tumor and metastatic lymph node invading the aorta before induction therapy in non-contrast-enhanced computed tomography (CT) scan. B, Primary tumor and metastatic lymph node invading the aorta before induction therapy in contrast-enhanced CT scan. C, The tumor immediately after induction therapy. D, The tumor after about 4 wks after induction therapy and the aortic wall has been shown to be a useful predictor, with an interface arc greater than 90 degrees, suggesting invasion, as reported by Picus et al. 5 Bronchoscopy was a useful method for evaluating tumor invasion of the trachea/bronchus. These modalities were used to comprehensively assess the presence or absence of tumor invasion of aorta and/or trachea/bronchus at a multidisciplinary conference (surgeons, gastroenterologists, oncologists, radiologists, and pathologists). Diseases were staged according to the UICC TNM grading system, 7th edition. 6 Although direct invasion of metastatic lymph nodes to the aorta and/or trachea/ bronchus are not clearly included in the UICC TNM system, it was included in this study as cT4b. All postoperative complications were graded based on the Clavien-Dindo classification (CDc), 7 and grade ≥ III events were documented as complications. This study was conducted with approval from the Institutional Review Board of Toranomon Hospital (approval number 2026).

| Non-surgical treatment
In the present study, induction therapy included chemoradiotherapy (CRT), chemotherapy (CT) alone, CT followed by CRT (CT-CRT), and CRT followed by CT (CRT-CT). The chemotherapy regimens in the induction CRT protocol were either 5-fluorouracil (5-FU) 700 mg/m 2 on days 1-4 and cisplatin 70 mg/m 2 on day 1 (FP) or docetaxel 75 mg/m 2 on day 1, cisplatin 75 mg/m 2 on day 1, and 5-fluorouracil 750 mg/m 2 on days 1-5 (DCF). In the radiation protocol, fraction of 2 Gy/day was given up to 40-60 Gy. DCRT was defined as chemotherapy combined with ≥50.4 Gy of radiation to the main tumor and detected metastases, and more than 40 Gy of prophylactic radiation to the regional lymph nodes. Induction CT regimen was either FP (800 mg/m 2 5FU on days 1-5, 80 mg/m 2 cisplatin on day 1) or DCF (75 mg/m 2 docetaxel on day 1, 75 mg/m 2 cisplatin on day 1, 750 mg/m 2 5 FU on days 1-5). Induction CT-CRT or CRT-CT protocol was mixed these chemotherapy and chemoradiotherapy protocols mentioned above.

| Tumor response
The clinical response was assessed by endoscopy with esophageal biopsy, computed tomography (CT, with PET-CT in some cases), and ultrasonography of the abdomen and neck. Especially, we diagnosed whether or not there was T4b relief after induction therapy using the same modalities before treatment such as enhanced CT, MRI, bronchoscopic and endoscopic examinations. The clinical response was determined using the RECIST (Response Evaluation Criteria In Solid Tumors) guidelines 8,9 and the 11th edition of the Japanese Classification of Esophageal Cancer parts II and III. 10 Responses were classified as follows: CR, PR (partial response), PD, (progressive disease), and SD (stable disease). We selected conversion esophagectomy when the tumor invasion is estimated to be relieved by these induction therapies. Our basic policy for management of patients presenting with cCR is patients' self-decision-making after receiving detailed informed consent regarding the risks and benefits

| Surgical procedure
We carry out esophagectomy with two-or three-field lymph node dissection depending on the degree of progression and surgical risk involved. 11 The operative thoracic approach is by video-assisted thoracoscopic surgery (VATS) or thoracotomy, and the abdominal approach is hand-assisted laparoscopic surgery (HALS) or open laparotomy depending on individual cases. We generally resected the thoracic duct (TD) in cases with cStage ≥ II for the purpose of lymphadenectomy. However, we try to preserve TD in patients with high risk particularly in hepatic or pulmonary function. We preserved bilateral bronchial arteries to maintain the bronchial blood flow, although they were resected when we suspected tumor invasion to the bronchial artery by primary tumor or metastatic lymph node. A manually sutured esophagogastric or esophago-ileal anastomosis in the neck was fashioned for all patients. We used three groups of lymph node basins defined in relation to the main tumor location by Japan Esophageal Society to describe the extent of lymph node dissection in an esophagectomy: D0 dissection: no or incomplete dissection of group 1 lymph nodes; D1 dissection: complete dissection of group 1 lymph nodes, but no or incomplete dissection of group 2 lymph nodes; D2 dissection: complete dissection of group 1 and group 2 lymph nodes, but no or incomplete dissection of group 3 lymph nodes; and D3 dissection: complete dissection of groups 1, 2, and 3 lymph nodes. 10

| Determination of treatment policy
In all the T4b cases, we assessed the esophageal cancer itself and the appropriate management of therapy at a multidisciplinary conference (surgeons, gastroenterologists, oncologists, radiologists, and pathologists). Therefore, at this conference, our hospital decided on the chemotherapy regimen, radiation dose, radiation field, and propriety of conversion surgery. Finally, management of therapy was decided by the patient after detailed informed consent was provided regarding the risks and benefits of each treatment option.

| Statistics
Cumulative rates of overall survival (OS) were analyzed by the Kaplan-Meier method. Prognostic factors involved in OS were evaluated using the log-rank test. In multivariate analysis, variables associated with OS were identified using stepwise Cox proportional hazards models. Variables identified using simple Cox proportional hazards models were selected for potential association with survival based on our clinical experience. Variables with significance of P < 0.05 in the simple Cox proportional hazards models were included in multifactorial Cox proportional hazard models. In multiple Cox hazards models, P < 0.05 was considered significant. All statistical analyses were carried out using Statistical Package for the Social Sciences (SPSS) version 19.0J for Windows (SPSS Inc., Chicago, IL, USA).  (Figure 2).
In contrast, if we could not perform R0 curative resection in the conversion surgery group after induction therapy, the survival rate did not differ significantly between the non-R0 resection in conversion surgery group and the non-CR in non-surgical treatment group ( Figure 2). The survival rate of patients who did not undergo R0 resection was significantly higher than that of patients with non-cCR in the non-surgical treatment group (P = 0.013  Table 3). Survival curves for patients who underwent conversion surgery within 4 months from start of induction therapy compared with those who were operated later than 4 months are shown in Figure 3A. The survival rate of patients who underwent conversion surgery within 4 months from start of induction therapy was significantly higher than that of patients who underwent conversion surgery later than 4 months (P = 0.017).
Survival curves for patients with the T4b tumor invasion by primary site compared with invasion by metastatic lymph node are shown in Figure 3B. The survival rate of the patients with T4b tumor invasion by primary site was significantly higher than that of patients with invasion by metastatic lymph node (P < 0.001). Table 4 shows the relationship between the origin of tumor invasion and histological type. The rate of poorly differentiated tumor was significantly higher in patients with invasion by metastatic lymph node as T4b reasons than that of primary tumor invasion (P = 0.005). In the present study, the survival rate of patients with poorly differentiated tumor was significantly lower than that of patients with well/moderately differentiated tumor (P = 0.047; Figure 3C).  Previous studies have reported that conversion surgery after induction therapy can be a potentially curative treatment option for selected patients with cT4b esophageal cancer. 12 The present study provided a similar result in that the patients who underwent Poorly differentiated 5 10

F I G U R E 4
The 42 patients who underwent two-or threefield lymph node dissection including prophylactic dissection. A, Regional lymph node metastasis (-). B, Regional lymph node metastasis (+ to neoadjuvant therapy is a nonstandard approach, but it has become more widely practiced with the advent of total neoadjuvant therapy and with increasing demand by patients in the context of a cCR. 19 However, we retrospectively analyzed the prognosis of esophageal cancer patients who achieved cCR after neoadjuvant therapy by making a comparison between the esophagectomy group and the non-surgical group. 20 These findings show that recurrernce-free survival and disease-specific survival were significantly better in patients who underwent esophagectomy than in patients who received nonsurgical treatment. Conversely, OS did not differ significantly between the two groups because of the higher risk of late effects (eg, respiratory complications after NACRT such as aspiration pneumonia/respiratory failure). 20 Further advances in the diagnostic accuracy of treatment modalities have increased the possibility of making a definitive diagnosis of cCR, with the choice of "Watch and Wait strategy" as a viable option.
The largest bias of this study is that the intraoperative decision to perform extended lymph node dissection should have been largely affected by intraoperative judgement of the curative status of the surgical margin. Our study has a single-center retrospective design and therefore there is a large selection bias from the stage of decision-making of treatment options, although the present data are based on a prospectively collated database for consecutive patients over a relatively short period. An external validation prospective study involving a sufficient number of patients would be needed to confirm our observations; a multicenter study with a larger number of cases is also warranted.

| CON CLUS IONS
In conversion esophagectomy for T4b esophageal cancer after induction therapy, R0 curative resection was found to be a favorable prognostic factor. The T4b tumor invasion by primary site and time to conversion surgery from the start of induction therapy within 4 months were all independent advantageous factors of R0 curative resection. In addition, standard esophagectomy including prophylactic 2-or 3-field lymphadenectomy should be performed if it is possible, while taking adequate care to prevent postoperative complications and to avoid damage to the aorta and trachea. An important issue for further research is to establish a method for more accurately diagnosing tumor resectability after induction therapy for cT4b esophageal cancer.

D I SCLOS U R E
Conflicts of Interest: None declared.