Treatment Strategy for Esophageal Squamous Cell Carcinoma With Endoscopic Intramural Metastasis

Purpose: Intramural metastasis (IM) in esophageal squamous cell carcinoma (ESCC) is sometimes found, and the prognosis of ESCC patients with pathologically diagnosed IM is known to be dismal. However, there are few reports on ESCC patients with clinically diagnosed IM. Methods: This study assessed 2,772 ESCC patients who underwent endoscopy for initial evaluation. Among them, 85 patients (3.1%) were diagnosed with endoscopic IM. In this study, we investigated these patients’ characteristics, survival among the groups stratified by the treatment modalities, and survival predictors. Results: Of 85 patients, 76 (89.4%) had T3 or T4 tumors, 73 (85.9%) had nodal metastases, and 36 (42.4%) had M1 diseases. Curative-intent treatment could be given to 63 patients (74.1%) with a median survival time (MST) of 15.6 months (95% CI: 10.7-20.4). As initial treatment, upfront surgery (US), neoadjuvant chemotherapy (NAC) using cisplatin and 5-fluorouracil (CF), neoadjuvant chemoradiotherapy, and definitive chemoradiotherapy (dCRT) were given to 17 (27.0%), 27 (42.9%), 2 (3.2%), and 17 patients (27.0%), respectively. dCRT was preferred for T4 tumors compared with US or NAC (P = 0.02). The MST of US and NAC patients was 19.3 (95% CI: 12.9-25.6) and 23.4 months (95% CI: 9.4-37.4), respectively. No significant difference was noted between US and NAC patients (P= 0.89). Conclusion: The prognosis of ESCC patients with endoscopic IM is poor even if curative-intent treatment is done. Moreover, no significant survival benefit of NAC with CF for these patients was observed when compared with US.


Introduction
Pathologically diagnosed intramural metastasis (IM) of esophageal squamous cell carcinoma (ESCC) is sometimes found with an incidence rate of 5.5-16.6%, and the prognosis of these patients was suggested to be dismal [1][2][3][4][5][6][7]. In our previous study, pathological IM has shown to be an indicator of lymphatic invasion and advanced cancer in ESCC patients [7]. Previous reports on IM were mainly on its pathological diagnosis after surgery [1][2][3][4][5][6][7], and few on the clinical features of ESCC patients with clinically diagnosed IM by endoscopy. Approximately half of the pathological IM cases were identified during preoperative examination [6,7], and the characteristics and outcomes of ESCC patients with clinically diagnosed IM remain unknown.
Multimodal treatment with a significant antitumor effect is required to improve outcomes of patients with IM; however, there is no determined standard treatment for ESCC patients with clinically diagnosed IM. Hokamura et al. evaluated the efficacy of neoadjuvant chemotherapy (NAC) that consisted of cisplatin and 5fluorouracil (CF) for ESCC with IM diagnosed by endoscopy [8]. They reported no NAC survival benefit when compared with their historical data. However, this study conducted a small-sample single-arm trial, with only 15 patients.
This study aimed to analyze the characteristics of ESCC patients with clinically diagnosed IM and assess their outcomes stratified by the treatment modalities and survival rates to explore the treatment strategy suited for these patients.

Definition of IM
Based on the endoscopic findings, clinical IM was characterized by the following (Figure 1): (1) separated from the primary tumor, (2) located in the esophagus or stomach wall, and (3) with the gross appearance of a submucosal tumor without intraepithelial component, occasionally with an erosive change [2]. Accompanied with these findings, the presence of an advanced primary lesion in the esophagus strongly suggests an IM tumor. Biopsies or boring biopsies to confirm cancer cells in the submucosal tumor were not performed routinely. Distinguishing a second primary lesion after it has grown to an elevated and ulcerated tumor is challenging; hence, such tumors from IM were excluded. Pathologically, IM was diagnosed as described in the previous study [7].

Treatment strategy for ESCC
All patients had undergone routine evaluations including endoscopy and computed tomography (CT). In patients showing T2 or higher T stage or possible metastatic nodes on CT, positron emission tomography was done. Endoscopic and endobronchial ultrasounds to evaluate clinical nodal involvement and extent of invasion were not performed routinely. Tumor staging was done, and treatment was decided with the multidisciplinary tumor board. In this study, the tumor stage was determined based on the Union for International Cancer Control TNM Classification of Malignant Tumours, 8th edition [9]. IM within the gastric wall was defined as distant metastasis (M1).
Patients were treated according to the Japan Esophageal Society guidelines as follows [10,11]: upfront surgery (US) for stage I disease, definitive chemoradiotherapy (dCRT) for T4b tumor or refusal to surgery irrespective of the stage, and salvage surgery for dCRT failure. However, during this study, treatment for stage II or III disease has changed owing to the result of the randomized trial JCOG9907 [12], and neoadjuvant chemotherapy (NAC) was introduced in 2009. NAC consists of two courses of CF: 80 mg/m2 cisplatin was given on day one and 800 mg/m2/day 5-fluorouracil on days one to five. Before that, NAC was not available. In the case of the tumor suspected to invade adjacent organs, but not definitely diagnosed as T4b disease, neoadjuvant chemoradiotherapy (NACRT) with CF and 40Gy irradiation was introduced. In this study, US, NAC, NACRT, and dCRT were defined as curative-intent treatment and chemotherapy alone, radiation alone, chemotherapy with palliative radiation, and best supportive care as palliative treatment, respectively. Supraclavicular lymph node metastases (M1) or IM within the gastric wall were not the contraindications of surgery.

Statistical analysis
All data are presented as median (range) or number (%). The Mann-Whitney U test and Fisher's exact test were appropriately used for statistical comparisons between groups. The overall survival was for the period from the date of diagnosis to either death or the last follow-up, and survival analysis was performed using the Kaplan-Meier method and log-rank test. A Cox proportional hazard model was used to evaluate the impact of variables on survival, and the hazard ratios (HRs) and 95% confidence intervals (CIs) were obtained. In the multivariate analysis, we included potential confounders, including sex, age, performance status, initial clinical T stage, N stage, M stage, and the treatment modalities. All statistical analyses were performed using the SPSS software package (version 23.0; IBM Corp., Armonk, NY). A P-value < 0.05 was considered statistically significant.

Impact of each treatment modality with curative-intent on survival
We investigated the clinical impact of treatment modalities on survival in patients who received curativeintent treatment, excluding NACRT patients (

Accuracy of clinical diagnosis for IM by endoscopy
Among the 17 US patients who were diagnosed as having IM by endoscopy, three patients (17.6%) did not have IM in pathological findings, indicating that the positive predictive value for clinical diagnosis of IM by endoscopy was 82.4%.

Pathological IM and preoperative treatment in patients who underwent surgery
Finally, we investigated the clinical impact of pathological IM on survival stratified by preoperative treatment in 42 patients who underwent surgery; 17 of US patients, 22 of NAC, one of NACRT, and two of dCRT, respectively. Patient survivals were calculated in pathological IM-positive patients (n = 13) and IMnegative patients (n = 12) who received preoperative treatment and US patients with pathological IM (n = 14). Despite the pathological IM-negative patients who received preoperative treatment possibly included patients with false-positive clinical IM, they attained long-term survival of approximately 50% (Figure 3). Also, the percentage of patients who did not have pathological IM in the NAC group (12 of 25, 48.0%) was higher than that in the US group (three of 17, 17.6%) significantly (P = 0.044).

Discussion
In this study, the clinical characteristics of patients with clinically diagnosed IM from ESCC and the impact of treatment modalities on survival were evaluated. It was observed that advanced tumors were more common in these patients, and the prognosis was poor. NAC with CF for these patients had no significant survival benefit when compared with US. Additionally, when US could be performed as a curative-intent treatment, the prognosis might improve compared to palliative treatment. Patients wherein pathological IM was not detected after preoperative treatment attained long-term survival of approximately 50% although they might include those with false-positive clinical IM.
In 1933, Watson firstly reported IM from esophageal cancer as cancer spreading through the submucosal lymphatic system [13]. It was reported that IM from ESCC is one of the important poor prognosticators and the reliable indicator for lymphatic invasion [1][2][3][4][5][6][7]. Although the IM in ESCC is not a factor to determine tumor, node, and metastasis (TNM) classification, it suggests highly aggressive biological behavior with a strong prognostic impact. Thus, multimodal treatment with a strong antitumor effect should be given.  [8]. Moreover, it has been shown that clinical IM is significantly associated with treatment failure in patients wherein NAC with CF was given [14]. Therefore, NAC with CF might be insufficient for these patients. Recently, the JCOG1109 trial showed the superiority of triplet NAC (docetaxel plus CF) over CF in terms of survival [15]. Also, in the CheckMate 577 trial, adjuvant immune checkpoint inhibitor (ICI) offered a better prognosis in ESCC patients who underwent NACRT followed by curative esophagectomy [16]. Henceforth, these promising treatment strategies should be evaluated.
Although NAC presented no significant survival benefit, 48.0% of patients whose pathological IM could not be detected after preoperative treatment achieved long-term survival. This study has several limitations. First, this was a retrospective observational study with a limited number of patients in a single institution. Second, as mentioned above, NAC with CF was introduced after 2009, suggesting that change in protocol influenced the results. However, despite advances in treatment strategy and management for ESCC, NAC, which is a recent treatment, did not show a survival benefit for these patients when compared with US. Third, this study was based on the past clinical diagnosis of IM, and the diagnostic accuracy and detailed findings of IM were insufficient. Therefore, a more extensive multicenter cohort study with a more accurate diagnosis of clinical IM is essential to validate our findings.

Conclusions
The prognosis of patients with a clinical diagnosis of IM from ESCC is poor even when curative-intent treatment is performed. Furthermore, NAC with CF does not contribute to prolonging survival when compared with US. Treatment strategies with more antitumor effects would be required for these patients.

Additional Information Disclosures
Human subjects: Consent was obtained or waived by all participants in this study. Institutional Review Board of Japanese Foundation for Cancer Research issued approval 2018-1192. This study was approved by the Institutional Review Board of Japanese Foundation for Cancer Research (number 2018-1192). Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue. Conflicts of interest: In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.