Perineural Invasion Is a Significant Indicator of High Malignant Degree and Poor Prognosis in Esophageal Cancer: A Systematic Review and Meta-Analysis

Background Perineural invasion (PNI) is a malignant metastatic mode of tumors and has been reported in many tumors including esophageal cancer (EC). However, the role of PNI in EC has been reported differently. This systematic review and meta-analysis aims to focus on the role of PNI in EC. Methods Eight databases of CNKI, VIP, Wanfang, Scopus, Wiley, ISI, PubMed, and EBSCO are used for literature search. The association of PNI with gender, pathological stages of T and N (pT and pN), lymphovascular invasion (LVI), lymph node metastasis, 5-year overall survival (OS), and 5-year disease-free survival (DFS) was examined in the meta-analysis by Revman5.0 Software. The pooled OR/HR and 95% CI were used to assess the risk and prognostic value. Results Sixty-nine published studies were screened for analysis of PNI in EC. The incidence of PNI in esophageal squamous carcinoma (ESCC) and esophageal adenocarcinoma (EAC) was different, but not statistically significant (p > 0.05). The PNI-positive patients had a significantly higher risk of pT stage (OR = 3.85, 95% CI = 2.45–6.05, p < 0.00001), pN stage (OR = 1.86, 95% CI = 1.52–2.28, p < 0.00001), LVI (OR = 2.44, 95% CI = 1.55–3.85, p = 0.0001), and lymph node metastasis (OR = 2.87, 95% CI = 1.56–5.29, p = 0.0007). Furthermore, the cumulative analysis revealed a significant correlation between PNI and poor OS (HR = 1.37, 95% CI = 1.24–1.51, p < 0.0001), as well as poor DFS (HR = 1.55, 95% CI = 1.38–1.74, p < 0.0001). Conclusion PNI occurrence is significantly related to tumor stage, LVI, lymph node metastasis, OS, and DFS. These results indicate that PNI can serve as an indicator of high malignant degree and poor prognosis in EC.


INTRODUCTION
Esophageal cancer (EC) is one of the top ten malignant tumors. According to global cancer statistics in 2020, EC ranks seventh in terms of incidence and sixth in mortality overall (1). The histological types of EC mainly contain esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC). ESCC is the most common histological type in China, accounting for more than 90%. In EC treatment, surgical excision is the best treatment for the early stage, and radiotherapy and chemotherapy are often used for the middle and late stage. Recently, neoadjuvant chemoradiation (nCRT) followed by esophagectomy is increasingly applied to locally advanced EC. However, the prognosis of EC is still poor after these treatments due to its insidious and highly invasive nature in the early stage. Most EC patients are prone to relapse and metastasis. In recent years, researchers found that there is a new metastasis pathway, perineural invasion (PNI), often happening in EC patients.
PNI refers to the phenomenon of cancer cells surrounding nerve fibers and entering the surrounding nerve, spreading local infiltration and metastasis. Now, the definition of PNI is that the tumor cells are in close contact with the nerve and surround at least 33% of the nerve periphery or invade any of the three layers of the nerve sheath, which is also taken as the current pathological diagnostic criteria for PNI (2,3). The occurrence of PNI not only is with incomplete resection of the tumor and recurrence of prognosis, but also often leads to pain in many cancers, such as prostate cancer, pancreatic cancer, and head and neck cancer (2)(3)(4)(5). However, the role of PNI in EC is differently reported. For example, PNI is associated with poor overall survival (OS) and can serve as an independent factor for OS in multivariate analysis (6)(7)(8), while Lee et al. thought PNI was not an important prognostic parameter in EC (9). These inconsistent conclusions may be due to the insufficient sample size. Hence, we collected a larger number of data from EC patients and used a systematic review and meta-analysis to obtain more accurate conclusions of PNI. The study determined the association of PNI with pathological parameters, OS and DFS, and then evaluated the role and effect of PNI on EC.

Literature Search
A literature search was performed by using the CNKI, VIP, Wanfang, Scopus, Wiley, ISI, PubMed, and EBSCO databases from January 1, 1990 to March 30, 2022. The main keywords in the abstract were "perineural invasion", "esophageal", and "cancer". The articles in CNKI, VIP, and Wanfang databases only come from the Chinese Core Journal. Duplicate articles were deleted and full articles were used for analysis. The articles with patient samples from the same institution in the repeated recruitment period, reviews, and case reports were excluded. The literatures of esophageal neuroendocrine carcinoma were also excluded. The quality of all studies was assessed by using the Newcastle-Ottawa Scale and was scored from 6 to 8 (full score = 9).

Literature Extraction
To analyze the positive rate of PNI in different pathological types of EC, the following information in the articles was extracted: first author, year of publication, country of study, patient samples of recruitment period, pathological types of EC (ESCC and EAC), and the number of samples who are PNI positive.
For meta-analysis to examine the association of PNI with gender, pT stage, pN stage, lymphovascular invasion (LVI), lymph node metastasis, 5-year OS, and 5-year disease-free survival (DFS), the extracted information contains the following: first author, year of publication, odds ratio (OR), hazard ratio (HR), and the corresponding 95% confidence interval (CI). The pooled HR and 95% CI were calculated using the method of inverse variance and the p-value threshold was set at 0.05. Some articles do not directly provide HR data, but provide RR data. HR and RR can be combined because without considering the time factor in the paper, they represent the same meaning.

Statistical Analysis
Statistical calculation was completed with SPSS21.0 software and statistical heterogeneity was tested using the Chi-square test. p < 0.05 indicated statistical significance. The forest and funnel plots of the meta-analysis were made using Review Manager 5.0 software (Revman5.0). p < 0.10 or/and I² > 50% were used to indicate heterogeneity.

Literature Search Results
The systematic search identified 764 potentially eligible articles, 392 of which were excluded due to duplication. Of the remaining 372 studies, 296 were excluded because they were reviews or specimens of the same origin. Five studies were excluded for PNI recurrence after neoadjuvant therapy or chemotherapy. Finally, 71 studies were finally included in this study: 69 studies were used to analyze the positive rate of PNI in different pathological types of EC; 9, 10, 6, 8, 7, 20, and 11 studies were used to analyze the correlation between PNI and gender, pT stage, pN stage, LVI, lymph node metastasis, OS, and DFS, respectively. The detailed screening process is shown in Figure 1.

DISCUSSION
PNI was identified in a variety of malignant tumors, such as prostate cancer, pancreatic cancer, and head and neck cancer (2)(3)(4)(5). The percentage of patients with PNI in pancreatic ductal adenocarcinoma is 70%-100%, and is closely related to the occurrence of ache (3). In colorectal cancer, PNI is an   independent risk factor of recurrence, which indicates a worse phenotype of tumor (3). EC is one of the common malignant tumors with high invasion. PNI often occurs in EC, but there are conflicting reports about the effects on EC of PNI (13,16,25,28). This review and meta-analysis was conducted to better understand the relationship of PNI with the development process and prognosis of EC. The esophageal nerve includes the vagal nerve and sympathetic nerve. The abundant nerve plexus is mainly distributed in the submucosa and smooth muscle layer and is often accompanied by blood vessels and lymphatic vessels. The development of PNI implies advanced tumor staging, the depth and range of LVI, and lymph node metastasis, as reported by studies (18,30). However, other studies indicated that there was no relation between PNI and tumor staging, LVI, and lymph node metastasis (6,54). According to our study, we found that the incidence of PNI in ESCC and EAC was different, but not statistically significant. PNI had a significant association with pT stage, pN stage, LVI, and lymph node metastasis in EC, which are well-known malignant characteristics of EC (77,78). Moreover, it is worth noting that in cancer tissues with PNI, researchers not only found abundant blood vessels and lymphatic vessels, but also found angiogenesis and lymphangiogenesis, further promoting the development and metastasis of tumor (79,80). Thus, these results further suggested that PNI was an important feature for the malignant degree of cancer.
It is well known that the malignant degree of EC has a significant association with poor prognosis of EC. Faiz et al. and Noble et al. reported that PNI is positively related to poor prognosis (8,18), while Li et al. and Dong et al. identified that it was negatively related to poor outcome (21,22). We evaluated the effect of PNI on 5-year OS and 5-year DFS of EC and found that there was a statically significant association between PNI and OS and DFS. These results indicated that PNI was an independent risk factor for the prognosis of EC. However, no matter what the treatment is, PNI is also significantly associated with worse OS and DFS, and can be evaluated as a prognostic predictor (42,81).
At present, it is believed that PNI is the result of the interaction between tumor cells and nerves. The occurrence of PNI is not only closely related to the distribution of nerves in tissues and tumor progression, but also associated with the regulation at the molecular level. In ESCC PNI, studies indicated that several genes, such as NF-KB (27), P53 (60), nuclear programmed cell death 4 (PDCD4) (82), and NK1R (83), were significantly positively associated with PNI development, while an inverse correlation was found between platelet counts and PNI (84). The expression of nuclear PDCD4 can predict the prognosis of EC. Moreover, nuclear PDCD4 expression was negatively correlated with PNI (82). Substance P (SP) plays an important role in several types of cancer promotion and progression by binding to its preferential neurokinin 1 receptor (NK1R). NK1R was upregulated, and its overexpression correlated with larger tumor size, deeper tumor invasion, more PNI, and eventually caused poorer OS (83). However, the exact molecular mechanism of PNI in EC remains unclear and is worth further exploring.
In brief, PNI is a dynamic pathological process, and its underlying molecular mechanisms need to be further investigated. Our study only proves that PNI plays an important role in EC. Moreover, our results suggested that PNI can be incorporated into patient stratification factors to make more accurate surgical or treatment plans. This not only greatly improves the survival rate and prognosis of patients, but also enables the further development of precision medicine.

CONCLUSION
This review and meta-analysis was conducted to better understand the relationship of PNI with the development   process and prognosis of EC. The results indicated that the effect of PNI on poor prognosis is not isolated and associated with gene expression, especially the presence of a number of adverse prognostic factors, such as depth of invasion, clinical stage, LVI, and lymph node metastasis. In general, PNI is a significant indicator of high malignant degree and poor prognosis in EC.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

AUTHOR CONTRIBUTIONS
LB provided the idea and logic of the article, and was responsible for the data collection and statistical processing of it. LY assisted LB to complete the production of pictures and tables, and put forward other ideas. The other authors put forward valuable suggestions and revised and polished the manuscript in the process of writing the article. All authors contributed to the article and approved the submitted version.