Combining perineural invasion with staging improve the prognostic accuracy in colorectal cancer: a retrospective cohort study

Colorectal cancer (CRC) is the third most frequently diagnosed cancer and the third most common cause of cancer-related deaths in the USA [1]. Meanwhile, new cases and deaths from CRC are ranked third among all malignant tumors in China, and the incidence rate and mortality of CRC in China are increasing every year [2]. The tumor-node-metastasis (TNM) classification system is the current standard for clinical prediction of survival and recurrence in CRC. In addition, several additional risk factors are used to further stratify the risk [3, 4]. To some extent, the use of these additional stratification factors suggests that the TNM system may need further improvement and supplementation. One of these stratification factors is perineural invasion (PNI). However, current guidelines only propose the importance of the PNI in stage II CRC. The prognostic value of PNI in other stages of CRC is ambiguous [5‐7].

PNI is the invasion of nerves by cancer cells [8]. Several previous studies have suggested that PNI is a potential pathway for cancer cell dissemination and metastasis in the same manner as vascular and lymphatic channels [9‐11]. The average detection rate of perineural invasion in CRC is 17%, ranging from 8 to 42% [12]. Previous studies have suggested that PNI may assisted in selecting patients with stage II colon cancer who could potentially benefit from adjuvant therapy [13‐15], and this reminds us of the potential clinical value of PNI. However, the risk factors of PNI remain unclear. Previous studies have found that predictors of CRC with PNI include lymphovascular invasion, poor tumor differentiation, and elevated CEA levels [5, 16]. Notably, these were retrospective studies with a sample size of less than 1000. In addition, they did not explore other risk factors, such as age, since early-onset colorectal cancer (EOCRC) has aroused widespread attention.

As with unclear risk factors, the relationship between PNI and CRC prognosis remains controversial. Hu et al. reported that PNI is not an independent poor prognostic factor in patients with CRC [5]. Nevertheless, many studies have shown that PNI is a prognostic factor for non-metastatic CRC [7, 12, 13, 17‐20]. Recently, some colleagues found that PNI promotes cancer progression [21, 22], suggesting that PNI might be a new metastatic spread of CRC, independent of lymphatic or vascular metastasis. In addition, there is another hypothesis that PNI may be the source of tumor deposits, which may help to improve the staging of colon cancer [23, 24]. Although the PNI has shown increasing importance, the current guidelines only purpose practical value in stage II CRC. We investigated whether PNI has influence on the prognosis of stage III CRC and found perineural invasion in stage III colorectal cancer suggests a worse survival. These indicated combined nerve-targeted therapy with chemotherapy may improve the prognosis of stage III CRC patients with PNI.

In this context, we aimed to describe PNI features and clarify the prognostic value of the PNI in CRC, especially in different tumor stages.

Perineural invasion was defined as the presence of cancer cells in the nerves or surrounding or pass-through nerves, tumor cells close to the nerve and surrounding at least 33% of the nerve periphery, or tumor cells invading any of the three layers of the neurolemma structure [25]. PNI in CRC specimens was observed in 12.6% of the patients in our single-center retrospective study. Deborde et al. found that PNI induced cancer cell dispersion and invasion [10], while Demir et al. considered that PNI was associated with cancer pain. However, few studies had focused on the risk factors for PNI [5, 16, 26, 27]. Previous studies had found that predictors of CRC with PNI include lymphovascular invasion, differentiation, and elevated CEA levels. In this study, multivariate logistic regression analysis demonstrated that factors predicting CRC with PNI included EOCRC, colon cancer, and pMMR. Moreover, the adverse effects of PNI on survival were more significant in patients with EOCRC. This finding indicated that nerve dissection might be more appropriate for treating EOCRC. Colon cancer had a higher PNI rate in our study. The colon innervated by nerve fibers originating from inferior mesenteric ganglia and superior mesenteric ganglia (sympathetic efferent), while the rectum innervated by nerve fibers originating from pelvic ganglia (parasympathetic efferent) [28]. Besides, sympathetic nerves facilitate cancer progression [29]. This may mean that CRC cells are more likely to invade the sympathetic nervous system. dMMR CRCs are characterized by a high tumor mutation burden that leads to abundant mutation-derived neoantigens that trigger a robust immune response in the tumor microenvironment with tumor-infiltrating lymphocytes [30]. In pMMR CRC, it is more likely that the neural microenvironment plays a major role in PNI, rather than tumor-infiltrating lymphocytes.

Many clinical guidelines [3, 4] and studies [13, 31] suggested that PNI was a risk factor for stage II colon cancer. To explore which factor contributed most to the prognosis among these risk factors (PNI, incomplete resection, lymphovascular invasion, T4, poor differentiation, less lymph node harvest, and bowel obstruction), a nomogram for risk factors for OS and DFS in stage II CRC patients indicated that the weight of PNI was second only to incomplete resection (Fig. 2A and B). This indicated that PNI was an important risk factor for stage II CRC. However, Baxter et al. found that the impact of adjuvant chemotherapy on survival in stage II patients with PNI was uncertain [31]. As the prognosis of stage II CRC with PNI is poor and the effect of adjuvant chemotherapy is uncertain, there is an urgent need for a new nerve-targeted therapy to improve the prognosis of stage II CRC patients with PNI.

PNI is considered both a form of local progression and a form of metastasis, since nerve invasion may extend proximally to reach the central nervous system [32, 33]. In our study, OS and DFS were not significantly different between stage II patients with PNI and III patients (Fig. 3A and B). PNI was an independent prognostic factor for stage III CRC recurrence (Table 2 and Table S5). PNI may be a source of tumor deposits, which improves CRC staging [23, 24]. Therefore, PNI is a form of metastasis that is parallel to lymph node and vascular metastases. The position of the PNI during staging should be improved and nerve-targeted therapy is likely to improve prognosis of stage III CRC patients with PNI.

In our study, female patients or patients without diabetes combined with PNI had a worse prognosis than the corresponding patients. Cancer cells secrete vascular endothelial growth factor (VEGF) A and platelet-activating factor in response to estrogens, enhancing proliferation and migration [34]. These results indicate that female estrogen may promote the recurrence and metastasis of nerve-infiltrating tumor cells. Peripheral nerve injury is a common complication in diabetes patients [35]. Patients with diabetes are accompanied by more nerve injury, which may hamper the invasion of cancer cells to nerve system. The adverse effects of PNI were more significant in patients without diabetes due to the complete nerve microenvironment.

Our study has several limitations. First, this single-center retrospective study design increased recall and information bias, which may limit the outreach of the conclusion. The prognostic value of the PNI was best investigated in a randomized controlled trial. Another potential source of error is pathology reports from single institutions without a complete review, and the study did not require specific expertise in the review of PNI. The advantages of the study include the large number of patients with detailed clinicopathological information and the timing of death and recurrence. Another advantage of this study is that it contains a specific analysis of subgroups of patients with EOCRC, women, and patients without diabetes who had never been studied.

In conclusion, the prognosis of stage II colorectal cancer combined with PNI is consistent with the prognosis of stage III colorectal cancer. Combining PNI with stage III colorectal cancer suggests a worse prognosis. The importance of PNI for recurrence and survival of risk factors is second only to incomplete resection in stage II CRC. EOCRC is more likely to occur with PNI, which predicts worse prognosis. Female patients or patients without diabetes combined with PNI predict a worse prognosis than the corresponding patients.