Perineural invasion (PNI) is the process of nerve invasion by cancer cells; however, a consensus regarding the definition is yet to be fully established. To date, the most widely cited definition of PNI is the invasion of tumor cells in, around, and through the nerves (1). However, Veness et al. (2) suggested that tumor cells must be present inside the perineural layer in order to call PNI, while Liebig et al. (3) proposed that it is appropriate to call PNI when tumor cells are present within any of three layers of the nerve sheath or when tumor foci are present outside of the nerve with the involvement of ≥33% of the nerve’s circumference.

Despite the discrepancy regarding the definition of PNI, it has emerged as a key pathological feature for predicting prognosis in many types of cancers, including pancreatic (4), prostate (5), colorectal (6), gastric (7), biliary tract (8), vulvar (9,10) and cervical cancers (10). In breast cancer, there are few reports investigating the correlation between PNI and prognosis (11-13). In the largest cases series reported by Narayan et al., PNI was found in 15.6% of cases, and it was an independent risk factor for locoregional recurrence (LRR) of invasive breast cancer (11). However, other studies have reported that PNI was found in 25.7% (12) and 1.14% (13) of cases, respectively, and that it had no prognostic value in patients with invasive breast carcinoma. Thus, the frequency of PNI in invasive breast carcinoma varies among studies, and the prognostic significance of PNI remains unclear, to date. Therefore, we aimed to explore the correlation between PNI and distant metastasis-free survival (DMFS) or disease-specific survival (DSS) in patients with invasive breast carcinoma of no special type (NOS) and to determine whether PNI can predict prognosis.

Patients and tumor specimens. The cohort comprised 292 consecutive female patients who underwent surgical resection of invasive breast carcinoma that was diagnosed according to the World Health Organization classification of breast cancer (14) from March 2013 to March 2017 at Tottori University Hospital (Yonago, Japan). Among these patients, 66 cases were excluded due to a history of neoadjuvant treatment (n=40), bilateral breast cancer (n=15), microinvasive carcinoma (n=7), distant metastasis (n=3) and short follow-up period (n=1). Thus, 226 patients were included in the first analysis. Of the included patients, 191 patients with invasive carcinoma of NOS were analyzed for clinicopathological characteristics and prognostic value. Clinicopathological data of the patients were obtained from the hospital medical records. Data regarding the histopathological factors, including the lymph node metastasis, Ki67 labeling index (LI), estrogen receptor (ER), progesterone receptor (PgR), and human epidermal growth factor receptor 2 (HER2) statuses, were retrieved from the pathology reports. Histological grade was determined according to the Elston–Ellis’s criteria (15). ER-positive or PgR-positive status was defined as ≥1% of immunoreactive cells. HER2-positive status was determined based on the Hercep Test (Dako Agilent Technology, Santa Clara, CA, USA) scored 3+. Cases that scored 2+ were considered as HER2-positive when the presence of HER2 amplification was detected through a fluorescent in situ hybridization analysis using PathVysion kit (Abott-Vysis, Inc., Downers Grove, IL, USA). Written informed consent was obtained from all the participants, and this study was approved by the Ethics Committee of the Faculty of Medicine, Tottori University (approval number: 22A038; approval date: 9 August 2022).

Evaluation of PNI. Hematoxylin and eosin tissue sections were used for the detection of PNI by two pathologists (K.H. and Y.U.). PNI was defined as the presence of carcinoma cells within any of the three layers of the nerve sheath (11).

Statistical analysis. Categorical variables are presented as count and percentage; continuous variables are presented as mean±standard deviation. The association between PNI status and clinicopathological factors was evaluated using non-parametric tests; chi-square and Fisher’s exact tests were used for categorical variables. For the survival analysis, two different endpoints, cancer relapse (distant metastatic recurrence) and cancer-related death were used to calculate DMFS and DSS, respectively. DMFS was defined as the period from the date of initial surgery to the date of clinical or pathological cancer relapse with distant metastasis. DSS was defined as the period from the date of initial surgery to the date of cancer-related death. Patients were censored at the time of their last cancer-free follow-up or at the time of death due to reasons unrelated to breast cancer. Survival curves were computed based on the Kaplan-Meier method and tested statistically using the log-rank test. The Cox proportional hazard regression model was used to perform univariate and multivariate analyses of several factors associated with DMFS and DSS. All tests were two-sided, and p-values of <0.05 were considered statistically significant in all tests. All statistical analyses were performed using the SPSS version 25 software (IBM SPSS Statistics; IBM Corporation, Armonk, NY, USA).

Correlation between PNI and histological type or intrinsic subtype. A representative hematoxylin and eosin image of PNI is shown in Figure 1. The frequency of PNI was 14.2% (32/226) in invasive breast carcinoma cases. Of 32 cases with PNI, 27 (84.4%) were invasive carcinoma of NOS (Table I). No significant difference was observed in the frequency of PNI status according to histological type (p=0.152) or intrinsic subtypes (p=0.970) (Table I).

Clinicopathological characteristics and associations with PNI status. The clinicopathological characteristics of the 191 patients with invasive carcinoma of NOS are summarized in Table II. Mastectomy and breast-conserving surgery were performed in 87 (45.5 %) and 104 patients (54.5%), respectively. Additionally, 109 (57.1%) and 73 patients (38.2%) received radiation therapy and chemotherapy, respectively. The frequency of PNI was 14.1% (27/191) and PNI-positive status was significantly correlated with large pathological tumor size (p=0.005), the presence of lymph node metastasis (p=0.001), and lymphatic invasion (p=0.009) (Table III).

Survival analysis according to PNI status. The median follow-up period was 70 months (range=5-104). Twenty-six patients (13.6%) experienced metastatic recurrence, 18 patients (9.4%) died due to breast cancer progression and 14 patients (7.3%) died due to causes, such as other cancers (n=3), pneumonia (n=4), cardiovascular disease (n=2), cerebral hemorrhages (n=2) and others causes (n=3). The DMFS and DSS curves are shown in Figure 2. The log-rank test showed that patients with PNI-positive status had significantly shorter DMFS (p=0.002) and DSS (p<0.001). The 5-year DMFS rates in the PNI-positive and PNI-negative groups were 69.0% [95% confidence interval (CI)=0.456–0.839] and 89.9% (95% CI=0.840–0.937), respectively, whereas the 5-year DSS rates in these groups were 74.5 % (95% CI=0.517–0.877) and 93.7% (95% CI=0.885–0.965), respectively.

The univariate analysis revealed a significant correlation between shorter DMFS and PNI-positive status (p=0.003), large pathological tumor size (p<0.001), high histological grade (p<0.001), high Ki67 LI (p=0.012) and the presence of lymph node metastasis (p=0.005) (Table IV). According to the multivariate analysis, PNI-positive status had a significant effect on DMFS [hazard ratio (HR)=2.621; p=0.037], as well as large pathological tumor size (HR=2.537; p=0.038), and high histological grade (HR=3.293; p=0.021) (Table IV). The univariate analysis showed a significant correlation between shorter DSS and PNI-positive status (p=0.001), large pathological tumor size (p=0.043), and negative-ER status (p=0.031) (Table V). The multivariate analysis showed that PNI-positive status had a significant adverse effect on DSS (HR=4.463; p=0.003), as well as negative ER status (HR=3.234; p=0.022) (Table V).

Although there have been several studies to date evaluating the role of PNI as a potential predictor of prognosis in various cancers, the clinical significance of PNI in patients with breast carcinoma remains controversial. In the present study, we aimed to clarify the prognostic value of PNI in invasive breast carcinoma and revealed that PNI was an independent poor prognostic factor for DMFS and DSS.

The frequency of PNI in invasive breast cancer varies from 1.14% to 34.2% according to previous studies (11-13,16). The main reason for the discrepancy in the PNI frequency among these studies may be the difference in cohort sizes and detection methods. In the present study, the incidence of PNI was 14.1%, which is similar to that in the largest case series (15.6%) reported by Narayan et al (11). To our knowledge, the present study was the first to find that there were no significant differences in the incidence of PNI according to intrinsic subtypes of breast carcinoma.

Duraker et al. reported that vascular invasion, axillary lymph node, and PgR positivity ratios were significantly higher in PNI-positive patients than in the PNI-negative ones (12). Kapak et al. reported that PNI was associated with higher T-stages, higher tumor grades, and lymphovascular invasion (LVI) (13). We also found that PNI-positive status was significantly correlated with large pathological tumor size, the presence of lymph node metastasis, and lymphatic invasion. It has been considered that PNI is a potential pathway for dissemination and metastasis of carcinoma cells in the same way as lymphatic and vascular channels (17). Additionally, it has been suggested that PNI can be observed before LVI (3) and that it is a pathway that eventually leads to lymphatic invasion (18). Angiogenesis and neurogenesis share a number of similarities and both processes are regulated by similar neurotrophic factors and transmitters (19); therefore, further studies are needed to clarify the relationship between PNI and LVI in invasive breast carcinoma. In future studies, we plan to investigate whether PNI is an independent prognostic factor in node-negative breast cancer patients, since lymphatic invasion was considered to be the only significant predictor of distant recurrence in these patients (20).

Many reports have concluded that PNI is a poor prognostic factor in several types of cancers (4-10). To our knowledge, there are few reports that have investigated the direct relationship between PNI and prognosis in breast cancer patients (11-13). Duraker et al. reported that PNI had no prognostic value in patients with invasive breast carcinoma; however, the cohort in their study is relatively outdated (from 1996 to 1999), and the study was conducted before the emergence of anti-HER2 therapy. Moreover, the frequency of ER and PgR positivity was very low (48.8% and 50.2%, respectively). Although the incidence of PNI was very low (1.14%), Karak et al. also stated that the significance of PNI as an independent poor prognostic factor remains questionable (13). Contrary, the largest case series reported by Narayan et al. revealed that PNI was an independent risk factor for LRR in invasive breast cancer and that the increased risk caused by PNI was similar in magnitude to that observed in the case of LVI or ER/PgR negativity (11). They pointed out that the reliability of detecting nodal recurrences before the appearance of distant metastases is uncertain. Therefore, we used DMFS, instead of LRR as recurrence-related endpoint. Although the number of patients in our study is very small compared to that in the report by Narayan et al., our study revealed, for the first time, that PNI was an independent unfavorable prognostic factor for DMFS and DSS.

The limitations of our study included the small sample size, its retrospective nature, and that the therapeutic methods were not the same among patients. Future studies that stratify the patients by the intrinsic subtype, pathological stage, or nodal status are warranted to clarify the importance of PNI as a prognostic factor in breast carcinoma.

This is the first study to demonstrate that PNI is an independent poor prognostic factor for DMFS and DSS in patients with invasive breast carcinoma. Although further studies with large size cohorts are necessary, our findings suggest that PNI could be useful in predicting aggressive phenotypes in breast cancer patients.

The Authors declare that there are no conflicts of interest regarding this study.

Conception and design: K.H., M.W. and Y.U.; acquisition of data: K.H. and M.W.; analysis and interpretation of data: K.H., M.W., K.I., and Y.U.; and writing, review, and/or revision of the manuscript: K.H., M.W., K.I. and Y.U. All Authors read and approved the final version.

The Authors are grateful to Kazuko Fukushima for their excellent technical assistance with the processing of the pathological specimens.