Introduction

Pancreatic ductal adenocarcinoma (PDAC) and cholangiocarcinoma (CCA) are aggressive cancers with only a limited response to chemotherapy. PDAC mortality is estimated to exceed the total breast, prostate, and colorectal cancer deaths and be the second leading cancer-related death by 2030 [1, 2]. PDAC and CCA share many clinical characteristics, which include high mortality rates and low treatment efficacy [3]. Unfortunately, survival rates have not improved even from recent novel therapeutic targets such as immune checkpoints [3,4,5,6,7]. Biologically PDAC and CCA are characterized by desmoplastic stroma and this stromal compartment is thought to be held responsible for the poor efficacy of chemotherapy. Today, surgery combined with chemotherapy is the only chance of cure [8].

The tumor microenvironment (TME) is a fervent area of research interest as it contains a host of nonmalignant cells that play an important role in carcinogenesis such as fibroblasts, immune cells, blood- and lymphatic vessels, and nerve fibers. In this review, we will focus on the pathways involved in neurogenesis and the interaction between nerve fibers and the other components of the TME.

Internal organs are innervated by the autonomic nervous system (ANS), which is composed of two components: the sympathetic nervous system (SNS) and the parasympathetic nervous system (PSNS). Increasing evidence shows that not only the internal organs are innervated by the PSNS, but solid tumors also depend on the development of nerves in the TME for growth and invasion in adjacent tissue [9,10,11].

Besides the aggressive behavior and poor response to treatment, another shared feature of these two cancer types is perineural invasion (PNI), which is defined as cancer cells surrounding at least 33% of the epineurial, perineural, and endoneurial space of the nerve sheath [12]. PNI describes the process of cancer cells invading the nerve, crossing all layers of the nerve sheath. Once the cancer cells are invaded in the nerve, they have reached a favorable environment to travel intraneural and contribute to the progression of the disease. Over time different definitions for PNI have been used. It has been described as cancer cells located in the endoneurium associated with the Schwann cells [13] or later on as the presence of cancer cells along one of the layers of the nerve sheath [12, 14,15,16]. For pathologists invasion in one of these nerve sheath layers is used to report PNI and often a mixture of invasion in different layers is seen in one histological slide. Intraneural invasion in PDAC has been associated with higher frequency of local/distant recurrence when compared to cases with PNI but without intraneural invasion [17]. This provides some evidence that defining the level of PNI matters clinically but up to now this is not recommended in the guidelines for the pathologists. The exact underlying mechanism of PNI remains unknown [12, 18]. A hypothesis is that the nerve fibers choose the path of “least resistance” and the cancer cells move along this low resistance path [14, 15]. A recent insight showed that PNI was activating signaling pathways when cancer cells attacked the perineural spaces of the surrounding nerves [12]. Even though PNI commonly occurs in many solid tumors [19,20,21,22], PDAC and CCA are “neurotropic cancers” and have a high frequency of PNI [23]. It has been reported that almost all PDAC lesions contain PNI and about 75% of CCA lesions showed PNI [23,24,25] (Fig. 1 presents the classical PNI pathological characteristics of PDAC and CCA).

Fig. 1: Histology of perineural invasion in neurotropic cancer.
figure 1

Histology slide of PDAC (a) and CCA (b) in Hematoxylin and Eosin (H&E) showing PNI, which is one of the shared pathological characteristics of both cancers. a PDAC slide with extended PNI and an almost identical histomorphology as CCA. b CCA with tumor cells massively invading the perineural space, surrounded by desmoplastic stroma and few small tumor glands in the stroma.

Full size image

A novel biological phenomenon is the cancer-related neurogenesis, which is described in prostate cancer. The nerve fiber density is increased in paraneoplastic and neoplastic prostate lesions [26]. It is not known whether this cancer-related neurogenesis also occurs in PDAC or CCA. Exploring the role of alterations in nerve fibers in PDAC and CCA has the potential to be of importance in develo** personalized medicine and finding an effective novel treatment strategy.

In this review, we aim to provide an overview of the current knowledge about nerve fiber crosstalk with cancer, and other components of the TME in PDAC and CCA.

Innervation and neurotransmitters

There is a complex nerve fiber network distributed around the pancreas, retroperitoneum and the biliary tree. Nerve fibers in the PDAC TME include axons originating from the sympathetic, parasympathetic, enteropancreatic or hepatic plexus, afferent nerve fibers and newly developed nerve fibers [94]. It was described that some MMPs (MT1-MMP) are shown to degrade the ECM and promote pancreatic cancer expansion, invasion, and progression to an advanced stage [95, 96]. It is reported that MMP9 is associated with more lymph node metastasis and a poorer survival in breast cancer [97]. PDAC cells in vivo undergoing chronic stress, were sensitive to neural signaling and pancreatic stromal cells were increased. This promoted tumor metastasis and cancer progression, β-adrenergic receptor blockade intervention therapy can block this neural signaling and be part of a combination therapy for PDAC [63].

Pathological features of nerve fibers in PDAC and CCA

PNI is considered as an important factor for poor prognosis in PDAC and CCA [24, 98]. It has been shown that PNI can be the reason for curative resection failure (shown in Table 1). Chatterjee et al. showed by examination of 212 PDAC slides, that the presence of PNI was directly correlated with tumor size, margin status, lymph node metastasis, and AJCC stages and inversely associated with disease free survival (DFS) and OS [17]. Shimada et al. found that the degree of intrapancreatic nerve invasion can be used as a predictor for recurrence of disease after surgery [99]. A phenomenon termed as “neural remodeling” is postulated in PDAC, characterized by the alterations in morphology of the nerve [18, 100,101,102]. It was shown that nerve fiber alterations including hypertrophic nerves, increased nerve fiber density and pancreatic neuritis were strongly associated with GAP-43 overexpression and abdominal pain [103]. In the perineural space, PNI induces reactive alterations in the morphology and function of the nerves. Morphological changes include changes of the nerve trunk and thickness [102]. The aggressiveness of PNI is related to neural remodeling, desmoplasia and cancer pain. Severe and enduring pain was strongly associated with poor survival in PDAC patients. However, these neural alterations did not have a significant association with survival [103]. NGF and Artemin play a fundamental role in neural modeling in pancreatic adenocarcinoma.

Table 1 Summary of current research neural remodeling in PDAC and CCA.
Full size table

Interestingly, lower intrapancreatic neural density in the tumor area was linked to shorter OS with multivariate analysis [73]. Related research in CCA mainly focused on PNI prevalence and patient survival [45, 99, 104]. In summary, these studies consistently showed that the presence of PNI was linked to shorter survival in patients with PDAC or CCA. The significance that PNI is an independent prognostic factor of poor outcome has been demonstrated. However, the influence of nerve remodeling especially the new outgrown small nerve fibers has not been fully explained.

Conclusions and future directions

With its frequent PNI, PDAC and CCA are two neurotropic cancers. The neurotropism of these cancers could be an explanation of their aggressiveness and poor response to treatment. In this review, the progress of recent research in the mechanism of PNI in PDAC and CCA is discussed. However, the crosstalk between the nervous system in PDAC and CCA is undiscovered. Different nerve fibers have a different function and the interaction with the components of the TME and the cancer are important to investigate.

In PDAC, the role of nerve fibers is divergent and nerves from the PSNS and SNS have a cancer stimulating and cancer inhibiting effect. To our current knowledge, detailed understanding of the underlying mechanisms of tumor and nerve fiber interaction is critical for the development of innovative therapeutic strategies for patients with these highly lethal cancers. The nerve outgrowth is part of the TME, in which cancer to stroma crosstalk takes place. It is likely that other components of the TME also influence the nerve outgrowth and immune cells and fibroblasts are key components in this process. Targeting nerves has the potential to be a new strategy for therapy for PDAC and CCA patients by influencing the TME, immune cells and fibroblasts, potentially influencing sensitivity to therapeutics. The newly formed nerve fibers are different from the more commonly used PNI. From our perspective, PNI originates in the pre-existing nerve fiber networks and the cancer uses the distribution network for cancer growth. This is a well-known sign of aggressive disease and is associated with poor survival. Small nerve fiber outgrowth can be used as a biomarker for a better survival, as a tool to stratify patients for treatment and as a target for therapeutics. More research is needed to investigate whether sensitivity to already existing immunotherapy can be achieved by targeting nerves.