Mixed ductal-neuroendocrine neoplasms of pancreas are defined as tumors composing of ductal adenocarcinomas and neuroendocrine neoplasms, with each part accounting for at least 30%. To date, the final pathological diagnosis relies mainly on postoperative specimen. To our knowledge, this was the largest cohort worldwide of pancreatic mixed ductal – neuroendocrine neoplasms diagnosed by EUS-FNA specimen.
EUS-FNA is often the first-line diagnostic tool used in the evaluation of solid pancreatic masses with a reported sensitivity of 85%-95%, specificity of 95%-98%, and diagnostic accuracy of 78%-95% [38, 39]. In general, the cytomorphologic feature of adenocarcinomas and NENs is easy to distinguish, however, the accurate cytological diagnosis of mixed tumors is limited by a potential sampling bias and overlapping the cytological characteristics, that should be taken a particular caution [5, 40–42]. Immunohistochemical stain should be considered when encountering a tumor showing mixed morphological features, particularly when the two components are poorly differentiated [40, 43]. The immunohistochemical markers of ductal differentiation include CK-7, CK-8, CK-18, CK-19, CA19-9, CEA, CA125, B72.3, MUC1 and MUC5AC, whereas neuroendocrine differentiation is identified by expression of synaptophysin, chromogranin A, CD56 and hormone production [2]. Tian et al. [44] reviewed 10 cases of pancreatic MiNENs in which an EUS-FNA and surgical treatment were performed, only 3 (30%) were accurately diagnosed by biopsy. Similarly, Sullivan et al. [5] revealed that only 2 out of 9 patients (22.2%) of mixed acinar-endocrine carcinoma were accurately diagnosed by EUS-FNA specimens before surgical resection. Of the previous reported cases of mixed ductal – neuroendocrine neoplasms, only 2 cases had two components observed in preoperative biopsy specimens [21, 33]. Furthermore, FNA cytology or biopsy cannot determine whether each component accounts for at least 30%, so it is difficult to make a definite diagnosis of MiNENs. When we encounter the tumor showing such mixed morphology features, we can diagnose combined tumor with mixed ductal and neuroendocrine features, suspicious of MiNENs.
Evidence based management strategies of pancreatic MiNENs have not been established because of the limit number of cases reported. Surgery is the treatment of choice in nearly all patients with localized MiNENs, which is also performed in approximately a quarter to one third of patients with advanced disease [45]. In unresectable cases, EUS-FNA is an important diagnostic tool to guide the clinical treatment. Some clinicians pointed out that treatment strategies for MiNENs should be based on the more dominant component (adenocarcinoma or NEC) [46, 47]. As most of the cases with a high grade neuroendocrine component, it is usually treated as pure NEC [47]. Kaji et al. [37] expressed that the NEC-targeted chemotherapy for MiNENs had been effective. EP (etoposide and cisplatin) therapy has been the most widely used in GEP NEC, which is typically followed by the treatment of the small-cell lung cancer[48]. Other drugs including oxaliplatin, temozolomide, and gemcitabine can also be assessed in mixed neuroendocrine with non-neuroendocrine component [49]. It has been reported that the molecular targeted drugs such as everolimus and sunitinib showing significantly prolonged progression – free survival among patients with advanced PanNETs [50]. In our cases, all of three patients targeted the predominant NENs component for treatment. Patients 2 and 3 achieved long-term survival with unresectable pancreatic tumor, which are far longer than the median overall survival of the current cohort. It suggests that our treatment strategy has been effective. hence, if the neuroendocrine component of these tumors was classified as NET G2, we propose appropriate adjuvant therapy rather than surgery.
The pathogenesis of MiNENs remains controversial and current molecular/genetic studies showed that the two components may have a common monoclonal origin [51–53]. In addition, the neuroendocrine components usually have a higher number of aberrations and a higher allele imbalance than non-neuroendocrine components, suggesting a more aggressive tumor biology [45]. The prognosis of MiNENs is not well known and the median overall survival (OS) ranging between 10.5 and 78 months [45]. Some scholars believed that the survival outcome of MiNENs was close to that of pure GEP-NEC, and it was more obvious in the advanced stage [54–56]. Similarly, Harada et al. [57] reported that the NEC component defined the prognosis, which was involved in most of the lymph node metastasis and vascular as well as perineural invasion in patients with MiNENs. In our study, the median OS of the reported cases of pancreatic mixed ductal – neuroendocrine neoplasms was 10 months, similar to that of panNEC [58].
In summary, we suggest that EUS-FNA is the most effective tool for diagnosis and subsequent guidance of therapy. It is difficult to diagnose mixed ductal-neuroendocrine neoplasms by EUS-FNA biopsy, however, it can identify mixed morphology and has high sensitivity. If the tumor grades of the neuroendocrine component were classified as NET G2 by EUS-FNA biopsy, we propose appropriate adjuvant therapy rather than surgery.