Epithelial neuroendocrine neoplasms (NENs) encompass both well-differentiated NENs, known as neuroendocrine tumors (NETs), and poorly differentiated NENs, referred to as neuroendocrine carcinomas (NECs) (1,2). The primary sites of epithelial NENs are the pancreas, gastrointestinal tract, and lungs (1). NENs originating in the head and neck (HN) region are exceedingly rare, accounting for less than 1% of HN malignancies (2,3), 2% of all NECs, and 0.6% of all NETs (2). According to the revised 2022 WHO classification of NENs, NETs are further divided into three categories: G1 NET (no necrosis and <2 mitoses/2 mm², Ki67 <20%), G2 NET (necrosis and/or 2-10 mitoses/2 mm², Ki67 <20%), and G3 NET (>10 mitoses/2 mm², Ki67 >20% without NEC cytomorphology). NEC includes small cell carcinoma and large cell carcinoma (4). In pancreatic and gastrointestinal NENs, clinical behavior and prognosis vary significantly across pathological categories, often holding greater importance than stage classification (5). However, due to the limited number of cases, epithelial NENs in the HN region are underreported, with most existing reports focusing on small cell carcinoma (3,6,7). No studies are exclusively dedicated to HN-NET, and only a few cases are mentioned as part of broader NEN studies (8,9). The absence of data comparing NET G1, G2, G3, and NEC in the HN region highlights challenges due to a lack of specialized research in diagnosis, prognosis, and treatment, which may not lead to optimal clinical management. Consequently, we still lack a clear understanding of the prognostic differences in these tumors. Furthermore, treatment approaches for NENs remain underexplored in some parts. For example, pancreatic and gastrointestinal neuroendocrine tumors are known to express high levels of somatostatin receptors (SSTR). The expression of high levels of SSTR in pancreatic and gastrointestinal neuroendocrine tumors has led to the approval of somatostatin analogs (SSA) (10) and peptide receptor radionuclide therapy (PRRT) (11,12) for treating unresectable or metastatic NETs. In cases where SSTR is negative or SSA is ineffective despite SSTR positivity, the use of Everolimus is expected to be beneficial (13). The ESMO guidelines include recommendations for treating rare NENs, with SSA, PRRT, and Everolimus highlight as potential treatments for HN-NET (2). However, HN cases were not included in the studies that led to the approval of SSA and PRRT (10,12), and the efficacy of SSA and PRRT in HN-NET is uncertain because these cases were excluded from key clinical trials.

In this study, we examined the prognostic differences across the various categories of NEN. Additionally, we analyzed cases in which SSA and PRRT, whose efficacy in the HN region remains uncertain, were administered. Based on these findings, we propose a treatment algorithm for HN-NEN that can be applied at present. This report is the first to evaluate the prognoses of NET G1, G2, and G3 individually and to document the clinical use of SSA and PRRT for HN-NEN in actual practice.

Patients. Twenty-four patients diagnosed with head and neck NENs (HN-NEN) based on histopathologic and immunohistochemical criteria were included in this retrospective study. They were diagnosed with NEN at Kyushu University Hospital (Fukuoka, Japan) between January 2007 and December 2023. The morphological evaluation included tumor cells with round nuclei and pale acidophilic, granule-shaped cytoplasm, with microvascular interstitium intercalated between the foci. Immunohistochemical staining for chromogranin A, synaptophysin, and SSTR was performed to confirm neuroendocrine differentiation. SSTR staining was introduced in 2022; earlier cases were evaluated without this marker, potentially influencing the diagnostic criteria.

For NETs, classification into G1, G2, and G3 was performed according to the 2022 WHO classification, based on measurements of mitotic count and the Ki-67 labeling index (MIB-1 Index). The observation period was defined as the time until death or a cut-off date (Dec 2024), with a median observation period of 37 months (range 1-156 months). The primary outcome was overall survival (OS) by tumor grade (G1, G2, G3), defined as the time from diagnosis to death. The secondary outcome was a response to SSA or PRRT.

The study was approved by the Institutional Review Board of Kyushu University (IRB No 22027). All patients provided written consent to participate in the study. The procedures adhered to the principles of the Helsinki Declaration.

Statistical analysis. Overall survival (OS) was estimated using the Kaplan-Meier method. Differences between groups were calculated using the log-rank test.

Overall survival outcomes by histological subtype of HN-NENs. Among the 24 cases of NEN, NET accounted for 29% (7/24) and NEC for 71% (17/24). The most common primary site was the sinonasal cavity (42%, 10/24), followed by the larynx (29%, 7/24) (Table I). The 7-year OS rates were 100% for NET G1, 100% for NET G2, 50% for NET G3, and 43% for NEC (Figure 1). There were no significant differences in OS between these groups (p=0.340).

NET response to novel SSTR-related treatments. Table II lists the seven cases of NET. Six cases underwent surgery, while one case opted for no treatment and was followed up with observation at the patient’s request. Except for two G1 cases, the remaining five cases experienced recurrence or were cancer-bearing, and among these five cases a disease-related death at 30 months was recorded. Notable treatments for recurrence were observed in Case 3 and Case 6. Case 3 involved a 56-year-old male with a primary site in the sinonasal cavity classified as G1 (Figure 2A). Initial treatment consisted of surgical resection. Three years post-resection, the disease recurred, prompting another surgical resection followed by chemoradiotherapy with cisplatin.

However, local recurrence occurred 6 months later (Figure 2B). The recurrent lesion in Case 3 was positive for SSTR on pathological immunostaining and somatostatin scintigraphy (Figure 2C). Treatment with SSA (Lanreotide Acetate; Somatuline^® 120 mg/body/month) was initiated. After three months of treatment, slight tumor shrinkage was observed (Figure 2D), and SSA therapy has been continued since. Case 6 involved a 52-year-old male with a primary site in the sinonasal cavity classified as G3 (Figure 3A). Initial treatment included surgical resection followed by chemoradiotherapy with etoposide and cisplatin. Two years later, recurrence was observed in the contralateral cervical lymph nodes, which were treated with surgical resection and postoperative radiation therapy. Eight years after the initial surgery (and 6 years after the previous recurrence), the patient developed multiple bone metastases. Somatostatin scintigraphy showed positivity (Figure 3B), and treatment with SSTR-targeted radionuclide therapy (¹⁷⁷Lu-DOTATATE; Lutathera^®) was performed. During treatment, more pronounced uptake was observed compared to the time of scintigraphy (Figure 3C). Although accurate evaluation through imaging was challenging due to bone metastases, symptoms such as pain improved, and the patient remained alive with the disease at the time of analysis.

Treatment and prognosis of NEC. Table III summarizes the characteristics, treatments, and outcomes of 17 NEC cases. The most common primary site was the sinonasal cavity, at 41% (7/17), followed by the larynx, at 30% (5/17). Regarding initial treatment, the primary modalities were surgery in 36% (6/17), radiation in 58% (10/17), and chemotherapy in 6% (1/17). Treatment with surgery alone was performed in 12% (2/17) of cases, while radiation therapy was administered at some point in 76% (13/17) of cases, and chemotherapy in 70% (12/17). Among chemotherapy regimens, a combination of cisplatin and etoposide was used in 83% (10/12) of cases, while cisplatin monotherapy was used in 17% (2/12). Among patients with NEC treated with primary surgery, 50% (3/6) achieved disease-free survival. For those treated with primary radiation, 30% (3/10) achieved disease-free survival. Distant metastases occurred either at the initial presentation or during the disease in 52% (9/17) of cases. At the time of analysis, 29% (5/17) of patients were in disease-free survival, 47% (8/17) had died from the disease, and 18% (3/17) were alive with the disease.

The most common site of NEN is the gastroenteropancreatic system, followed by the lung and the thymus, while HN-NEN is a very rare site of NEN (1). A characteristic feature of NEN is that prognosis depends on the pathological classification of NET G1/G2/G3 and NEC rather than on stage classification. The prognosis of NEN is more dependent on the pathological classification of NET G1/G2/G3 and NEC than on stage classification, and effective treatment is differentiated among these groups. In fact, a study of 2,813 gastroenteropancreatic neuroendocrine neoplasms reported a significant difference in prognosis between the groups, with a five-year OS of 85.8% in NET G1, 73.1% in G2, 43.0% in G3 and 25% in NEC (5). However, few reports have comprehensively examined HN-NEN (8,9), and whether the same trend exists for HN-NEN is unknown. HN-NEC has been reported previously (3,6-9) as small cell carcinoma with more cases than NET, and its clinical behavior is somewhat known. Accordingly, while some cases of NEC can be cured with aggressive treatment, the majority relapse, and the 3-year survival rate is approximately 30-50%, reflecting its poor prognosis (7,14). However, reports on HN-NET are limited, and the clinical behavior and prognosis of these tumors remain unclear. Shi et al. reviewed 80 cases of NEC and 13 cases of NET, reporting a 5-year OS of 58.2% for NEC and 100% for NET, with NET showing a significantly better prognosis (p=0.004) (8). Bal et al. found that among 16 cases of NEC and nine cases of G2 NET, the disease-related death rate was 25% for NET G2 and 54.5% for NEC (9). The number of cases in these reports is limited, and they primarily focus on the overall prognosis of NET or a subset of NET cases. In our study, despite the small number of cases, we examined the prognosis of G1, G2, and G3 tumors individually. The results showed that long-term survival was possible for G1 and G2 tumors; however, G1 cases survived without disease and G2 cases survived with the disease. It was inferred that although G2 cases take a long time before affecting life expectancy, complete control of the disease is difficult. In contrast, for G3 cases, the prognosis was nearly identical to that of NEC. These findings suggest similar trends to those observed in NENs from other sites (5), and the study revealed that HN-NEN shares similar clinical behavior and prognosis with NENs from other regions. However, due to the small sample size, this is not definitive.

The treatment of gastroenteropancreatic, lung, and thymic NENs is well established. Resection is the primary treatment for NETs, while etoposide, cisplatin (or carboplatin), and radiotherapy are the main treatment recommendations for NECs. For unresectable NETs, treatment options have expanded in recent years and include somatostatin analogs (octreotide and lanreotide), peptide receptor radionuclide therapy (PRRT), and the molecularly targeted agents Everolimus and Sunitinib may be considered (15). In the gastrointestinal and pancreatic regions, treatment with lanreotide acetate (Somatuline^®) has been shown to significantly extend the 2-year progression-free survival rate to 65.1%, compared to 33.0% in the placebo group (10). For metastatic G1/G2 NET cases that are SSTR-positive, SSA therapy is the first-line treatment (15).

Furthermore, for intestinal NETs unresponsive to first-line SSA treatment, PRRT, specifically ¹⁷⁷Lu-DOTATATE (Lutathera^®), a lutetium-177 (Lu-177)-labeled somatostatin analog, is considered the second-line treatment. The PRRT group has shown a significantly longer progression-free survival than the octreotide treatment group, with a 20-month progression-free survival rate of 65.2% compared to 10.8% (11). In addition, for progressive G2/G3 NETs in the gastrointestinal and pancreatic regions, the administration of ¹⁷⁷Lu-DOTATATE (Lutathera^®) significantly extended the median progression-free survival to 22.8 months, compared to 8.5 months in the SSA treatment group. Therefore, it is recommended that ¹⁷⁷Lu-DOTATATE should be considered the new standard of care as a first-line treatment for this patient population, as presented in 2024 (12). Furthermore, for NETs in the lungs and gastrointestinal tract, Everolimus treatment resulted in a median progression-free survival of 11.0 months in the treatment group, compared to 3.9 months in the placebo group. This significant improvement in progression-free survival with Everolimus (13) suggests that for NETs that do not respond to SSA/PRRT therapies, targeted therapies such as Everolimus are selected (13,15). However, even for pancreatic and gastrointestinal NETs, the history of PRRT is still relatively short, and the precise differentiation between SSA and PRRT use remains unclear.

Currently, it remains unclear whether the same treatment regimen is appropriate for HN-NEN as for other NEN. The reasons include the anatomical difficulty of complete resection, unlike in other organs, and the small number of patients who have not been included in many clinical trials. According to the ESMO guidelines and other reports, for HN-NEC, surgery may be considered for very localized cases. However, in general, treatment focusing on etoposide, cisplatin (or carboplatin), and radiation therapy is recommended (2,6).

In contrast, for NET, surgical resection is the primary treatment. Postoperative radiation therapy may be considered for cases with macroscopic residual disease, though its efficacy remains uncertain. Cisplatin (or carboplatin)-etoposide chemotherapy is not recommended for NET (2). Thus, the treatment strategies for NEC and NET differ significantly, and due to the rarity of HN-NET, treating HN-NET in clinical practice remains challenging. Additionally, HN cases were not included in the studies mentioned above; therefore, it remains uncertain whether these treatments can be considered viable treatment options for HN-NET. Nonetheless, in this study, a certain degree of effectiveness was observed for both SSA and PRRT in HN cases. To the best of our knowledge, this is the first report showing that SSTR is expressed in HN-NET as well as in other regions, anti-SSTR treatment was effective. This observed effectiveness is encouraging news, suggesting that SSA and PRRT could become viable treatment options for HN-NET as well. At this point, based on guidelines and clinical trial results for NETs in other regions (2,10-13,15), we propose a treatment algorithm for HN-NEN, presented in Figure 4. For inoperable, progressive, or metastatic HN-NET, the first-line treatment for G1 is SSA, while for G2, both SSA and PRRT can be considered (although, from the perspective of adverse events, SSA may be preferred). For G3, PRRT is considered the first-line treatment. Regarding NEC, the treatment regimen centered on etoposide, cisplatin (or carboplatin), and radiation therapy is considered the first-line option. Note that this algorithm is only preliminary and should be validated in a larger prospective study.

Study limitations. The sample size is small due to the rarity of the disease. Additionally, the long-term follow-up of cases treated with SSA and PRRT could not be fully conducted. The retrospective nature of the study may limit the ability to control for confounding factors. Selection bias and heterogeneity in treatment regimens may occur due to the extended study period. This study highlights the prognosis differences according to the pathological categories of HN-NEN and proposes a treatment algorithm.

HN-NEN is rare, and NETs other than small cell carcinoma are especially uncommon. It is necessary to accumulate more cases to clarify these clinical entities and establish effective treatment strategies. In this study, we reported the clinical realities of various categories of HN-NEN and propose a treatment algorithm drawing on insights from NENs in other regions. This report is expected to be a valuable resource for clinicians dealing with the treatment of rare HN-NENs.

The Authors declare that they have no competing interests in relation to this study.

Conceptualization: Mioko Matsuo; Data curation: Yusuke Miyamoto, Ryunosuke Kogo, Noritaka Komune, Masanobu Sato, Shogo Masuda; Investigation: Kenji Tsuchihashi; Supervision: Takashi Nakagawa; Validation: Kenji Tsuchihashi, Kazuki Hashimoto; Writing - original draft: Mioko Matsuo.

The Authors thank Editage (www.editage.com) for English language editing.

Not applicable.

No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.