Neuroendocrine neoplasia (NEN) represents a challenging field, continuing the traditional concept of ‘karzinoide’ introduced by Oberndorfer in 1907, that was followed by the concept of neuroendocrine tumors (NETs) and then NEN according to different WHO (World Health Organization) classifications (mainly 2010, 2017, 2019) that currently takes into consideration the Ki67 and mitotic index for grading (1-3). Basically, according to the 2019 WHO classification, NETs are included in the group of NENs of gastrointestinal tract origin in addition to neuroendocrine carcinoma (NEC), also defining mixed non-neuroendocrine and neuroendocrine differentiation tumors (MiNENs) (4).

Any organ has the potential to host a tumor of neuronal embryologic origin with endocrine features, more or less associated with clear clinical and biochemical expression such as classical carcinoid syndrome. Yet, actually, NENs, most of the time portend lung and gastro-entero-pancreatic (GEP) NENs (5). The heterogeneity of the domain and new discoveries continuously require novel nomenclature, while a wide area of different practitioners are dedicated to the field of NENs (4-6). Impressive progress has been registered currently, thus an improved prognosis and a longer survival rate are reported in regards to many subtypes (7,8). The tumors underline a large spectrum of malignant potential, although the traditional benign-malign cutoffs are not really applicable in this area; however Ki67 labeling based on immunohistochemistry reports, overall NEN grading, remains the key indicator of aggressive behavior in association with the type of originating tissue, to local and distance spreading, and to the therapeutic approach that is feasible for a patient in point (7,8).

This is a narrative review of literature focusing on NEN and bone status, in terms of both skeletal metastases but also fractures and the associated diagnosis of osteoporosis. The main (but not exclusive) research tool was PubMed. The key words were ‘neuroendocrine neoplasia’, but also, due to changes of terminology over the years, ‘neuroendocrine tumors’ and ‘carcinoid tumors’ in addition to various combinations of ‘bone’, ‘skeletal’, ‘metastases’, ‘fractures’, ‘osteoporosis’. In addition, a collateral research included ‘ectopic Cushing syndrome’, ‘pheocromocytoma’, and ‘paraganglioma’ and bone field. The objective was to provide practical points for different practitioners in this complex, trans-disciplinary topic. The cited articles were mostly published between 2010 and 2020. A number of 119 references were introduced. The data were organized in micro-chapters reflecting the current status of the NEN field, as well as osteoporosis; then the relationship between NEN-associated bone metastases, respective carcinoid-related osteoporosis; and finally two topics of overlap between NEN and skeleton aspects: pheocromocytoma/paraganglioma (PHEO/PG) and ectopic Cushing syndrome.

Reports of NEN incidence vary, but generally propose an increasing trend due to endocrine disruptors, environmental triggers and advances in detection and screening protocols (9-11). The low incidence in the global population has increased to 6 times in the US during the last 30 years (9). A study on 6,179 NETs showed that the incidence between 1995 and 2012 went from 3.1 to 7.1/100,000 cases (or 3.96 to 6.61, depending on database); the most frequent age interval was between 50 and 64 years; most often NET origins are 30.6% for lung, respective 16.82% for the small intestine (9).

Another database between 1995 and 2014 included 85,133 NETs and it showed that the G1 NET median overall survival was 233 months with a 5-year NEN-specific survival rate of 97.6%; median survival decreased with age (poor prognosis was registered especially in individuals over 70 years); 34% of persons over 70 years had metastases at diagnosis and 40.8% of these patients presented with G3 NET (10). Data from the Swedish Cancer Registry revealed the following related to the metastatic potential of 7,334 NENs. Liver was the most frequent site (representing 82% of all the patients with NEN-related metastases); 66% of patients with lung metastases had synchronous hepatic spreading; small intestine followed by pancreatic-hepatic-biliary tract were the most common histological types that were associated with metastases while the lowest metastatic potential was for appendix and rectal NENs; males had more frequent metastases than females; the most severe prognosis was in subjects with metastases from unknown primary NEN; and the metastatic behavior mostly depended on the primary tumor site (11).

As this study, and also another study on 302 NENs, brain metastases were most frequently associated with lung NEN, and most probably at that point, liver and other sites such as the lymph nodes and bone metastases were synchronously positive (11,12). Bone metastases were described in one fifth of the patients diagnosed with NEN (13). The skeleton affinity was regulated by intrinsic factors of epithelial-mesenchymal transition (EMT) such as C-X-C chemokine receptor 4 (CXCR4) expression, in collaboration with other cofactors such as CXCL12(13). To date, the complex biomolecular mechanisms of skeleton spreading are less understood, yet the point is that bone colonization impairs the quality of life of the patient and it serves as a surrogate for a reduced survival rate (14).

Whether there is a relationship between NENs without bone metastases and associated low bone mass (a potential new cause of secondary osteoporosis) and primary osteoporosis (menopausal or age-related) is still difficult to establish based on the epidemiological data we have to date. Contrary to the low NEN incidence in the global population (despite the recent massive increasing trend), primary osteoporosis and associated low-trauma (osteoporotic) fractures have a high prevalence in the general population, especially in the aging population, in postmenopausal women and seniors (which also represent the more exposed population for the diagnosis of NENs) (15-17).

According to data from the International Osteoporosis Foundation (IOF) site, the number of people at high fracture risk from 2010 will double in 2040; 1 in 3 females and 1 in 5 males over 50 years will suffer an osteoporotic fracture (18). The economic burden related to low-trauma fractures is reflected in associated costs of $17.9 up to £4 billion/year in the UK and the US due to the fact that currently osteoporosis is undertreated (19). These days, less than one fifth of osteoporotic patients receive medication for this diagnosis, and the current rate is not increasing due to multiple causes such as the presence of other co-morbidities (especially age-dependent) that also require medication where osteoporosis is not considered a priority; the fear of side effects of anti-osteoporotic drugs; and difficulties of anti-osteoporosis medication reimbursement protocols depending on countries (20).

Secondary prevention (treatment of osteoporosis after a first osteoporotic fracture) also is suboptimal (21). For instance, a picture of how the real world medical approach has emerged, is reflected by a study in 2019 on 2,933 patients over 50 years of age with a recent incident fragility vertebral fracture (between 2008 and 2014). This study showed that 93% of these patients were naïve of anti-osteoporosis medication during the prior 2 years and the following year after the diagnosis of the fracture (21). Only 7% of patients immediately started medication after fracture and another 5% of subjects were offered drugs during the first year, while 73% of them were never treated (21). A total of 38% of these patients had a second fracture within the following 24 months (21). Screening programs may be helpful. For example, a controlled study from 2020 included UK people aged between 70 and 85 years showing that the screening of 1,000 subjects saved 29 fractures (hip/non-hip ratio of 9/20) during the following 14 years (remaining life time) with a significant improvement in the associated costs (22). A multitude of factors are contributors to damaged bone mass, especially after the age of 50 years. Based on the epidemiological data that we have to date, NENs are not yet an input in this playlist but convincing evidence is increasing.

NENs may affect the skeleton status through different mechanisms (23,24). On the one hand, there are well-established malignancy-related aspects which include: bone metastases potentially associated with pathological fractures (such as vertebral fractures or even atypical femoral fractures), vertebral/cord compression due to vertebral fractures, and paraneoplasic hypercalcemia (23,24). The natural history of such incidental events largely varies due to the multiple subtypes of NENs; thus, a different behavior is expected, but the most important aspect in secondary dissemination is actually the tumor aggressive pattern rather than a particular histological type (25,26).

Generally, the presence of bone metastases in patients with NENs is regarded as a poor prognostic indicator and the local behavior typically does not differ in regards to the bone metastases of other non-NEN malignancies (23). If liver and lung metastases are already identified, then skeleton and brain metastases should be checked (27). A potential explanation of the fact that some NENs have a high bone affinity is the overexpression of CXCR4 at the level of the tumor and its serial identification through immunohistochemistry after surgical removal or after biopsy may become a predictive factor of skeleton dissemination (26).

A multi-institution, 5-year study analyzed the time-to-event parameters in 691 patients with NENs and 82 of them (representing 12%) had bone metastases (28). A total of 55% of the patients with bone metastases were males; median age at diagnosis was 49 years; among NENs: 25% were malignant pheocromocytoma and paragangliomas, 20% were G3 NENs, while 8% had pancreatic origin (28). The complications of metastases were the following: local symptoms, mostly pain (60%), cord compression (10%), secondary fractures (9%), and hypercalcemia (4%) (28). Their overall management included: local radiation therapy (50%), bisphosphonates (45%), bone/spine surgery (18%), denosumab (5%); as for malignant PHEO/PG additional I¹³¹-MIBG (radioiodine 131-metaiodobenzylguanidine) therapy was introduced (accounting for 11% of all NEN patients with bone metastases) (28).

A prior review published by Hori et al showed that the average age of patients with carcinoid tumor-related skeleton metastases was 54.9 years, with male predominance (66%); most frequent bone site was the vertebra; the survival rate was lower in osteolytic type compared to osteosclerotic metastases (29). Another study published in 2020 on 2,005 NEN patients showed a prevalence of 7% for bone metastases, representing an independent factor of poor survival based on a hazard ratio (HR) of 1.412 (95% CI: 0.965-2.065) (30). Another study from 2019 included 3,909 pancreatic NENs and it showed that, while the most metastatic organ was the liver, the median survival time when bone metastases were present was 56 months and in this situation the 5-year overall survival was 42.7% (31).

An imaging based study from 2020 analyzed bone metastases detected with ¹⁸F-DOPA PET in 155 metastatic midgut NEN patients (admitted between 2011 and 2018) and it showed that: 29.7% of NENs had skeleton colonization; bone dissemination was independently associated with carcinoid syndrome; the majority (78%) were of osteoblastic subtype; less than 9% were symptomatic; they represented an independent prognostic factor (32). The concept of ‘oligo-bone metastases’ was introduced for lesions associated with less than 5 criteria at ¹⁸F-DOPA PET (vs. ‘poly-bone metastases’) and they were not associated with a poor prognosis (opposite to ‘poly’ subtype) (32). Overall, NENs may embrace a metastatic profile, nevertheless including the bone, thus a level of awareness is essential.

On the other hand, non-metastatic skeleton damage may include the traditional diagnosis of osteoporosis/osteopenia according to central dual-energy X-ray absorptiometry (DXA) and potentially complicated with (low-trauma or even spontaneous) osteoporotic fractures, in tumors such as lung and GEP-NEN, due to multiple hormonal, mechanical, nutritional, and metabolic factors (33). We propose the term carcinoid-related osteoporosis, even though the actual traditional terminology of ‘carcinoid’ tumor has been replaced by NEN/NET; yet, we consider that this expression clarifies a new clinical entity, different from neuroendocrine mechanisms underlying bone status under normal and abnormal conditions (34,35). The pathogeny and epidemiologic data are still a matter to discuss, while a correlation with NEN status itself or with neuroendocrine marker levels is ongoing (34,35).

A few studies have shown that in NENs, serum serotonin, not its urinary metabolite hydroxyindolacetic acid (5-HIIA), is inversely correlated with hip BMD (34). But the skeleton influence consists in a cocktail of factors produced by the tumor, including non-serotonin growth factors or hormones with bone tropism (35). Up to this moment, carcinoid-related osteoporosis is not a standardized cause of secondary osteoporosis, but there is an increasing amount of medical evidence that points out this well-deserved specification. Case-finding strategies are still lacking in order to address the bone health in these cases, especially in those with a good prognosis, and in the absence of bone metastases. The importance of approaching the bone issues in addition to medical and surgical management in NEN subjects is related to increased quality of life knowing that adequate protocol of multimodal investigations as well as modern multiple-level NEN therapies have massively improved the overall survival during the last decade (36,37).

The tools of bone and fracture risk assessment, as seen in non-NEN cases, are mostly similar: the evaluation of phosphorus-calcium metabolism: serum calcium and phosphate levels, mineral hormones (parathormone (PTH) and 25-hydroxyvitamin D [25(OH)D]], bone turnover markers [of formation (alkaline phosphatase, osteocalcin, P1NP] and of resorption (CrossLaps) (38,39). Also, traditional fracture risk assessment includes central DXA (especially in menopausal women and men over 50 years old), while the Fracture risk assessment tool (FRAX) model does not include as distinct input the diagnosis of NEN (40,41). As regard to the Trabecular Bone Score (TBA), data are also missing, except for an extrapolation for a diabetic subpopulation (especially in post-menopausal females) of NEN subjects due to impaired glucose profile via the secretion of hyperglycemic hormones such as glucagon, also for syndromic circumstances that are associated with primary hyperparathyroidism or for Cushing syndrome as seen in paraneoplastic type accompanying some NENs (42,43). In some NEN multi-leveled regimes, glucocorticoids are used and thus TBS-related changes are expected (44).

The identification of prevalent fractures starts from a simple screening plane X-ray of the spine (for instance, profile thorax-lumbar X-ray) up to computed tomography (CT) or magnetic resonance imaging (MRI) (that are usually applied in the non-NEN general population in selected cases) or in addition to detailed investigations for NENs such as whole body bone scintigraphy, somatostatin receptor scintigraphy (SRS) or positron emission tomography/computed tomography (PET-CT) which may also serve for differentiation of a metastasis-related fracture from an osteoporotic fracture (45,46).

In terms of skeleton health, patients with NENs may suffer from a multitude of factors, not only from direct malignant dissemination to the bone (47,48). For instance, secondary diabetes mellitus due to pancreatic and duodenal tumors such as somatostatinomas (with or without underlying neurofibromatosis type 1), or pheocromocytomas or due to paraneoplastic hypercortisolemia may cause a deterioration of bone quality like that reflected by TBS in addition to abnormal turnover that increase fracture risk (49-51). In addition, some worsening of the glucose profile is expected after high doses of somatostatin analogues, a powerful medication for NEN control of secretion and growth, as seen (to a lesser extent though) in acromegaly (52,53).

Weight loss is another mechanism of bone mass decrease, being caused by long standing disease-related anorexia, different oncological drugs such as chemotherapy, or intestinal/gastric resection because of tumor location (54). Weight loss may be associated with other nutritional deficiency and anemia of different mechanisms (55).

Low body mass index is associated with low BMD and high fracture risk, as also reflected by the algorithm FRAX, as generally applicable in individuals over 40 years of age (56,57). Weight loss, advanced age, some drugs (such as checkpoint inhibitors), uncontrolled carcinoid syndrome (clinically manifested with diarrhea) and severe illness may aggravate a process of sarcopenia (deficiency of skeletal muscle mass and strength) that is a supplementary contributor to the risk of fall and an independent indicator of poor prognosis (58). Some chemotherapy regimens and radiotherapy are associated with dexamethasone therapy or similar glucocorticoid products, thus a component of glucocorticoid osteoporosis may be overlapped under these particular circumstances (59). In NENs of prostatic or ovarian origin, the co-presence of hormonal-dependent adenocarcinoma may require induction of hypogonadism, either chemically or surgically, causing low BMD (60). Hypovitaminosis D, a particular risk factor of poor bone status, is related to malnutrition, surgery of the intestinal area, malabsorption, liver metastasis, low sun exposure, and different concurrent medications (61). Of course, NEN itself may produce hormones that have a negative impact on the skeleton such as serotonin or parathormone-related protein (PTHrP) (62,63).

PHEO, respective PG, represent tumors with adrenal, respective extra-adrenal enterochromaphin cells which produce catecholamine (norepinephrine and epinephrine) and others amines like dopamine. These tumors are typically candidates to surgery in order to be cured, associating a high cardiovascular risk of complications before, during and after surgery (64,65). Neoplasia has a low incidence in the global population (2-8/1 million) but they represent up to 6 cases out of 1,000 hypertensive persons (66). However, in otherwise asymptomatic patients, they may actually be accidentally detected as an incidentaloma in 10-50% of cases, depending on the definition of incidentaloma (67). Metastases define the malignant potential, and not necessarily the post-operative histological report (66,68). The percent of distance spreading varies between 2 and 60%, most frequently 3-10% depending on the study, usually PG having a higher potential than PHEO (66-69).

Bone is the most frequent site of metastases, involving different areas such as the vertebras to unusual locations such as the mandibular or orbital (70,71). If succinate dehydrogenase (SDHB) is positive, the metastatic rate is higher than in sporadic cases; other sites of malignant colonization are the lung and liver with a more severe prognosis (69). Skeletal metastases may be positive at first admission or at a more distant time (even after 2 decades since diagnosis) (72). Some cases of metastatic spine paraganglioma inducing cord compression have been reported (73). Skeletal-related events may also include hypercalcemia due to bone metastases or PTHrP tumor production, distinct to the synchronous presence of primary hyperparathyroidism (PHP) in cases with type 2A multiple endocrine neoplasia (MEN) syndrome (74,75). Removal of the primary tumor improves the prognosis, while prevention of metastases is less feasible, but prompt extended interventional radiology methods may help (76).

In non-malignant PHEO/PG, several studies have shown a higher prevalence of osteoporosis/osteoporotic fracture compared with the general population, related to the tumor hormonal over-secretion with the bone targeted and to increased sympathetic activity which is a key player of bone microarchitecture/quality (TBS assay may reflect this particular aspect) (77-79). Thus an improvement in bone status after tumor removal has been observed (77-79).

A retrospective, single-center study in 2020 showed in 49 patients with PHEO vs. 61 subjects with non-secretor adrenal tumors that prevalence of vertebral fractures was higher (43 vs. 16%) after adjustment for age and sex, with a 3.8% increase in lumbar BMD after tumor removal (79). A controlled study in adrenal tumors (incidentaloma/PHEO ratio of 266/29) showed a 2.9% lumbar TBS reduction in subjects with catecholamine-producing tumors who also had a statistically significant reduced lumbar BMD and an increased collagen-derivate bone resorption marker (78).

Another study from 2021 published by Q-AND-A study group also showed a lower TBS in the PHEO/PG group vs. non-functional adrenal tumors; TBS being statistically significant lower in subjects with vertebral fractures vs. those without vertebral osteoporotic fractures; TBS was independent to BMD decrease; TBS was inversely correlated with hormone excess and tumor size (80). The same study group showed that PHEO patients had a higher rate of abdominal aortic calcifications when associated with vertebral fractures, thus proving the co-presence of osteoporosis and atherosclerosis in PHEO (81). In primary menopausal osteoporosis, a link between metabolism disturbances such as hyperglycemia or hyperlipidemia and low BMD and/or low TBS has already been established (82,83). Specific protocols of BMD/TBS screening in PHEP/PG patients, especially over the age of 50 years or menopausal woman are still deficient.

Endogenous Cushing syndrome, of either adrenal, pituitary or cancer-related cause (including NEN), is a combination of obesity, diabetes mellitus, high cardiovascular risk, hypokalemia, coagulation disturbances, hirsutism, male hypogonadism, high frequency of infections, secondary glucocorticoid osteoporosis and associated low-trauma fractures such as vertebral fractures, and also aseptic necrosis of different regions as the hip or shoulders (84,85). The rapid evolution of the symptoms, and a phenotype likely associated with imbalances in serum electrolytes, hyperpigmentation, weight loss, and proximal myopathy are more suggestive of paraneoplastic Cushing syndrome (86). The neoplasia mass itself and potential local effects of the originating tumor may be identified or the underlying cause remains occult, thus multiple investigative protocols are necessary to locate it, including functional imaging (87). In addition to identification of the primary source of non-pituitary adrenocorticotropic hormone (ACTH), tests are first necessary to confirm that it is an ectopic Cushing syndrome (such as steroid metabolomic assays) (88). As management, ideally the removal of NEN is recommended to control hypercortisolemia, but in many cases it is not feasible thus steroidogenesis inhibitors, mitotane or bilateral adrenalectomy are indicated to control the large area of complications due to persistent exposure to hypercorticism (89).

Fragility fractures occur in Cushing syndrome due to anomalies of bone turnover as reflected by bone turnover markers (increased bone resorption, reduced bone formation), anomalies of growth hormone [insulin-like growth factor (IGF1)], of sclerostin, and potentially due to the effects of diabetes mellitus on bone, and even weight changes in association or not with hypovitaminosis D (90,91).

A cohort study between 2000 and 2019 included 130 subjects with ectopic Cushing syndrome (female/male ratio of 93/37) and it showed that the prevalence of osteoporosis was 25 vs. 48% (females vs. males), while the prevalence of vertebral fractures was 5 vs. 16%, suggesting the fact that actually males are more affected than women when it comes to bone status (92). Moreover, the fracture risk was dependent to some extend on low BMD according to central DXA and/or decreased values of TBS; but the main contributor to fracture risk was found to be the level of hypercortisolemia and the time of exposure, not necessary the pathogenic type of Cushing syndrome or the source of ectopic ACTH over-secretion (meaning that it seemed less important that there was an underling NEN in association with ectopic ACTH production in terms of glucocorticoid osteoporosis) (93).

The European Registry on Cushing's syndrome (ERCUSYN) collected the data of 481 patients with Cushing syndrome, from 36 centers, respective 23 countries, between 2000 and 2008, and 5% of them had ectopic ACTH excess (94). Overall, the prior mentioned observation was consistent; male vs. women had a statistically significant higher frequency of osteoporosis based on lumbar BMD, and more often were diagnosed with vertebral fractures (52 vs. 18%), and also rib fractures (34 vs. 23%) (94).

Among NENs, patients with bronchial carcinoid were more frequently reported to develop ectopic Cushing syndrome, potentially complicated with osteoporotic fractures especially at the vertebral site, even complicated with vertebral compression that required surgical procedures such as vertebroplasty (95). Another scenario included the synchronous presence of NEN-related paraneoplastic Cushing syndrome with potential skeleton impairment and an NEN with a gastro-intestinal origin, or even of thymus origin or medullary thyroid cancer, with bone metastases (96,97). Generally, when bone metastases are positive, NEN liver spreading is already present, thus there is also a higher risk of an associated clinically manifested carcinoid syndrome due to interferences in NEN-released serotonin hepatic metabolism (98,99). The differentiation of the etiology of the fractures (osteoporotic or due to malignancy spreading) may be assessed by using advanced technology tools such as gallium (Ga)-DOTATATE PET/CT or ¹⁸F-FDG (18-F-fluoro-deoxy-glucose) PET (100).

Several points need to be mentioned concerning this current complex topic since they are still a matter of discussions.

Lately, due to a more complex and efficient therapeutic approach toward NENs, patient survival has improved; thus, there is a need for increasing the quality of patient life, including the bone status, but not only related to skeleton metastases, but also in those cases with a good prognosis related to carcinoid-related osteoporosis, knowing the burden of low-trauma fractures. Protocols of serial checking of fracture risk are imperious, especially if clinical risk factors that are more frequent in NEN patients are already identified (advanced age, diabetes mellitus, extended period of time in menopause, induced male/female hypogonadism, prolonged exposure to glucocorticoids). As simple tools of bone status assessment, biochemistry parameters such as serum calcium and bone turnover markers are extremely useful, in addition to other measurements such as DXA-BMD or even TBS (especially in menopausal diabetic women, PHEO/PG or ectopic Cushing syndrome). Up to this moment, carcinoid-related osteoporosis is not a standardized cause of secondary osteoporosis, but this remains an open topic and the future will provide us with specific data that require a multidisciplinary medical team once again, as usually NENs do.