Colorectal neuroendocrine neoplasms (CRNENs) are diagnosed with increasing frequency, mainly due to the widespread use of colonoscopy, including screening examinations [1–9]. Most of these lesions are located in the rectum. An increasing body of evidence suggests that rectal neuroendocrine neoplasms (RNENs) and colon neuroendocrine neoplasms (CNENs) are separate diseases. RNENs are characterised with a low to moderate degree of malignancy, good prognosis, and most of the lesions can be treated endoscopically. On the other hand, CNENs are often aggressive, poorly differentiated neoplasms with a higher degree of malignancy, and unfavourable prognosis, and surgery is the treatment of choice [5, 10].
1. Epidemiology
CNENs and RNENs constitute 17.6% and 26.3%, respectively, of all neuroendocrine neoplasms of the gastrointestinal tract [11, 12].
According to the 2015 Surveillance, Epidemiology and End Result (SEER) register, the incidence of CNEN is 0.35/100,000/year [13]. The mean age of CNEN onset is the seventh decade of life. Women suffer twice as often [14]. The most common location of colon tumours is the caecum [15].
There has been a ten-fold increase in the incidence of RNEN over the past 30 years, and it now amounts to 1.5 cases per 100,000 per year [13]. They are detected once in every 1000–2000 endoscopic examinations [1, 5, 6, 8, 16]. The average age of a patient diagnosed with RNEN is 56 years, and they are predominantly men [17].
2. Clinical characteristics
The clinical picture of a CRNEN case is related to its location and stage. These tumours are characterised by the lack of specific hormone secretion [8].
If the neuroendocrine neoplasm (NEN) is located in the colon, the main symptoms include a change in the nature of bowel movements (most often diarrhoea), and in advanced disease, abdominal pain and weight loss [7]. Abdominal pain is caused by the effect of the tumour mass or overgrowth of the connective tissue (desmoplastic effect). Weakness and decreased exercise tolerance related to blood loss through the gastrointestinal tract may also occur. Gastrointestinal obstruction is possible. At diagnosis, the mean CNEN size is approximately 5 cm. Distant metastases occur in 16–40% of patients [18]. In the case of CNEN, the five-year survival is the lowest among all gastrointestinal neuroendocrine neoplasms and refers to 33–42% of patients after surgical treatment [18].
Most RNENs are asymptomatic and are diagnosed incidentally during colonoscopy [5, 19–21].
Symptoms may include rectal bleeding, itching and discomfort in the area, change in bowel movements, and painful faecal urgency. The vast majority (75–85%) of the lesions are local at diagnosis. Five-year survival is 75–100% depending on the grade of histological maturity, the proliferation index, and the stage of clinical advancement [22]. Most RNENs do not exceed 10 mm at diagnosis. The tumours of this size have a low metastatic potential, but it should be kept in mind that metastases to regional lymph nodes and distant metastases occur in 3% and 1.6% of patients, respectively [23]. The risk of metastasis increases with the lesion size. In tumours of 11–19 mm in size, regional metastases occur in 66% and distant metastases in 50% of patients. However, in tumours of ≥ 20 mm in size, the risk of distant metastases is almost 100% [23].
Carcinoid syndrome in RNEN is almost unheard of due to the very rare occurrence of serotonin-secreting cells in this location (0.1%) [2, 24].
A second neoplasm develops in 13% of patients with CNEN [2, 4, 5, 25]. The gastrointestinal system, including the colon, is the most common location for synchronous tumours, while metachronous neoplasms mainly affect the lungs, prostate, and urinary system.
3. Diagnostics
3.1. Biochemical diagnostics
We do not have a biochemical marker specific for CRNEN. Determination of serum chromogranin A (CgA) is still the most valuable method of monitoring the disease treatment and prognosis. The CgA concentration may increase and correlate with the severity of the neoplastic disease [26].
Because tumours located in this part of the gastrointestinal tract rarely secrete serotonin, the concentration of 5-hydroxyindoleacetic acid (5-HIAA) in the 24-hour urine collection usually remains within the normal range.
The possibility of using other tests in the diagnosis and monitoring the course of the disease and treatment effects is described in the article Update of the diagnostic and therapeutic guidelines for gastro-entero-pancreatic neuroendocrine neoplasms (recommended by the Polish Network of Neuroendocrine Tumours) [27].
The minimal consensus statement on biochemical tests:
3.2. Pathomorphological diagnosis
Similarly to other gastrointestinal NENs, CRNENs include neoplasms, the common element of which is the ability to produce neuroendocrine markers. These properties are most often confirmed by immunohistochemical tests, demonstrating mainly the presence of synaptophysin and CgA in neoplastic cells. The ability to produce neuroendocrine markers is not a factor determining the biology of these tumours, but only their common property. Based on their clinical course and the response to treatment, they can be divided into two main groups. These are neuroendocrine tumours (NETs) and neuroendocrine carcinomas (NECs). This division of pancreatic neuroendocrine neoplasms (PanNENs), initially proposed in 2017, is based on the assessment of two factors: proliferative activity determined by the Ki-67 proliferation index and the mitotic index, as well as by the histoformative features of neoplastic cells [28, 29]. This division effectively reflects the biological differences stemming from various genetic disorders in these groups. For example, in the NET group there are no mutations in the TP53 and RB genes, which are observed in the NEC group. Moreover, NECs, unlike NETs, respond well to platinum-based chemotherapy [30].
The group of well-differentiated NETs, due to proliferative activity defined by the Ki-67 proliferation index and the mitotic index, is divided into three subcategories: NETs G1 with the Ki-67 proliferation index < 3% and the mitotic index < 2/2 mm2, NETs G2 with the Ki-67 proliferation index in the range 3–20% and the mitotic index in the range 2–20/2 mm2, and NETs G3 with higher parameters of proliferative activity. It should be noted that we are not grading the group of poorly differentiated NENs, or NECs, previously graded as NENs G3 [30]. This has led to confusion in the interpretation of NETs G3. It is also worth noting that currently the group of NETs G3 and NECs has the same lower limit of the Ki-67 proliferation index (above 20%) and no upper limit of the Ki-67 proliferation index. Very rarely, neoplasms with the Ki-67 index of, e.g., 80% can be classified into the NET G3 group, and neoplasms with the Ki-67 index of, e.g., 30% can be calssified into the NEC group. The differences between these two groups of neoplasms, which are important from the point of view of treatment and prognosis, are based on cytological and histoformative features.
There are two groups of NECs classified according to their cytological and histological structure: small cell carcinoma (SCC) and large cell neuroendocrine carcinoma (LCNC). The nomenclature suggests significant morphological differences, while in practice this distinction is sometimes difficult, sometimes even impossible, especially when we are dealing with a small biopsy specimen. Due to the currently used treatment regimen, which is the same for both types of carcinomas, the appropriate procedure in difficult cases and the availability of only a small specimen is the diagnosis of poorly differentiated neuroendocrine carcinoma (without taking into account the division into two groups).
The Ki-67 proliferation index in neuroendocrine tumours is often uneven; therefore, the assessment is performed in the areas of the highest intensity, the so-called hot spots. For this reason, when we are dealing with a small section of neoplastic tissue, and not the entire tumour, the degree of differentiation may change after assessing the entire tumour in the postoperative material. The method of assessment of the mitotic index and Ki-67 proliferation index is presented in the pathomorphological part of the Update of the diagnostic and therapeutic guidelines for gastro-entero-pancreatic neuroendocrine neoplasms (recommended by the Polish Network of Neuroendocrine Tumours) [27].
As in other parts of the gastrointestinal tract, there also occur tumours in the large intestine with a mixed histological structure, including neuroendocrine and non-neuroendocrine tissues. Such tumours are classified as mixed neuroendocrine — non-neuroendocrine neoplasms (MiNENs). In the large intestine, these neoplasms occur mainly as tumours composed of components of adenocarcinoma and neuroendocrine carcinoma. For this reason, they were formerly referred to as mixed adenoneuroendocrine carcinomas (MANECs). To classify the neoplasm to the MiNEN group, it is assumed that each component should constitute at least 30% of the tumour tissue. In cases that do not meet the quantitative criteria, the presence of a minority component should be considered in the diagnosis, especially when it is a neuroendocrine carcinoma component. In MiNEN, both components must be invasive. Thus, the presence of NEN tissue within the adenoma does not classify it as mixed tumour. There are rare cases when both neoplastic tissues do not constitute a single tumour with two components, but two independent tumours. Such tumours are not classified as MiNEN; they are referred to as “collision tumours”, and each component is described and graded separately. Neuroendocrine differentiation sometimes occurs in relapses of non-neuroendocrine cancers after treatment. These tumours also do not belong to the MiNEN group.
The presence of cells with neuroendocrine properties can be confirmed in immunohistochemical tests in 50% of cancers typical for the gastrointestinal system, but they do not constitute a basis for the diagnosis of NET. The histopathological structure, confirmed in immunohistochemical tests, is of decisive importance. Colon NETs are most often composed of serotonin-producing cells (EC-cells), which, histologically, form nest structures with peripheral nuclear palisades, similarly to tumours composed of histamine-producing cells (ECL-cell). These cells with eosinophilic cytoplasm demonstrate a strong positive reaction with the presence of CgA and synaptophysin. More rarely, NETs are composed of L-cells forming trabecular structures that are characterised with a positive immunohistochemical reaction to the presence of synaptophysin and, only focally, a weakly positive reaction to the presence of CgA, similarly to NEC. Tumours composed of these cells are more common in the rectum; therefore, synaptophysin is a more sensitive marker of immunohistochemical neuroendocrine differentiation for tumours located in this area.
CRNETs often demonstrate the presence of CDX2 in immunohistochemical tests, which is sometimes useful in the search for a primary tumour in the presence of metastases of a neuroendocrine tumour from an unknown primary site. Unfortunately, some of these tumours also demonstrate the presence of TTF1, an antigen associated mainly with lung and thyroid cancers.
The 2019 World Health Organization (WHO) histopathological classification of colorectal neuroendocrine neoplasms is presented in Table 1 [32].
Neuroendocrine tumour NOS (8240/3) |
Neuroendocrine tumours, grade 1 (G1) (8240/3) Neuroendocrine tumours, grade 2 (G2) (8249/3) Neuroendocrine tumours, grade 3 (G3) (8249/3) L-cell tumour (8152/3) Glucagon-like peptide-producing tumour (98152/3) PP/PYY-producing tumour (8241/3) Enterochromaffin-cell carcinoid (8241/3) Serotonin-producing tumour (8241/3) |
Neuroendocrine carcinoma (8246/3) |
Large cell neuroendocrine carcinoma (8013/3) Small cell neuroendocrine carcinoma (8041/3) |
Mixed neuroendocrine — non-neuroendocrine neoplasm (MiNEN) (8154/3) |
The final postoperative histopathological report must include an assessment of the tumour–node–metastasis (TNM) staging. In the case of neuroendocrine tumours, due to biological differences in relation to carcinomas, separate staging classification is used. Currently, the 2019 WHO TNM classification of gastrointestinal tumours is accepted, which in the case of colon tumours coincides with the 8th edition of the American Joint Committee on Cancer (AJCC) and Union for International Cancer Control (UICC) classifications [32–34] (Tab. 2). NEC and MiNEN are graded as non-neuroendocrine carcinomas.
T — primary tumour |
|
pTX |
The main tumour has not been assessed |
pT1 |
The tumour is confined to the mucosa and submucosa and does not exceed 2 cm in diameter |
pT1a |
The tumour is confined to the mucosa and submucosa and does not exceed 1 cm in diameter |
pT1b |
The tumour is confined to the mucosa and submucosa and its diameter is 1–2 cm |
pT2 |
The tumour infiltrates the muscularis propria or is larger than 2 cm in diameter |
pT3 |
The tumour infiltrates the subserous tissue or the pericolonic and perirectal tissues not covered by the peritoneum |
pT4 |
The tumour infiltrates the peritoneum and other organs |
N — regional lymph nodes |
|
pNX |
Regional lymph nodes have not been assessed |
pN0 |
No regional lymph node involvement |
pN1 |
Regional lymph node involvement |
M — distant metastases |
|
pM0 |
No distant metastases |
pM1 |
The presence of distant metastases
|
An important parameter in the CRNEN histopathological report is the assessment of the proximal, distal, and peripheral margins. The peripheral margin is assessed in the parts of the gastrointestinal tract not covered with serosa. It must be emphasised that during the macroscopic assessment of the surgical specimen, it should be marked with ink. It is recommended that the distance from the most deeply infiltrating tumour foci to the peripheral margin line is specified. The presence of residual neoplastic tissue following treatment, denoted by the letter R, is not a necessary factor for the assessment of pTNM and clinical advancement. However, this assessment is very important for further therapeutic decisions, especially in patients with rectal tumours. Due to doubts regarding the minimum width of the surgical margin relative to the tumour, in order to determine the lack of residual tissue at the tumour resection border, this assessment can be characterised according to the scheme proposed by Wittekind [35] (Tab. 3).
Assessment of residual tumour tissue (R) after treatment |
Presence of tumour tissue |
Rx |
Presence of residual neoplastic tissue cannot be assessed |
R0 > 1 mm |
No neoplastic tissue in the resection margin located at a distance larger than 1 mm from the tumour |
R0 ≤ 1 mm |
No neoplastic tissue in the resection margin located at a distance smaller than or equal to 1 mm from the tumour |
R1-dir |
Neoplastic tissue present in the resection margin — the resection line runs through the neoplastic infiltration |
R2a |
Neoplastic tissue visible in macroscopic picture at the site of tumour resection (locally) |
R2b |
Neoplastic tissue visible in macroscopic picture at a site distant from the tumour (e.g. in a metastatic lymph node cut within the margin of resection) |
R2c |
Neoplastic tissue visible in macroscopic picture at both sites |
Statement on pathomorphological examinations
The macroscopic assessment method and the principles of collecting specimens for histopathological examination are the same as in the case of other neoplasms. It is included in the standards for the postoperative material processing published by the Polish Society of Pathologists. It is recommended that the final pathology report be a synaptic report, which includes the following:
This way of result reporting is transparent, first of all, to clinicians and oncologists. At the same time, because laboratories use various numbering systems for blocks and organs from which the specimens were collected, it is necessary for the report to include information on the labelling of preparations collected from a specific area. This avoids any ambiguity in the case of pathomorphological consultations.
Minimal consensus statement on pathomorphological examinations
3.3. Location diagnostics
3.3.1. Endoscopic diagnostics
“Classic” white light colonoscopy combined with biopsy for morphological assessment is an essential tool in the diagnosis of CRNENs.
Diagnostic efficiency is increased by the use of next-generation image-enhanced endoscopy (IEE), including narrow-band imaging (NBI) and magnifying endoscopy (ME). Image enhancement endoscopy increases the sensitivity of early lesion detection. Additionally, it helps in the differential diagnosis (distinguishing NENs from epithelial lesions), which in turn determines the selection of the optimal treatment method [36]. The most accessible and most frequently used is endoscopic image enhancement with the use of a narrow light band.
In recent years, artificial intelligence has been a significant achievement. Equipping the devices with this type of software significantly increases the sensitivity and specificity of endoscopy [37, 38].
Imaging of the large intestine is also possible with the use of capsule endoscopy; however, in the diagnostics of CRNENs, this method has no practical application at present [39].
In most patients (83%) with CRNEN (compared to other subepithelial lesions), the biopsy is positive because NENs originate from the muscularis mucosae [40]. In patients qualified for endoscopic treatment, biopsy is not recommended, because fibrosis occurring after collecting the specimens may make the procedure more difficult and increase the risk of complications. The collection of material for pathomorphological examinations is necessary before surgical treatment.
The characteristic endoscopic morphological features of CRNEN include polyps growing from a broad base, with a smooth surface, covered with unchanged or injected mucosa with dilated vessels, often yellow/white (in the case of rectal lesions). CNENs are most often advanced lesions macroscopically resembling cancer infiltration, diagnosed early in the form of a polyp/subepithelial tumour. Most RNENs (80%) have the above-described characteristic morphological features [5–8]. The atypical features, occurring in about 20% of cases, include the following: a semi-pedunculated shape, a form resembling a mushroom cap with a central depression, a flat lesion, reddening of the mucosa, and the presence of erosions or ulcerations on the surface. Atypical features occur mainly in lesions larger than 1 cm. Minor NEN lesions (up to 5–7 mm) may resemble a hyperplastic polyp. The use of endoscopy with NBI and/or magnification is helpful in differentiating between both lesions due to the different pit pattern and microvessel image [36]. RNENs are usually single, most often occurring in the middle part. The majority of RNENs (70–80%) are up to 10 mm in size at diagnosis [24]. In 2020, Chen et al. described a new RNEN endoscopic assessment system [41]. The size, shape and surface of the mucosa are considered in the scale. The division of individual features is as follows: size < 1 cm, ≥ 1 cm, ≥ 2 cm; shape — sessile, semi-pedunculated, flat/mushroom-shaped; mucosa — smooth, with a depression, erosion/ulceration (Tab. 4). The individual features are assigned points, with a cut-off value of 110. The accuracy of the scale in identifying poor NEN prognosis is high and reaches 95% (change ≥ 110 points).
Endoscopic picture |
Number of points |
Lesion size |
|
< 1 cm |
0 |
≥ 1 cm |
76 |
≥ 2 cm |
100 |
Lesion shape |
|
Non-pedunculated |
0 |
Semi-pedunculated |
34 |
Flat/mushroom-shaped |
75 |
Surface of the mucosa |
|
Smooth/unchanged |
0 |
With a depression |
21 |
With erosion/ulceration |
86 |
In the case of NENs located in the rectum, endoscopic ultrasound (EUS) is an important complement to colonoscopy. The latest ENETS (European Neuroendocrine Tumour Society) guidelines recommend EUS in RNENs starting from 5 mm [9]. EUS is recommended in three main clinical situations: at diagnosis — before planned treatment, in the case of incomplete lesion removal — before selecting “rescue” therapy, and as a follow-up examination. The examination is primarily used to determine the degree of local advancement (depth of infiltration of the intestinal wall, assessment of regional lymph nodes for metastases) and is useful in the differential diagnosis [8, 9, 42]. The sensitivity and specificity of this examination in determining the depth of infiltration are 87% and 93%, respectively [6]. A typical picture of NEN in EUS is a well-limited, hypoechoic, homogeneous lesion originating from the muscularis mucosa [43, 44]. After NEN removal, EUS is of little use in assessing the completeness of the procedure or looking for recurrence in the intestinal wall; it is mainly used to assess regional lymph nodes [45].
Minimal consensus statement on endoscopic examinations
3.3.2. Imaging diagnostics
Imaging diagnostics in patients with CRNEN is primarily used to determine the stage of the disease, and to assess and monitor the effects of treatment. Assessment of the primary lesion in the large intestine is usually based on endoscopic examinations. The following methods are used in imaging diagnostics.
Computed tomography (CT) before and after intravenous administration of a contrast agent
Multiphase CT examination (native, usually involving the abdominal cavity, in particular the liver) before and after the intravenous administration of a contrast agent, involving the abdominal cavity and pelvis in the arterial phase (approx. 25–30 s) and portal venous phase (approx. 55–75 s), because some highly vascularised NENs are only visible in one phase or the other after intravenous (i.v.) administration of the contrast agent. This examination is performed to detect metastatic lesions to regional lymph nodes and distant metastases, as well as in the assessment of NEN infiltration of the neighbouring organs [46, 47].
If there is a suspicion of dissemination of the neoplastic process (in particular, when the primary feature of the primary tumour — pT is assessed as T3–4, and/or N1, Mx), CT of the chest and base of the neck is performed to assess potential metastases.
In the case of lung parenchyma assessment, the examination is performed without administering a contrast agent, in the case of mediastinal assessment, examination before and after the i.v. administration of a contrast agent is preferable. Generally, the CT scan of the CNEN is performed in the same way as a standard CT scan for colorectal cancer. In the case of therapy, the standard examination description must include a summary of the effectiveness of therapy based on the Respons Evaluation Criteria In Solid Tumors (RECIST) 1.0 or RECIST 1.1 criteria [48].
Magnetic resonance imaging (MRI) before and after intravenous administration of a contrast agent
In patients with CRNETs, MRI mainly concerns the assessment of the pelvis and the abdominal cavity. In selected cases, MRI of the spine, central nervous system (CNS), or whole body should be performed. MRI is very helpful in the case of suspected metastases in the course of CRNENs due to the excellent contrast resolution, which is superior to that of CT.
MRI is the preferred examination in the assessment of local advancement in the case of RNENs that do not meet the size criterion of the primary lesion as T1a, size < 1 cm with R0 > 1 mm or R0 ≤ 1 mm. Rectal MRI includes all other T1 resections with R1 and R2, and at least T1 moderately differentiated NETs G2. MRI should be performed before and after intravenous administration of a contrast agent (DCE — dynamic contrast enhancement) along with diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) maps.
MRI is the optimal examination to assess the degree of local advancement, it is preferable to CT, it is more easily accessible, and it has a broader scope than EUS in the assessment of pelvic structures. Pelvic MRI should be performed according to a dedicated protocol to assess the rectum, mesorectum, and mesorectal fascia and the presence of regional lymph node metastases.
MRI is more sensitive and specific than CT in detecting metastases to the liver, pelvic lymph nodes, bones, and the CNS.
It is also recommended that MRI of the abdominal cavity be performed alongside chest CT in order to obtain a comprehensive assessment of possible distant metastases, especially small lesions in the liver. If liver metastases can be detected only by MRI, MRI should be used to assess their presence and continued as the main follow-up examination [46].
In patients with liver metastases, additional MRI of the spine or MRI of the whole body may be performed. Whole body MRI is less sensitive in assessing potential metastasis, so segmented MRI is preferred. After the identification of liver metastases, extended MRI is used to identify other potential distant metastases, especially to bone or peritoneal implants. In the case of the latter, the preferred examination is CT with i.v. administration of a contrast agent, without filling the loops of the intestine with the contrast agent. In every case, a neutral contrast medium is administered orally (water). The sensitivity of MRI in the detection of focal (metastatic) lesions seems to be similar to that of positron emission tomography/computed tomography (PET/CT) with the use of [68Ga]Ga-DOTATATE/TOC (91% vs. 92%), except for the liver, for which MRI sensitivity is higher, 99% vs. 92% and bone 96% vs. 82%, respectively [49].
CNS MRI is recommended in patients with symptoms suggesting metastases to the CNS, which, however, are rare even in gastroenteropancreatic neuroendocrine carcinomas (GEP-NECs) (in 4% of patients) [50]. The exception is advanced RNEN, in which metastases to the CNS should be ruled out each time, in almost every patient.
There is currently no consensus on the need for systematic performance of MRI of the CNS at first diagnosis in patients with advanced CNEC and distant metastases [51].
Minimal consensus statement on CT and MRI examinations
3.3.3. Radioisotope diagnostics
Somatostatin receptor imaging (SRI) is performed by means of scintigraphy (planar, single photon emission computed tomography (SPECT) or single photon emission computed tomography/computed tomography (SPECT/CT), or PET/CT.
PET/CT imaging with [68Ga]Ga-DOTA-SSA (somatostatin analogue) demonstrates higher sensitivity than scintigraphic examination and should be performed in cases of doubtful or negative results of scintigraphy as a conclusive examination, especially in the case of lesions < 1 cm in size [52, 53]
To assess the stage of advancement, SRI may be considered, especially in patients with T2 or higher clinical progression. In the case of radical surgery, there are no indications for patient follow-up using SRI.
SRI should be performed in all patients qualified for treatment with somatostatin analogues (SSAs) and peptide receptor radionuclide therapy (PRRT) [54].
Additionally, in the case of negative [68Ga]Ga-DOTA-SSA PET/CT and rapidly growing NETs and NECs, 2-deoxy-2-[fluorine-18]fluoro-D-glucose ([18F]FDG) PET/CT is indicated.
High accumulation of [18F] FDG is considered a hallmark of malignant neoplasms with high metabolism and proliferation. It correlates with higher stage of advancement and is an independent negative prognostic factor [53, 55, 56].
Minimal consensus statement of radioisotope examinations
4. Treatment
4.1. Surgical treatment of CNENs
Recommendations for the surgical treatment of CNENs are analogous to the recommendations for the treatment of colon adenocarcinoma [8]. It is recommended that a resection procedure (open or laparoscopic approach) with lymphadenectomy be performed in patients with tumours without distant metastases. In the case of CNENs G1 and G2 with distant metastases (usually to the liver), palliative resection with regional lymphadenectomy or, if possible, cytoreduction of the tumour are recommended, even if complete resection cannot be achieved.
In the case of infiltration of adjacent organs, if technically possible, multi-organ excision combined with left or right hemicolectomy or extension of the procedure according to the extent of the lymphatic drainage is also proposed [8].
4.1.1. Surgical treatment of RNENs
Most RNENs are detected early and can be treated endoscopically.
In the case of lesions that do not qualify for endoscopic treatment, the method of choice is local full-thickness resection using transanal endoscopic microsurgery (TEM) or resection of the rectum with the mesorectum using a classic (open) approach or with the use of minimally invasive techniques. The recommended treatment algorithm for rectal neuroendocrine neoplasms is presented in Figure 1 [8].
TEM is mainly used in so-called salvage therapy [57–60]. This treatment is used in the event of failure to achieve radical resection confirmed in pathomorphological examination or uncertainty regarding radical resection following an endoscopic procedure or in the case of local recurrence. Recent data show that endoscopic full-thickness resection (EFTR) is as effective as TEM in these patients [61].
In tumours larger than 2 cm in diameter (and according to some authors even > 1.5 cm), the muscularis mucosa is frequently infiltrated. In these cases, resection is recommended, preserving the sphincters if possible. The procedure of choice is anterior rectal resection with total mesorectal excision (TME) and the possible creation of a protective stoma [8, 62].
Radical resection is also recommended in the presence of other risk factors, including metastases to regional lymph nodes, such as the following: high mitotic index of the tumour (G2, G3, Ki-67 > 2%), infiltration of lymphatic and blood vessels, and positive HES77 expression [23, 63].
In the case of T3 and T4 tumours with local lymph node involvement, it is possible to achieve oncological radicality if there are no distant metastases.
If the tumour is located in the lower part or infiltrates the sphincters, it is advisable to perform abdomino-sacral or abdomino-perineal amputation.
In special situations, especially when the patient does not consent to radical surgery, the optimal solution is to use the TEM technique. However, the patient should be informed that radical rectal resection may be necessary in the event of an unfavourable histopathological examination result of the tumour resected using TEM.
One work has questioned the advantage of radical resection over local excision in the case of RNENs of 10–20 mm in diameter with and without lymph node involvement, because radical excision may have a negative impact on the quality of life [64].
In locally and systemically advanced tumours with distant metastases, radical resection surgeries are not recommended because in this group the survival is 6–9 months from diagnosis [9].
The indication for palliative surgery is tumour bleeding with ineffective local haemostasis or bowel obstruction. In RNETs G1 and G2 with metastases confined to the liver, local tumour excision followed by liver parenchyma resection (metastasectomy) may be considered or, in selected cases, liver transplantation [65].
The minimal consensus statement on surgical treatment
4.2. Endoscopic treatment of CRNENs
Patients with negligible risk of metastasis are eligible for endoscopic treatment of CRNEN. Surgical treatment is recommended in CNEN, even in the case of small/early diagnosed lesions, due to the high risk of metastases [66, 67]. In contrast, most RNENs can be treated endoscopically. At diagnosis, most of these lesions are up to 1 cm in size, which is associated with a very good prognosis [68, 69].
According to the latest reports, the highest R0 resection rates (over 90%) are provided by three main groups of endoscopic methods: modified endoscopic mucosal resection (mEMR), endoscopic submucosal dissection (ESD), and EFTR [61, 70–80]. RNEN removal cannot be performed using classic polypectomy or biopsy forceps due to the very low rate of R0 resections (approximately 30%). Classic endoscopic mucosal resection (EMR), consisting only of injecting fluid under the lesion followed by excision with a standard diathermy loop, should not be performed either, because comparative studies have shown that the modified EMR (mEMR) allows for significantly better results [81].
In rescue therapy (incomplete/uncertain radicality), ESD or EFTR is recommended — with the optimal procedure being full-thickness resection.
Indications for endoscopic treatment of RNENs should primarily take into account the size, stage (T and N), and proliferation index of the tumour (G). The optimal situation assumed by most guidelines is that the physician performing the colonoscopy correctly recognises the neuroendocrine lesion and refers the patient for EUS examination, and then for optimal treatment. However, in clinical practice, the correct diagnosis of RNENs at the time of the first endoscopy concerns only 18–30% of patients. Most lesions are removed incorrectly and non-radically, using a diathermy loop or biopsy forceps — the R0 resection rate is 24–30% [24, 82]. It may also occur that the correct endoscopic method of NEN treatment is used but R0 resection is not achieved.
In the case of RNENs of 5 mm in size and above, EUS should be performed before the planned endoscopic treatment, with assessment of the depth of infiltration and the condition of regional lymph nodes.
Patients with RNEN lesions up to 10 mm in size without risk factors are eligible for endoscopic treatment [74, 75]. The risk factors are as follows: infiltration of the muscular membrane proper, involvement of regional lymph nodes, infiltration of lymphatic and/or blood vessels, and a proliferation index above 2%. In the case of lesions of 10–20 mm in diameter, without muscularis propria infiltration and without metastases to regional lymph nodes, the tumour can be removed endoscopically, and after histopathological evaluation it is possible to decide on further treatment, taking into account (as well as the pathomorphological examination result) the patient’s age, general condition, and comorbidities. RNENs greater than 20 mm in size, all NENs G3, and lesions with markers of poor prognosis should undergo radical surgical treatment.
The scale developed by Chen et al. [41] may be useful in the initial endoscopic assessment of RNEN prognosis (Tab. 4).
One of the recommended methods of endoscopic treatment of RNENs is the above-mentioned ESD [75–78]. Considering the fact that ESD is a technically difficult and time-consuming procedure, recent reports on the very good effectiveness of the modified EMR (mEMR) method, achieving R0 resection in over 93% of cases, facilitate the use of this easier procedure, especially in the case of lesions with a diameter of up to 10 mm. There are three main types of mEMR: cap-assisted endoscopic mucosal resection (EMR-C), endoscopic mucosal resection with ligation (EMR-L), and endoscopic mucosal resection with precut of the mucosa around the lesion (EMR-precut) [70–74].
The highest R0 resection rate was achieved using EMR-L.
According to the latest publications, EFTR may become the method of choice in the treatment of RNEN due to its high efficiency (100% of R0 resection) and its safety [61, 79, 80].
In a situation where the rectal neuroendocrine lesion has been removed endoscopically and the histopathologist identifies the lack of radicality or cannot confirm the radicality following resection, the patient should be referred for a “rescue” procedure. A rescue endoscopic procedure is possible if there is a visible scar after NEN removal.
The rescue endoscopic procedures include ESD and EFTR. The use of EFTR provides a higher rate of R0 resection (100%) than ESD (90%) [61]. The effectiveness of EFTR is comparable to that of TEM, with a shorter procedure time (18 minutes on average), without the need for general anaesthesia or the use of an operating room [79]. It is worth emphasising that even in the case of small (up to 5 mm) RNENs that had been incorrectly removed with a diathermy loop, neoplastic infiltration was found in the material obtained during rescue treatment in more than 20% of patients [83]. The RNEN endoscopic treatment algorithm is presented in Figure 2 [84].
Minimal consensus statement on endoscopic treatment
4.3. Pharmacological treatment
4.3.1. Biotherapy
Somatostatin analogues
SSAs are used in NENs of the ascending colon. The CLARINET study demonstrated the antiproliferative activity of SSAs in all non-functional, highly or moderately differentiated GEP-NENs (Ki-67 < 10%). Hence, SSAs are the first-line treatment in NETs G1 and G2 [85, 86].
Both octreotide and lanreotide are effective in controlling the hormonal symptoms associated with NETs and inhibiting tumour growth. In clinical trials of the use of SSAs in the treatment of tumours located in the hindgut, there has been little evidence of their efficacy, among others, due to the size of the study group being too small. However, in the case of lesions with exposed somatostatin receptors, the use of SSA should be considered.
Targeted therapy — m-TOR inhibitors (everolimus)
Analysis of the subgroup of patients with advanced functional CRNETs (G1/G2) in the RADIANT-2 study demonstrated that the combination of everolimus and octreotide resulted in a significant increase in progression-free survival [87]. Additionally, the data from the RADIANT-4 study, including patients with non-functional CRNETs G1 and G2, confirmed the effectiveness of m-TOR inhibitors in this group of patients [88].
The m-TOR inhibitors are effective in the treatment of unresectable, locally advanced, and/or metastatic neoplasms that demonstrate progression during active follow-up or SSA therapy, and are recommended by scientific societies. The drug is approved in G1/G2 neoplasms, but currently it is not reimbursed in this indication [89].
4.3.2. Chemotherapy and immunotherapy
The importance of chemotherapy (ChT) in neoplasms with well- and intermediate grade of differentiation (G1/G2) is limited, and its application should apply to neoplasms resistant to other forms of systemic treatment, radioisotope therapy, or radiotherapy (NET G2 with Ki-67 >10 %, aggressive — progression according to RECIST within 3–6 months, or without SSTR expression).
ChT is primarily indicated in the treatment of patients diagnosed with NEC, as well as those who show features of NET G3 progression. In the case of NECs, the use of platinum derivatives should be standard practice [89].
As well as the combination of cisplatin and etoposide (important especially in small cell neoplasms with high Ki-67), the use of oxaliplatin (FOLFOX, XELOX) or irinotecan (FOLFIRI, IP) regimens can be considered [90–94]. Temozolomide- and/or capecitabine-based regimens are justified especially in patients with NETs G2/G3 [95, 96].
In some patients with neuroendocrine colorectal neoplasms with a higher degree of biological aggressiveness (including NEC, NET G3), DNA repair disorders, including microsatellite instability (MSI), are identified [97, 98]. Despite many reports on the possibility of using PD-1 inhibitors in this population and the site-agnostic approval of pembrolizumab in this indication, it is argued that the effectiveness of therapy is not equal in every organ location. Considering that the large intestine location is associated with potentially higher efficacy in the treatment of adenocarcinomas, the MSI assessment is indicated in situations where it is necessary to seek effective systemic treatment of advanced disease. Due to the approval of pembrolizumab as monotherapy for metastatic colorectal cancer with high microsatellite instability (MSI-H) or deficient DNA mismatch repair (dMMR), with very limited data in the population of patients with neuroendocrine neoplasms, consideration of this therapy must include a comprehensive patient assessment and evaluation of the tumour’s histoclinical and molecular features. The presence of the MSI-H phenotype in patients after radical treatment may indicate a potentially better prognosis and speak against the considered NEC/NET G3 adjuvant treatment.
Minimal consensus statement on pharmacological treatment:
4.4. Radioisotope therapy — PRRT
There are no prospective randomised clinical trials in patients with CRNENs. Few reports come from retrospective studies [99–102]. The observed survival parameters in a Polish multicentre study in this group of patients amounted to PFS 40.6 months and OS 131.2 months and were longer than those previously observed in the literature [100–104].
Qualification for treatment takes place according to the principles described in the Update of the diagnostic and therapeutic guidelines for gastro-entero-pancreatic neuroendocrine neoplasms (recommended by the Polish Network of Neuroendocrine Tumours) [27].
Minimal consensus statement on radioisotope therapy
4.5. Treatment monitoring
An endoscopic and imaging follow-up after R0 resection in patients with RNENs is presented in Table 5 [105]. Serum CgA provides additional information: if the elevated concentration observed before tumour resection normalises after surgery and then increases, it may suggest NET recurrence. According to the SEER Norwegian Register Cancer database, RNENs are associated with the best prognosis of all gastrointestinal tumours, with a five-year survival rate of 74–88% [106]. GEP-NENs are typically slow-growing tumours, so tumour marker determination should be performed infrequently, i.e. at intervals ranging from 4 to 12 months; and after radical surgery, every 6 months up to 2 years [20]. According to 2020 NANETS GUIDELINES, monitoring of CgA levels and 5-HIAA excretion should be considered if these results were abnormal at the onset of the disease [107].
Tumour size |
Grading |
Examination |
Incidence |
Duration |
< 1 cm |
1 or 2 |
No need for follow-up |
|
|
< 1 cm |
3 |
Colonoscopy |
Annual |
For 5 years |
1–2 cm |
Regardless of grade |
Colonoscopy EUS MRI |
After 12 months then every 5 months After 12 months After 12 months |
For 5 years |
> 2 cm |
1 or 2 |
Colonoscopy EUS MRI |
Every year Every year Every year |
For 5 years For 5 years For 5 years |
> 2 cm |
3 |
Colonoscopy EUS MRI |
Every 4–6 months for a year, then every year |
For 5 years For 5 years For 5 years |
Evidence quality and the strength of recommendations
#Evidence quality and the strength of recommendations has been established on the basis of the following tables according to ESMO and NCCN [108, 109] (Supplementary File — Tab. S1 and S2).
Conflict of interest
For the Conflict of Interest Statement, please see the Supplementary File.
Author contributions
All authors contributed to the idea, gathered the information, interpreted the data, and wrote and accepted the final version of the manuscript.