Abstract
There is no standardized treatment for grade 3 neuroendocrine tumors (G3 NETs). We aimed to describe the treatments received in patients with advanced G3 NETs and compare their efficacy. Patients with advanced digestive G3 NETs treated between 2010 and 2018 in seven expert centers were retrospectively studied. Pathological samples were centrally reviewed, and radiological data were locally reviewed. We analyzed RECIST-defined objective response (OR), tumor growth rate (TGR) and progression-free survival (PFS) obtained with first- (L1) or second-line (L2) treatments. We included 74 patients with advanced G3 NETs, mostly from the duodenal or pancreatic origin (71.6%), with median Ki-67 of 30%. The 126 treatments (L1 = 74; L2 = 52) included alkylating-based (n = 32), etoposide-platinum (n = 22) or adenocarcinoma-like (n = 20) chemotherapy, somatostatin analogs (n = 21), targeted therapies (n = 22) and liver-directed therapies (n = 7). Alkylating-based chemotherapy achieved the highest OR rate (37.9%) compared to other treatments (multivariable OR 4.22, 95% CI (1.5–12.2); P = 0.008). Adenocarcinoma-like and alkylating-based chemotherapies showed the highest reductions in 3-month TGR (P < 0.001 and P = 0.008, respectively). The longest median PFS was obtained with adenocarcinoma-like chemotherapy (16.5 months (9.0–24.0)) and targeted therapies (12.0 months (8.2–15.8)), while the shortest PFS was observed with somatostatin analogs (6.2 months (3.8–8.5)) and etoposide-platinum chemotherapy (7.2 months (5.2–9.1)). Etoposide-platinum CT achieved shorter PFS than adenocarcinoma-like (multivariable HR 3.69 (1.61–8.44), P = 0.002) and alkylating-based chemotherapies (multivariable HR 1.95 (1.01–3.78), P = 0.049). Overall, adenocarcinoma-like and alkylating-based chemotherapies may be the most effective treatments for patients with advanced G3 NETs regarding OR and PFS. Etoposide-platinum chemotherapy has poor efficacy in this setting.
Introduction
The classification of neuroendocrine neoplasms (NENs) relies on differentiation (well-differentiated neuroendocrine tumors (NETs) or poorly-differentiated neuroendocrine carcinomas (NECs)) and the proliferation index, which is determined by the Ki-67 index and/or the mitotic count. The 2010 WHO classification recognized well-differentiated G1 NETs (Ki-67 ≤ 2%), G2 NETs (Ki-67 between 3 and 20%) and poorly-differentiated G3 NECs (Ki-67 > 20%) (Rindi et al. 2010). Within this classification, patients with high Ki-67 (>20%) but with well-differentiated morphology were not considered as a separate entity and were mainly grouped together with G3 NEC regardless of morphology. However, recently G3 NENs have shown to be a heterogeneous group, with approximately 20% of them being well-differentiated (Sorbye et al. 2013, Vélayoudom-Céphise et al. 2013, Basturk et al. 2015, Heetfeld et al. 2015, Milione et al. 2017, Raj et al. 2017). Subsequently, the existence of G3 NETs, distinct from NECs, has been recognized within G3 NENs in the WHO classification, in 2017 for pancreatic NEN and 2019 for all digestive NEN (Klöppel et al. 2017, WHO Classification of Tumours 2019). While in G1 and G2 NENs the therapeutic options have been increasing over the last decade (Pavel et al. 2016), the etoposide-platinum chemotherapy (CT) combination has long been the reference therapy for G3 NENs (regardless of morphology) based on the assumption that their clinical behavior is similar to that of small-cell lung carcinomas (Coriat et al. 2016, Garcia-Carbonero et al. 2016, Stelwagen et al. 2021).
The recognition of G3 NETs is of paramount importance because their molecular features and prognosis largely differ from those of NECs and is much closer to those of G2 NETs (Coriat et al. 2016, Tang et al. 2016, Konukiewitz et al. 2017). However, the most appropriate management for patients with G3 NETs is currently undefined, due to currently limited specific data. They appear to be much less sensitive to etoposide-platinum CT than NECs (Hijioka et al. 2017, Raj et al. 2017, Elvebakken et al. 2020, Lacombe et al. 2021). Conversely, G3 NETs may rather be treated as G2 NETs since they seem to correspond to a more biologically aggressive counterpart of the NET continuum (Coriat et al. 2016). The objective of this study was to describe the treatments received in patients with centrally reviewed advanced G3 NETs, and compare their efficacy, in an international multicenter retrospective cohort.
Patients and methods
Patients
We included all consecutive patients with digestive (or unknown primary) well-differentiated G3 NETs, locally advanced unresectable or metastatic, treated between 2010 and 2018 in seven European centers with recognized expertise on NENs. Each patient had to undergo regular follow-up using conventional imaging, with a 3 ± 1 months interval. Patients were eligible for screening if pathology slides were available for central review. All G3 NETs confirmed by central pathological review, as defined by the coexistence of a well-differentiated morphology and a G3 (Ki-67 > 20% and/or mitotic count > 20/10 high-power fields), were included.
Exclusion criteria were the absence of tissue available for central pathology review, NEC, G1/G2 NET, patients with genetic predisposition, patients with other active cancers, and patients without regular follow-up by conventional imaging. G3 NET patients who were cured by surgery or ablation could not be included during the initial phase of management, but later during metachronous metastatic recurrence. Patients with G3 NETs arising in the setting of grade progression were not included in this study.
Data collection
This study was performed among the NET-CONNECT network, which consists of an international initiative of clinical research and educational programs. This study was performed according to the Helsinki convention and following Institutional Review Board approval (CEERB Paris Nord, IRB no. 00006477-15-073). The study protocol was approved by the ethics or audit committees at each institution, where required. Consent was obtained from each patient after a full explanation of the purpose of the study.
Epidemiological, clinical, and tumor data at baseline were collected anonymously for each patient. In addition, detailed information on the first-line (L1) and second-line (L2) of non-surgical treatments of the metastatic disease were collected, as well as follow-up data. Treatments were regrouped as alkylating-based CT (temozolomide-, dacarbazine- or streptozotocin-based), etoposide-platinum CT, adenocarcinoma-like CT (FOLFOX- or FOLFIRI-based), somatostatin analogs (lanreotide autogel or octreotide LAR), targeted therapies (everolimus, sunitinib or peptide-radionuclide radiation therapy (PRRT)) and liver-directed therapy (transarterial (chemo)-embolization or selective internal radiation therapy).
A central virtual pathological review was performed for all cases by two experienced pathologists (JC, AC). For each case, one hematoxylin and eosin stain slide and one Ki-67 slide were digitalized and reviewed. To distinguish well- from poorly-differentiated NENs, we strictly followed the descriptions provided by the WHO classification of both lung and digestive/pancreatic NENs. The criteria are concordant between the two localizations and were recently completed by criteria proposed by pathological revision of the cases included in the NORDIC study (Elvebakken et al. 2020). NETs were defined by organoid growth patterns including nesting, trabecular, gyriform and/or pseudo-glandular architecture without necrosis. Tumor cells showed minimal to moderate atypia. NECs were defined by large expansile confluent nests, cribriform nests, vessels distant to tumor and by desmoplastic stroma accompanied by frequent necrosis. In large-cell type NECs, tumor cells were polygonal containing large nuclei with vesicular chromatin and frequent prominent nucleoli. In small-cell type NECs, cells had scant cytoplasm, round or elongated nuclei, fine granular chromatin and inconspicuous nucleoli. All cases were reviewed by the two experienced pathologists at the same time. When the diagnosis proposal was discordant, a consensus was reached by discussion according to the description proposed above.
Radiological review of imaging was performed locally by experienced radiologists. Target lesion identification and follow-up were performed using the same imaging technique (CT or magnetic resonance) for each patient, as per RECIST 1.1 recommendations (Eisenhauer et al. 2009). As part of metastatic extent, the liver involvement was classified as 0–10%, 11–25%, 26–50% or ≥50% (Zappa et al. 2017). All contrast-enhanced imaging scans were reviewed to assess the best objective response and progression as defined by RECIST (Eisenhauer et al. 2009). In addition, tumor evolution was also assessed using the tumor growth rate (TGR), which was calculated as previously published (Lamarca et al. 2019a) and expressed as the percentage change in tumor size over 1 month (%/month). Three TGRs were calculated: at baseline (comparing imaging performed at baseline and within the 6 previous months), at 3 months (comparing imaging performed at 3 months and baseline), and at the best response (comparing imaging showing the best RECIST-defined response and at baseline). 18Fluorodeoxuglucose positron-emitting tomography and SST receptor nuclear imaging were reviewed and positive tumor uptake (tumor uptake at least equal to that of the liver) was noted.
Statistical analyses
The primary endpoint was the description of the treatments received as L1 or L2 in patients with G3 NET. Secondary endpoints included RECIST-defined objective response (OR) rate, TGR, progression-free survival (PFS) and overall survival (OS).
The categorical variables were described as frequencies (percentages) and compared using the Chi-2 test. The continuous variables were described as medians (interquartile range (IQR)) and compared using the Mann–Whitney U-test. Variables associated with each main treatment group were explored using backward stepwise multivariable logistic regression models. OR rate, TGR and PFS analyses were performed for treatments received as L1 and L2. Factors associated with OR were explored using univariable and backward stepwise multivariable logistic regression models. Comparisons in TGR at baseline and TGR at 3 months for each treatment were performed using the Wilcoxon matched-pairs signed rank test. OS was assessed for L1 treatments only.
PFS was measured from the date of the first cycle of L1 (or L2) to the date of progression or death or censored at the date of last follow-up without progression or death. OS was measured from the date of beginning of L1 to the date of death or censored at the date of last follow-up without death. Median PFS and median OS were estimated using the Kaplan–Meier method. The 95% CIs for median PFS and OS were derived from log hazards. Surviving patients were censored at the last follow-up. The cut-off date was set at July 30, 2020. Survival rates were compared using the log-rank test. Factors associated with PFS and OS were explored using Cox proportional hazard multivariable regression models.
All statistical tests were two-tailed. Any P-value < 0.05 was considered to be statistically significant. Analyses were performed using Prism® (v.6, Graphpad™) and SPSS® (v.20, IBM™) softwares.
Results
Patients
Among eighty-six patients identified, five were not included because no pathological material was available, and seven patients were excluded following central pathological review (three patients had G2 NETs and four patients had NECs) (Fig. 1). Among the 74 patients identified, 47 (63.5%) were male and median age was 55.8 years (Table 1). G3 NETs had a median Ki-67 of 30% (IQR, 25–35.8) and were mostly of duodenal or pancreatic origin (71.6%). At initial diagnosis, most patients (66.2%) had cancer-related symptoms and 17.6% of patients presented with the functioning syndrome. The majority of patients (90.5%) had synchronous metastases, with 51.4% having extra-hepatic dissemination and 22.2% of patients with liver burden ≥ 50%. Thirteen patients had undergone previous primary tumor surgery, including six patients who had synchronous metastases. Among 38 patients with available pre-baseline imaging, 28 (73.8%) had documented progressive disease and 10 patients had stable disease at the time of initiation of treatment. 18Fluorodeoxuglucose positron-emitting tomography showed tumor uptake in 32/36 patients (88.9%). SST receptor nuclear imaging showed tumor uptake at least equal to that of the liver in 54/61 patients (88.5%), of which 36 (59%) had homogeneous uptake higher than that of the liver (Table 1).
Flow chart of the study. NEC, poorly-differentiated neuroendocrine carcinoma; NET, well-differentiated neuroendocrine tumor.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Flow chart of the study. NEC, poorly-differentiated neuroendocrine carcinoma; NET, well-differentiated neuroendocrine tumor.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Flow chart of the study. NEC, poorly-differentiated neuroendocrine carcinoma; NET, well-differentiated neuroendocrine tumor.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Baseline characteristics of 74 patients with advanced G3 neuroendocrine tumors.
| Characteristics | n total | Result |
|---|---|---|
| Age (years), median (IQR) | 74 | 55.8 (44–64.1) |
| Male gender, n (%) | 74 | 47 (63.5) |
| Immunohistochemistry, n (%) | ||
| Chromogranin A positive | 74 | 69 (93.2) |
| Synaptophysin positive | 66 | 66 (100) |
| Ki-67 (%), median (IQR) | 74 | 30 (25–35.8) (range, 21–80) |
| Primary tumor, n (%) | 74 | |
| Pancreas/duodenum | 53 (71.6) | |
| Small or large bowel | 12 (16.2) | |
| Other/unknown | 9 (12.2) | |
| Previous primary tumor resection, n (%) | 74 | 13 (17.6) |
| Performance status, n (%) | 73 | |
| PS-0 | 39 (53.4) | |
| PS-1 | 27 (37) | |
| PS-2/3 | 7 (9.6) | |
| Cancer-related symptoms, n (%) | 74 | 49 (66.2) |
| Functioning syndrome, n (%) | 74 | 13 (17.6)a |
| Tumor stage, n (%) | 74 | |
| Liver metastases only | 35 (47.3) | |
| Extrahepatic metastases | 38 (51.4) | |
| Synchronous metastases | 73 | 66 (90.4) |
| Liver involvement, n (%) | 72 | |
| 0–10% | 26 (36.1) | |
| 11–25% | 16 (22.2) | |
| 26–50% | 14 (19.4) | |
| ≥50% | 16 (22.2) | |
| RECIST-defined progression within 12 months, n (%) | 38 | 28 (73.8) |
| FDG uptake on PET, n (%) | 36 | 32 (88.9) |
| Positive SST receptor imaging, n (%) | 61 | 54 (88.5) |
aCarcinoid syndrome (n = 10), gastrinoma (n = 4), VIPoma (n = 2), calcitonin-producing (n = 2), ACTH-producing (n = 2).
FDG, 18fluorodeoxyglucose; IQR, interquartile range; PET-CT, positron-emitting tomography; SST, somatostatin.
Treatments received
Overall, 126 L1 and L2 non-surgical treatments (n = 74 and 52, respectively) were studied, as depicted in Table 2. CT was the prominent therapeutic modality and consisted in alkylating-based (n = 32), etoposide-platinum (n = 22) or adenocarcinoma-like CT (n = 20). The latter category mostly comprised FOLFOX or FOLFIRI regimen, of which 15 were associated to bevacizumab. Alkylating-based (n = 17/53, 32%) and etoposide-platinum CT (n = 16/53, 30%) were the most frequent L1 treatment options for patients with duodenal or pancreatic G3 NETs, while SST analogs (n = 9/21, 43%) were more commonly selected for other primary sites (Table 2). Median Ki-67 index for patients grouped by first-line therapy is shown in Supplementary Table 1 (see section on supplementary materials given at the end of this article). Targeted therapies and adenocarcinoma-like CT were the most frequent L2 treatments. Among the 50 patients with both L1 and L2 documented, no preferred therapeutic sequence could be clearly noted (Fig. 2).
Description of the 126 treatments received as first- or second-line therapy in 74 patients with advanced G3 neuroendocrine tumors.
| Treatment groups | n | L1 + L2 (n = 126) | Pancreas/duodenal primary (n = 92) | Other primary (n = 34) |
|---|---|---|---|---|
| Alkylating-based chemotherapy (n = 32) | ||||
| Streptozotocin + 5FU or adriamycin | 6 | 32 (25.4%) | 26 (28.3%) | 6 (17.6%) |
| L1: 17 | L1: 6 | |||
| Temozolomide ± capecitabine | 20 | |||
| L2: 9 | L2: 0 | |||
| Dacarbazine ± 5FU | 6 | |||
| Etoposide-platinum chemotherapy (n = 22) | ||||
| Etoposide-platinum | 22 | 22 (17.5%) | 17 (18.5%) | 5 (14.7%) |
| L1: 16 | L1: 3 | |||
| L2: 1 | L2: 2 | |||
| Somatostatin analogues (n = 21) | ||||
| Lanreotide autogel | 14 | 21 (16.7%) | 11 (12%) | 10 (29.4) |
| Octreotide LAR | 7 | L1: 10 | L1: 9 | |
| L2: 1 | L2: 1 | |||
| Adenocarcinoma-like chemotherapy (n = 20) | ||||
| FOLFOX | 4 | 20 (15.9%) | 14 (15.2%) | 6 (17.6%) |
| FOLFOX + bevacizumab | 13 | L1: 6 | L1: 2 | |
| FOLFIRI + bevacizumab | 2 | L2: 8 | L2: 4 | |
| FOLFIRINOX | 1 | |||
| Targeted therapy (n = 22) | ||||
| Sunitinib | 10 | 22 (17.5%) | 20 (21.7%) | 2 (5.9%) |
| Everolimus | 5 | L1: 3 | L1: 0 | |
| Peptide receptor radionuclide therapy | 7 | L2: 17 | L2: 2 | |
| Liver-directed therapy (n = 7) | ||||
| Transarterial embolization | 6 | 7 (5.6%) | 2 (2.2%) | 5 (14.7%) |
| Selective internal radiation therapy | 1 | L1: 1 | L1: 1 | |
| L2: 1 | L2: 4 | |||
| Other treatments (n = 2) | ||||
| Lurbinectedin (PM1183-B-005-14 trial) | 1 | 2 (1.6%) | 2 (2.2%) | 0 |
| Durvalumab + Tremelimumab (DUNE trial) | 1 | L1: 0 | ||
| L2: 2 | ||||
5FU, 5-fluorouracile; L1, first-line therapy; L2, second-line therapy.
Sankey diagram presenting treatments performed successively in 50 patients who received first- and second-line therapy for an advanced G3 neuroendocrine tumor. ADK-like, adenocarcinoma-like; CT, chemotherapy; SST, somatostatin.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Sankey diagram presenting treatments performed successively in 50 patients who received first- and second-line therapy for an advanced G3 neuroendocrine tumor. ADK-like, adenocarcinoma-like; CT, chemotherapy; SST, somatostatin.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Sankey diagram presenting treatments performed successively in 50 patients who received first- and second-line therapy for an advanced G3 neuroendocrine tumor. ADK-like, adenocarcinoma-like; CT, chemotherapy; SST, somatostatin.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
No factors were clearly associated with the use of alkylating-based CT in comparison with other treatments, although it tended to be more frequently used in first-line (OR 2.12, 95% CI (0.88–5.09), P = 0.093) and in patients with cancer-related symptoms at diagnosis (OR 1.91, 95% CI (0.77–4.74), P = 0.165) (Fig. 3). Etoposide-platinum CT was more likely to be used for G3 NETs with Ki-67 ≥ 30% (OR 3.10, 95% CI (1.07–8.94), P = 0.037) and in the first-line (OR 6.27, 95% CI (1.69–23.27), P = 0.006), and tended to be less used in patients with cancer-related symptoms at diagnosis (OR 0.42, 95% CI (0.16–1.16), P = 0.094). SST analogs were significantly more used in first-line (OR 9.39, 95% CI (2.02–43.53), P = 0.004), and tended to be more frequently used for NET from intestinal/other/unknown rather than from pancreatic/duodenal origin (OR 0.42, 95% CI (0.14–1.30); P = 0.131). Finally, adenocarcinoma-like CT was more frequently used for patients with cancer-related symptoms at diagnosis (OR 3.6, 95% CI (1.01–12.91), P = 0.049) and less frequently used in case of positive SST receptor imaging (OR 0.20, 95% CI (0.05–0.77), P = 0.020) (Fig. 3).
Factors associated with the use of different systemic treatments in patients with advanced G3 NETs. Plain arrows/circles: factors significantly associated (P < 0.05) at multivariable analysis. Doted arrows/circles: factors with statistical tendency for association (P < 0.20) at multivariable analysis.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Factors associated with the use of different systemic treatments in patients with advanced G3 NETs. Plain arrows/circles: factors significantly associated (P < 0.05) at multivariable analysis. Doted arrows/circles: factors with statistical tendency for association (P < 0.20) at multivariable analysis.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Factors associated with the use of different systemic treatments in patients with advanced G3 NETs. Plain arrows/circles: factors significantly associated (P < 0.05) at multivariable analysis. Doted arrows/circles: factors with statistical tendency for association (P < 0.20) at multivariable analysis.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Effect of treatments on tumor size and growth
Among 115 evaluable L1 or L2 treatments, those that obtained the highest RECIST-defined OR rates were alkylating-based CT (11/29, 37.9%), liver-directed therapy (3/7, 42.9%) and adenocarcinoma-like CT (5/20, 25%) (Table 3 andFig. 4). The best response was obtained after a median time on treatment of 4.2 months (IQR, 3–6.2), without significant differences between treatment modalities (data not shown). The highest disease control rates (OR + stable disease rates) were achieved by adenocarcinoma-like CT (18/20, 90%) and targeted therapy (20/22, 90.9%) subgroups. In the subgroup of patients with the highest Ki-67 (last quartile, ≥ 36%) and evaluable response, OR was observed in 2/5, 1/6, 0/4, 3/7 and 1/3 of those receiving alkylating-based CT, etoposide platinum CT, SST analogs, adenocarcinoma-like CT and targeted therapies, respectively.
Best morphological response and progression-free survival achieved by treatments received as first- or second-line in patients with advanced G3 NET.
| Best response (L1 + L2) | PFS (L1 + L2) | PFS (L1) | PFS (L2) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| N | OR, n (%) | SD, n (%) | PD, n (%) | N | Median (months), (95% CI) | N | Median (months), (95% CI) | N | Median (months), (95% CI) | |
| Alkylating-based CT | 29 | 11 (37.9) | 12 (41.4) | 6 (20.7) | 32 | 7.8 (4.2–11.5) | 23 | 6.4 (1.6–11.3) | 9 | 7.8 (5.6–10.1) |
| Etoposide-platinum CT | 17 | 2 (11.8) | 12 (70.6) | 3 (17.6) | 22 | 7.2 (5.2–9.1) | 19 | 6.6 (4.3–8.9) | 3 | 7.4 (na–na) |
| SST analogs | 20 | 0 | 15 (75) | 5 (25) | 21 | 6.2 (3.8–8.5) | 19 | 6.2 (3.9–8.4) | 2 | 3.5 (na–na) |
| Adenocarcinoma-like CT | 20 | 5 (25) | 13 (65) | 2 (10) | 20 | 16.5 (9.0–24.0) | 8 | 23.5 (12.8–34.1) | 12 | 7.7 (0–16.8) |
| Targeted therapy | 22 | 3 (13.6) | 17 (77.3) | 2 (9.1) | 22 | 12.0 (8.2–15.8) | 3 | 8.4 (0.7–16.0) | 19 | 13.5 (10.6–16.5) |
| Liver-directed therapy | 7 | 3 (42.9) | 3 (42.9) | 1 (14.3) | 7 | 9.7 (0.3–19.0) | 2 | 9.7 (na–na) | 5 | 3.4 (0–25.8) |
| Overall | 115 | 24 (20.9) | 72 (62.6) | 19 (16.5) | 124 | 8.8 (0.9–7.0) | 74 | 8.8 (7.0–10.5) | 50 | 11.9 (8.1–15.8) |
CT, chemotherapy; L1, first-line therapy; L2, second-line therapy; N, number of patients; OR, objective response; PD, progressive disease; PFS, progression-free survival; SD, stable disease; SST, somatostatin.
Waterfall plot of the maximal change in size of target lesions, obtained with 115 treatments performed in first-line (black) or second-line (grey) for 74 patients with advanced G3 neuroendocrine tumor. CT, chemotherapy.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Waterfall plot of the maximal change in size of target lesions, obtained with 115 treatments performed in first-line (black) or second-line (grey) for 74 patients with advanced G3 neuroendocrine tumor. CT, chemotherapy.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Waterfall plot of the maximal change in size of target lesions, obtained with 115 treatments performed in first-line (black) or second-line (grey) for 74 patients with advanced G3 neuroendocrine tumor. CT, chemotherapy.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
At multivariable analysis, alkylating-based CT was independently associated with a significantly higher probability of OR in comparison with other treatment modalities (OR 4.22, 95% CI (1.47–12.17), P = 0.008), adjusted for Ki-67, primary NET origin, L1 or L2 setting and previous primary tumor surgery (Supplementary Table 2).
Imaging scans performed at pre-baseline, baseline and 3-month were available for 71 treatments. The median baseline TGR was significantly higher when CT treatments were used (6.3%/month, IQR (2–28.7)) compared with other treatment types (4.6%/month, IQR (0–9.3), P = 0.03) (Fig. 5
Difference of tumor growth rate (TGR) measured at baseline and at 3 months, for each treatment modality. CT, chemotherapy; s.d., standard deviation; TGR, tumor growth rate.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Difference of tumor growth rate (TGR) measured at baseline and at 3 months, for each treatment modality. CT, chemotherapy; s.d., standard deviation; TGR, tumor growth rate.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Difference of tumor growth rate (TGR) measured at baseline and at 3 months, for each treatment modality. CT, chemotherapy; s.d., standard deviation; TGR, tumor growth rate.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Progression-free survival
The longest median PFS was obtained with adenocarcinoma-like CT (16.5 months, 95% CI (9.0–24.0)) and targeted therapies (12.0 months, 95% CI (8.2–15.8)), while the shortest median PFS was observed for SST analogs (6.2 months, 95% CI (3.8–8.5)) and etoposide-platinum CT (7.2 months, 95% CI (5.2–9.1)) (Table 3 and Fig. 6A). Because these different subgroups are not directly comparable due to high selection bias (discrepancies in baseline prognostic characteristics associated with the selection of individual treatment options), we further explored PFS among patients who received CT.
Among the 74 CT treatments received in any line (L1 + L2), adenocarcinoma-like CT achieved longer PFS than etoposide-platinum CT (median 16.5 months, 95% CI (9.0–24.0) vs 7.2 months, 95% CI (5.2–9.1; P = 0.002), while it was not statistically different with alkylating based CT (median 7.8 months, 95% CI (4.2–11.5); P = 0.304) (Fig. 6B). Similar results were observed when focusing on the 50 CT treatments received in L1 (Fig. 6C). In patients with pancreatic or duodenal G3 NETs, etoposide-platinum CT was associated with shorter PFS (median 6.6 months, 95% CI (4.3–8.9)) in comparison with adenocarcinoma-like CT (median 18.1 months, 95% CI (9.8-26.5); P = 0.005), but also with alkylating-based CT (median 9.3 months, 95% CI (5.9–12.6); P = 0.046), without difference between the two latter (Fig. 6D). On multivariable analysis adjusted for age, Ki-67, tumor origin and treatment line (L1 or L2), etoposide-platinum CT was significantly associated with a higher risk of progression, when compared with both adenocarcinoma-like CT (HR 3.69, 95% CI (1.61–8.44), P = 0.002) and alkylating-based CT (HR 1.95, 95% CI (1.01–3.78), P = 0.049) (Table 4). Conversely, no significant differences between alkylating-based and adenocarcinoma-like CT (HR 1.89, 95% CI (0.90–3.97), P = 0.092) were identified.
Progression-free survival for (A) all treatments received as first- or second-line in patients with advanced G3 NET and in patients who received chemotherapy (B) in all lines (L1+L2) for G3 NET of all origins, (C) in first line for G3 NET of all origins, and (D) in all lines for pancreatic or duodenal G3 NET. ADK, adenocarcinoma; ALK, alkylating; EP, etoposide-platinum; SST, somatostatin.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Progression-free survival for (A) all treatments received as first- or second-line in patients with advanced G3 NET and in patients who received chemotherapy (B) in all lines (L1+L2) for G3 NET of all origins, (C) in first line for G3 NET of all origins, and (D) in all lines for pancreatic or duodenal G3 NET. ADK, adenocarcinoma; ALK, alkylating; EP, etoposide-platinum; SST, somatostatin.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Progression-free survival for (A) all treatments received as first- or second-line in patients with advanced G3 NET and in patients who received chemotherapy (B) in all lines (L1+L2) for G3 NET of all origins, (C) in first line for G3 NET of all origins, and (D) in all lines for pancreatic or duodenal G3 NET. ADK, adenocarcinoma; ALK, alkylating; EP, etoposide-platinum; SST, somatostatin.
Citation: Endocrine-Related Cancer 28, 8; 10.1530/ERC-21-0109
Multivariate analysis of baseline variables influencing the risk of progression, achieved by 74 chemotherapy treatments received as first- or second-line for G3 neuroendocrine tumors.
| Multivariate Cox model | |||
|---|---|---|---|
| HR | 95% CI | P-value | |
| Age (each added year) | 1.03 | 1.01–1.05 | 0.013 |
| Ki-67 ≥ 30% (vs < 30%) | 0.71 | 0.41–1.25 | 0.238 |
| Pancreas/duodenum primary NET (vs intestinal/other/unknown) | 1.13 | 0.53–2.42 | 0.758 |
| Second-line treatment (vs first line) | 1.13 | 0.60–2.11 | 0.714 |
| Treatment group | |||
| Alkylating-based CT (vs adenocarcinoma-like CT) | 1.89 | 0.90–3.97 | 0.092 |
| Etoposide-platinum CT (vs alkylating-based CT) | 1.95 | 1.01–3.78 | 0.049 |
| Etoposide-platinum CT (vs adenocarcinoma-like CT) | 3.69 | 1.61–8.44 | 0.002 |
CT, chemotherapy; HR, hazard ratio; NET, neuroendocrine tumor.
Overall survival
From the beginning of L1, 43 patients died within a median follow-up of 24.2 months (IQR, 12.5–41.3). The estimated median OS was 36.0 months (95% CI, 17.4–54.5), with 1- and 5-year OS rates of 81.7 ± 4.6% and 25.1 ± 7.0%, respectively (Supplementary Fig. 2A). At multivariable analysis adjusted for age, tumor origin and liver involvement, previous primary tumor surgery (HR 0.28, 95% CI (0.25–0.97), P = 0.024) and performance status of 0 (HR 0.49, 95% CI (0.67–2.52), P = 0.039) were significantly associated with a prolonged OS (Supplementary Table 3). Among the 50 patients who received CT as L1, no significant difference existed between the different CT regimens in terms of OS (Supplementary Fig. 2B).
Discussion
In this original and large international multicentric series, which included patients with advanced G3 NETs after central pathology review, alkylating-based and adenocarcinoma-like CT appeared to be the most effective treatments in the first or second line. This work provides some useful characterization of patients with advanced G3 NETs and factors associated with the choice of different treatments. It also describes benchmark regarding estimated OS (median 36 months), which is concordant with previously described series (22 to 41 months) and longer than for patients with NEC (Vélayoudom-Céphise et al. 2013, Nuñez-Valdovinos et al. 2018, Rogowski et al. 2019, Sahu et al. 2019, Elvebakken et al. 2020). Hence, patients with G3 NETs will likely receive more than one line of treatment, which highlights the importance of adequate treatment selection.
The evaluation of antitumor efficacy by RECIST criteria showed that alkylating-based CT achieved the highest OR rate (37.9%), which was significantly superior to the other treatments at multivariable analysis. In previous studies, alkylating-based CT achieved response rates of 30–50% and PFS of approximately 9–15 months (Raj et al. 2017, 2019, Sahu et al. 2019, Apostolidis et al. 2020, Liu et al. 2021). Hence, it may be relevant for patients with G3 NETs, especially in cases where tumor shrinkage is the main therapeutic objective, for example, in patients with symptoms related to bulky liver enlargement, or in those who could be considered for secondary surgical resection in case of sufficient tumor shrinkage.
Adenocarcinoma-like CT achieved a lower OR rate of 25% and a 90% disease control rate. However, the relevance of RECIST criteria has long been challenged in the field of NETs (de Mestier et al. 2014). In this view, TGR may better evaluate the antitumor effect of treatments and predict PFS (Lamarca et al. 2019a,b). Herein, adenocarcinoma-like CT was associated with the highest decreases in TGR between baseline and at 3 months. Moreover, it provided the longest PFS in comparison with other treatments (median 16.5 months). Previous data regarding the role of adenocarcinoma-like CT in G3 NETs are very limited, with one preliminary study reporting an OR rate of 53% and a PFS of 6 months with FOLFOX (Apostolidis et al. 2020). In another recent study, FOLFOX was reported in seven G3 NET patients with an OR rate of 29% and PFS of 13 months (Liu et al. 2021). Of note, in our study, adenocarcinoma-based CT mostly consisted in oxaliplatin-based CT, while irinotecan-containing combinations only accounted for a minority, which does not allow to draw any conclusion regarding the efficacy of the latter. Finally, the use of oxaliplatin-based chemotherapy beyond 4–6 months likely results in peripheral neurotoxicity. Hence, such treatment should be used cautiously, especially since patients with NETs may have prolonged survival, and maintenance chemotherapy with fluoropyrimidine alone should be considered after initial treatment with oxaliplatin.
Of note, most patients in the adenocarcinoma-like CT subgroup in our study also received bevacizumab, and it was not possible to measure the specific impact attributable to this antiangiogenic agent. The BETTER phase II study reported an OR rate of 56% and a median PFS of 23.7 months in pancreatic G1/G2 NETs (Ducreux et al. 2014), and one retrospective cohort described a 64% OR rate and PFS of 14 months in 11 patients with advanced NECs (Collot et al. 2018). Conversely, the role of bevacizumab-associated CT has not been explored in G3 NETs to date but is currently under evaluation by the BETTER-2 trial (NCT03351296). Overall, in this retrospective analysis, adenocarcinoma-like CT appears as one of the most effective treatment options for advanced G3 NETs, especially when the main therapeutic objective is achieving prolonged tumor control.
Among the three CT regimens, etoposide-platinum achieved the lowest (12%) OR rate, a limited decrease in TGR and a significantly shorter PFS than the other CT regimens, although no significant difference in OS existed between the different CT regimens. Several comparative studies reported significantly lower OR rates with etoposide-platinum CT in G3 NETs (0–20%) than in NECs (35–50%) (Vélayoudom-Céphise et al. 2013, Heetfeld et al. 2015, Hijioka et al. 2017, Raj et al. 2017, Elvebakken et al. 2020, Lacombe et al. 2021). Similarly, a preliminary study reported that etoposide-platinum CT achieved significantly shorter PFS (median 5.2 months) in comparison with other treatments (mainly alkylating agents or FOLFOX, median 9 months, P = 0.011) (Apostolidis et al. 2020). Overall, we can probably conclude that etoposide-platinum CT is not a relevant option for the early treatment of patients with advanced G3 NETs and should be reserved for NECs.
The use of SST analogs was higher than originally expected in this study. However, its use achieved no OR, and the lowest TGR variation and PFS (median 6.2 months). Similarly, in a recent NET CONNECT study, the median PFS was low in patients with pancreatic G3 NETs treated with SST analogs (4 months), and significantly lower compared to pancreatic G2 (Ki-67 ≥ 10%) NETs (12.4 months, P = 0.0007) (Merola et al. 2020). Hence, the role of SST analogs appears limited in this setting. However, SST analogs have low toxicity and although PFS appears short, this could be an option for lower volume, SST-receptor positive disease. Occasionally, some patients might experience disease stability over several months and assuming close follow-up, this may be an appropriate option following a discussion with the patient, especially when relatively non-toxic treatment is desired (McGarrah et al. 2020).
In the present study, the targeted therapy subgroup was heterogeneous since it included sunitinib, everolimus and PRRT. Unfortunately, a limited number of patients did not allow for individual analysis of each of these treatments and further larger series in G3 NET will be required to clarify their individual role. Encouraging outcomes were previously reported with sunitinib for advanced G3 NETs, with a possibly higher OR rate (above 50%) in comparison with G1/G2 NETs (Mizuno et al. 2018, Pellat et al. 2018). The role of cabozantinib, another tyrosine kinase inhibitor, is being explored in patients with G3 NETs in the ongoing CABINET trial (NCT03375320). Similarly, PRRT seems relevant for SST-receptor-expressing advanced G3 NETs. While specific data are scarce in this setting, a recent multicenter study reported an OR rate of 42% and median PFS of 19 months, however, without central pathology and imaging review of cases (Carlsen et al. 2019). The ongoing NETTER-2 trial will further explore the role of PRRT in G3 NETs in a prospective randomized setting (NCT03972488). Finally, although data on everolimus are scant in G3 NETs, its role appears limited with a median PFS of 6 months in one study of 15 patients (Panzuto et al. 2017).
These considerations highlight the need for designing clinical trials specifically for patients with G3 NETs to overcome the bias from retrospective data. In addition to treatment efficacy and sequences, future studies should explore molecular biomarkers predictive of treatment efficacy in G3 NETs, in order to move toward personalized management. In our study, the level of Ki-67 index did not seem to impact the outcomes of treatments within all patients with G3 NETs. Low expression of the O6-methylguanine-methyltransferase, or methylation of its promoter, seems associated with increased efficacy of alkylating agents in digestive NETs (de Mestier et al. 2020), although contradictory results exist in the literature (Cives et al. 2016). Its theranostic role is currently under evaluation in the MGMT-NET trial, which includes patients with G3 NETs (NCT03217097), and in the ECOG E2211 clinical trial (NCT01824875). Besides, loss of Rb expression correlates with increased effectiveness of etoposide-platinum CT in NECs and might be helpful in highly proliferating G3 NETs, notably in case of other treatments failure (Hijioka et al. 2017, Lacombe et al. 2021).
This study is limited by its retrospective design. In particular, the comparison of different treatments was limited by a selection bias, which we attempted to take into account by analyzing the factors associated with each treatment type, and by adjusting analyses on the main prognostic factors. In addition, due to a small number of patients for each treatment group, statistical power for subgroups analyses was limited. Although all radiological data were reviewed locally without a central radiology review, all participating centers have recognized expertise in the management of NETs, including experienced radiologists. Conversely, besides being one of the largest series of advanced G3 NETs reported so far, one of the main strengths of this study is the systematic centralized pathological review of cases, which is especially important since G3 NEN are known to be classification-challenging (Tang et al. 2016; Elvebakken et al. 2020). In our study, 7/81 cases classified as G3 NETs were finally reclassified as G2 NETs (n = 3) or NECs (n = 4), highlighting the importance of centralized, expert pathological review (Merola et al. 2021).
In conclusion, in this large retrospective multicenter study, adenocarcinoma-like and alkylating-based CT were found to be the most effective treatments for patients with advanced G3 NETs. Further biomarker-based prospective studies comparing these different treatments are needed to determine the best treatment strategy in this setting.
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/ERC-21-0109.
Declaration of interest
L de Mestier: AAA, Ipsen, Keocyt, Novartis, Pfizer, Sirtex (consulting); Member of the Knowledge Network and NET-CONNECT Initiatives funded by Ipsen. A Lamarca: travel and educational support from Ipsen, Pfizer, Bayer, AAA, SirtEx, Novartis, Mylan and Delcath; Speaker honoraria from Merck, Pfizer, Ipsen, Incyte, AAA and QED; Advisory honoraria from EISAI, Nutricia Ipsen, QED and Roche; Member of the Knowledge Network and NET-CONNECT Initiatives funded by Ipsen. J Hernando: speaker honoraria from Eisai, Novartis, AAA, Angelini, Roche; Consultant or Advisory Role: IPSEN, AAA, Roche, Pfizer, Sanofi, Bayer, Janssen, Eisai, Astellas, BMS, MSD, Lilly; Member of the NET-CONNECT Initiative funded by Ipsen. W Zandee: member of the NET-CONNECT Initiative funded by Ipsen. T Alonso Gordoa: member of the NET-CONNECT Initiative funded by Ipsen. A Walenkamp: institutional financial support for advisory role from Polyphor, IPSEN, Karyopharm, and Novartis; Unrestricted research grants from IPSEN and Novartis; Study budgets from Abbvie, BMS, Genzyme, Karyopharm Therapeutics, Roche. M Ronot: member of the Knowledge Network initiative funded by Ipsen. T Brabander: speaker fees from AAA/Novartis and Ipsen; Advisory board: AAA/Novartis. G Cadiot: AAA, Ipsen, Keocyt, Novartis (consulting). J Capdevila: speaker and consultancy fees from Ipsen, Novartis, AAA, Pfizer, Merck, Lilly, Exelixis, Bayer, Eisai, Sanofi; Member of the NET-CONNECT Initiative funded by Ipsen. M Pavel: member of the NET-CONNECT Initiative funded by Ipsen; Speaker and consultancy fees from IPSEN, Novartis, AAA, Pfizer, Boehringer Ingelheim, Riemser. J Cros: Member of the NET-CONNECT Initiative funded by Ipsen.
Funding
NET CONNECT is supported by an Independent Educational Grant from IPSEN. The program is, therefore, independent, and the content is not influenced by IPSEN and is under the sole responsibility of the authors. Dr Louis de Mestier was part-funded by the Société Nationale Française de Gastro-Entérologie. Dr Angela Lamarca was part-funded by The Christie Charity.
Acknowledgements
The authors thank NET CONNECT for having made this study possible through logistical and organizational support provided by COR2ED.
References
Apostolidis L, Dal Buono A, Merola E, Jann H, Jaeger D, Wiedenmann B, Winkler EC & Pavel M 2020 Multicenter analysis of treatment outcomes for well differentiated grade 3 neuroendocrine tumors (NET G3). Journal of Clinical Oncology 38 4607. (https://doi.org/10.1200/JCO.2020.38.15_suppl.4607)
Basturk O, Yang Z, Tang LH, Hruban RH, Adsay V, McCall CM, Krasinskas AM, Jang KT, Frankel WL, Balci S, et al. 2015 The high-grade (WHO G3) pancreatic neuroendocrine tumor category is morphologically and biologically heterogenous and includes both well differentiated and poorly differentiated neoplasms. American Journal of Surgical Pathology 39 683–690. (https://doi.org/10.1097/PAS.0000000000000408)
Carlsen EA, Fazio N, Granberg D, Grozinsky-Glasberg S, Ahmadzadehfar H, Grana CM, Zandee WT, Cwikla J, Walter MA, Oturai PS, et al. 2019 Peptide receptor radionuclide therapy in gastroenteropancreatic NEN G3: a multicenter cohort study. Endocrine-Related Cancer 26 227–239. (https://doi.org/10.1530/ERC-18-0424)
Cives M, Ghayouri M, Morse B, Brelsford M, Black M, Rizzo A, Meeker A & Strosberg J 2016 Analysis of potential response predictors to capecitabine/temozolomide in metastatic pancreatic neuroendocrine tumors. Endocrine-Related Cancer 23 759–767. (https://doi.org/10.1530/ERC-16-0147)
Collot T, Fumet JD, Klopfenstein Q, Vincent J, Bengrine L & Ghiringhelli F 2018 Bevacizumab-based chemotherapy for poorly-differentiated neuroendocrine tumors. Anticancer Research 38 5963–5968. (https://doi.org/10.21873/anticanres.12943)
Coriat R, Walter T, Terris B, Couvelard A & Ruszniewski P 2016 Gastroenteropancreatic well-differentiated Grade 3 neuroendocrine tumors: review and position statement. Oncologist 21 1191–1199. (https://doi.org/10.1634/theoncologist.2015-0476)
de Mestier L, Dromain C, d’Assignies G, Scoazec JY, Lassau N, Lebtahi R, Brixi H, Mitry E, Guimbaud R, Courbon F, et al.2014 Evaluating digestive neuroendocrine tumor progression and therapeutic responses in the era of targeted therapies: state of the art. Endocrine-Related Cancer 21 R105–R120. (https://doi.org/10.1530/ERC-13-0365)
de Mestier L, Couvelard A, Blazevic A, Hentic O, de Herder WW, Rebours V, Paradis V, Ruszniewski P, Hofland LJ & Cros J 2020 Critical appraisal of MGMT in digestive NET treated with alkylating agents. Endocrine-Related Cancer 27 R391–R405. (https://doi.org/10.1530/ERC-20-0227)
Ducreux M, Dahan L, Smith D, O’Toole D, Lepère C, Dromain C, Vilgrain V, Baudin E, Lombard-Bohas C, Scoazec JY, et al. 2014 Bevacizumab combined with 5-FU/streptozocin in patients with progressive metastatic well-differentiated pancreatic endocrine tumours (BETTER trial) - a phase II non-randomised trial. European Journal of Cancer 50 3098–3106. (https://doi.org/10.1016/j.ejca.2014.10.002)
Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, Dancey J, Arbuck S, Gwyther S, Mooney M, et al.2009 New Response Evaluation Criteria in Solid Tumours: Revised RECIST Guideline (1.1 version). European Journal of Cancer 45 228–247. (https://doi.org/10.1016/j.ejca.2008.10.026)
Elvebakken H, Perren A, Scoazec JY, Tang LH, Federspiel B, Klimstra DS, Vestermark LW, Ali AS, Zlobec I, Myklebust TÅ, et al. 2020 A consensus developed morphological re-evaluation of 196 high-grade gastroenteropancreatic neuroendocrine neoplasms and its clinical correlations. Neuroendocrinology [epub]. (https://doi.org/10.1159/000511905)
Garcia-Carbonero R, Sorbye H, Baudin E, Raymond E, Wiedenmann B, Niederle B, Sedlackova E, Toumpanakis C, Anlauf M, Cwikla JB, et al. 2016 Enets consensus guidelines for high-grade gastroenteropancreatic neuroendocrine tumors and neuroendocrine carcinomas. Neuroendocrinology 103 186–194. (https://doi.org/10.1159/000443172)
Heetfeld M, Chougnet CN, Olsen IH, Rinke A, Borbath I, Crespo G, Barriuso J, Pavel M, O’Toole D, Walter T, et al. 2015 Characteristics and treatment of patients with G3 gastroenteropancreatic neuroendocrine neoplasms. Endocrine-Related Cancer 22 657–664. (https://doi.org/10.1530/ERC-15-0119)
Hijioka S, Hosoda W, Matsuo K, Ueno M, Furukawa M, Yoshitomi H, Kobayashi N, Ikeda M, Ito T, Nakamori S, et al.2017 Rb loss and KRAS mutation are predictors of the response to platinum-based chemotherapy in pancreatic neuroendocrine neoplasm with Grade 3: A Japanese multicenter pancreatic NEN-G3 study. Clinical Cancer Research 23 4625–4632. (https://doi.org/10.1158/1078-0432.CCR-16-3135)
Klöppel G, Couvelard A, Hruban RH, Klimstra DS, Komminoth P, Osamura RY, Perren A & Rindi G 2017 Neoplasms of the neuroendocrine pancreas. In WHO Classification of Tumours of the Endocrine Organs, 4th ed. pp. 210–239. Lyon: IARC Press.
Konukiewitz B, Schlitter AM, Jesinghaus M, Pfister D, Steiger K, Segler A, Agaimy A, Sipos B, Zamboni G, Weichert W, et al.2017 Somatostatin receptor expression related to TP53 and RB1 alterations in pancreatic and extrapancreatic neuroendocrine neoplasms with a Ki67-index above 20%. Modern Pathology 30 587–598. (https://doi.org/10.1038/modpathol.2016.217)
Lacombe C, De Rycke O, Couvelard A, Turpin A, Cazes A, Hentic O, Gounant V, Zalcman G, Ruszniewski P, Cros J, et al.2021 Biomarkers of response to etoposide-platinum chemotherapy in patients with Grade 3 neuroendocrine neoplasms. Cancers 13 643. (https://doi.org/10.3390/cancers13040643)
Lamarca A, Crona J, Ronot M, Opalinska M, Lopez Lopez C, Pezzutti D, Najran P, Carvhalo L, Franca Bezerra RO, Borg P, et al.2019 aValue of tumor growth rate (tgr) as an early biomarker predictor of patients’ outcome in neuroendocrine tumors (net)-the GREPONET study. Oncologist 24 e1082–e1090. (https://doi.org/10.1634/theoncologist.2018-0672)
Lamarca A, Ronot M, Moalla S, Crona J, Opalinska M, Lopez Lopez C, Pezzutti D, Najran P, Carvhalo L, Bezerra ROF, et al. 2019 bTumor growth rate as a validated early radiological biomarker able to reflect treatment-induced changes in neuroendocrine tumors: the GREPONET-2 study. Clinical Cancer Research 25 6692–6699. (https://doi.org/10.1158/1078-0432.CCR-19-0963)
Liu AJ, Ueberroth BE, McGarrah PW, Buckner Petty SA, Kendi AT, Starr J, Hobday TJ, Halfdanarson TR & Sonbol MB 2021 Treatment outcomes of well-differentiated high-grade neuroendocrine tumors. Oncologist 26 383–388. (https://doi.org/10.1002/onco.13686)
McGarrah PW, Hobday TJ, Starr JS, Kendi AT, Graham RP, Sonbol MB & Halfdanarson TR 2020 Efficacy of somatostatin analog (SSA) monotherapy for well-differentiated grade 3 (G3) gastroenteropancreatic neuroendocrine tumors (NETs). Journal of Clinical Oncology 38 617–617. (https://doi.org/10.1200/JCO.2020.38.4_suppl.617)
Merola E, Alonso Gordoa T, Zhang P, Al-Toubah T, Pellè E, Kolasińska-Ćwikła A, Zandee W, Laskaratos F, de Mestier L & Lamarca A et al. 2020 Somatostatin analogs for pancreatic neuroendocrine tumors: any benefit when Ki-67 is ≥10%? Oncologist 26 294–301 (https://doi.org/10.1002/onco.13633)
Merola E, Zandee W, de, Mestier L, Klümpen HJ, Makulik K, Geboes K, van Velthuysen ML, Couvelard A, Cros J, van Eeden S, et al.2021 Histopathological revision for gastroenteropancreatic neuroendocrine neoplasms in expert centers: does it make the difference? Neuroendocrinology 111 170–177. (https://doi.org/10.1159/000507082)
Milione M, Maisonneuve P, Spada F, Pellegrinelli A, Spaggiari P, Albarello L, Pisa E, Barberis M, Vanoli A, Buzzoni R, et al. 2017 The clinicopathologic heterogeneity of Grade 3 gastroenteropancreatic neuroendocrine neoplasms: morphological differentiation and proliferation identify different prognostic categories. Neuroendocrinology 104 85–93. (https://doi.org/10.1159/000445165)
Mizuno Y, Kudo A, Akashi T, Akahoshi K, Ogura T, Ogawa K, Ono H, Mitsunori Y, Ban D, Tanaka S, et al. 2018 Sunitinib shrinks NET-G3 pancreatic neuroendocrine neoplasms. Journal of Cancer Research and Clinical Oncology 144 1155–1163. (https://doi.org/10.1007/s00432-018-2636-2)
Nuñez-Valdovinos B, Carmona-Bayonas A, Jimenez-Fonseca P, Capdevila J, Castaño-Pascual Á, Benavent M, Pi Barrio JJ, Teule A, Alonso V, Custodio A, et al.2018 Neuroendocrine tumor heterogeneity adds uncertainty to the World Health Organization 2010 classification: real-world data from the Spanish tumor registry (R-GETNE). Oncologist 23 422–432. (https://doi.org/10.1634/theoncologist.2017-0364)
Panzuto F, Rinzivillo M, Spada F, Antonuzzo L, Ibrahim T, Campana D, Fazio N & Delle Fave G 2017 Everolimus in pancreatic neuroendocrine carcinomas G3. Pancreas 46 302–305. (https://doi.org/10.1097/MPA.0000000000000762)
Pavel M, O’Toole D, Costa F, Capdevila J, Gross D, Kianmanesh R, Krenning E, Knigge U, Salazar R, Pape UF, et al.2016 Enets consensus guidelines update for the management of distant metastatic disease of intestinal, pancreatic, bronchial neuroendocrine neoplasms (NEN) and NEN of unknown primary site. Neuroendocrinology 103 172–185. (https://doi.org/10.1159/000443167)
Pellat A, Dreyer C, Couffignal C, Walter T, Lombard-Bohas C, Niccoli P, Seitz JF, Hentic O, André T, Coriat R, et al.2018 Clinical and biomarker evaluations of sunitinib in patients with Grade 3 digestive neuroendocrine neoplasms. Neuroendocrinology 107 24–31. (https://doi.org/10.1159/000487237)
Raj N, Valentino E, Capanu M, Tang LH, Basturk O, Untch BR, Allen PJ, Klimstra DS & Reidy-Lagunes D 2017 Treatment response and outcomes of Grade 3 pancreatic neuroendocrine neoplasms based on morphology: well differentiated versus poorly differentiated. Pancreas 46 296–301. (https://doi.org/10.1097/MPA.0000000000000735)
Rindi G, Klimstra DS, Arnold R, Klöppel G, Bosman FT, Komminoth P & Perren A 2010 Nomenclature and classification of neuroendocrine neoplasms of the digestive system. In WHO Classification of Tumours of the Digestive System, 4th ed. pp. 13–14. Lyon: IARC Press.
Rogowski W, Wachuła E, Gorzelak A, Lebiedzińska A, Sulżyc-Bielicka V, Iżycka-Świeszewska E, Żołnierek J & Kos-Kudła B 2019 Capecitabine and temozolomide combination for treatment of high-grade, well-differentiated neuroendocrine tumour and poorly-differentiated neuroendocrine carcinoma - retrospective analysis. Endokrynologia Polska 70 313–317. (https://doi.org/10.5603/EP.a2019.0010)
Sahu A, Jefford M, Lai-Kwon J, Thai A, Hicks RJ & Michael M 2019 CAPTEM in metastatic well-differentiated intermediate to high grade neuroendocrine tumors: a single centre experience. Journal of Oncology 2019 9032753. (https://doi.org/10.1155/2019/9032753)
Sorbye H, Welin S, Langer SW, Vestermark LW, Holt N, Osterlund P, Dueland S, Hofsli E, Guren MG, Ohrling K, et al.2013 Predictive and prognostic factors for treatment and survival in 305 patients with advanced gastrointestinal neuroendocrine carcinoma (WHO G3): the Nordic NEC study. Annals of Oncology 24 152–160. (https://doi.org/10.1093/annonc/mds276)
Stelwagen J, de Vries EGE & Walenkamp AME 2021 Current treatment strategies and future directions for extrapulmonary neuroendocrine carcinomas: a review. JAMA Oncology 7 759–770. (https://doi.org/10.1001/jamaoncol.2020.8072)
Tang LH, Basturk O, Sue JJ & Klimstra DS 2016 A practical approach to the classification of WHO Grade 3 (G3) Well-differentiated Neuroendocrine Tumor (WD-NET) and Poorly Differentiated Neuroendocrine Carcinoma (PD-NEC) of the Pancreas. American Journal of Surgical Pathology 40 1192–1202. (https://doi.org/10.1097/PAS.0000000000000662)
Vélayoudom-Céphise FL, Duvillard P, Foucan L, Hadoux J, Chougnet CN, Leboulleux S, Malka D, Guigay J, Goere D, Debaere T, et al.2013 Are G3 Enets neuroendocrine neoplasms heterogeneous? Endocrine-Related Cancer 20 649–657. (https://doi.org/10.1530/ERC-13-0027)
WHO classification of tumours 2019. Digestive System Tumours. Lyon: IARC.
Zappa M, Hentic O, Vullierme MP, Lagadec M, Ronot M, Ruszniewski P & Vilgrain V 2017 Is visual radiological evaluation of liver tumour burden in patients with neuroendocrine tumours reproducible? Endocrine Connections 6 33–38. (https://doi.org/10.1530/EC-16-0092)
