Treatment sequencing in gastroenteropancreatic neuroendocrine tumors

In the management of advanced or metastatic NET, treatment sequencing is one of the major unresolved questions. This malignant disease stands out for its heterogeneity in terms of primary tumor location, presentation, clinical course, and prognosis. To some extent, this is reflected in the fact that ENETS has recently published updated guidance papers separately for different tumor sites and hormonal syndromes observed in NET [1‐4]. These guidelines provide a detailed update on diagnosis and multidisciplinary treatment, whereas the latest European Society for Medical Oncology (ESMO) guidelines for gepNET, last updated in 2020, are more narrowly focused on systemic treatment [5]. The majority of gepNET are slowly progressing tumors, and patients are likely to receive multiple lines of therapy over their disease course.

In contemporary treatment algorithms, primary tumor site, tumor grading, expression of somatostatin receptors (SSTR), progression status, and tumor load are the main clinical decision factors [2, 4‐6]. The grading system for gepNET established by the World Health Organization (WHO) is based on Ki-67 index and mitotic count and distinguishes NET G1 (Ki-67 < 3% and/or < 2 mitoses/2 mm²), NET G2 (Ki-67 3–20% and/or 2–20 mitoses/2 mm²), and NET G3 (Ki-67 > 20% and/or > 20 mitoses/2 mm²) [7]. A higher grade indicates a higher biological aggressiveness and is associated with decreased SSTR expression and worse survival. [8, 9] In contrast, neuroendocrine carcinomas (NEC) are poorly differentiated and have an even grimmer prognosis [5]. About 20% of intestinal NET patients and 10–30% of pancreatic NET patients present with hormonal syndromes, and symptom control is primarily achieved through somatostatin analogs (SSA), with other therapies also playing a role [5]. In addition, locoregional treatment options such as surgery in a curative setting are worth mentioning but are not further discussed here [5]. Finally, tumor burden is an important prognostic factor and may influence the treatment decision in favor of a cytoreductive versus a disease-stabilizing treatment approach.

Current guidelines consider somatostatin analogs (SSA) the main first-line option for slowly growing metastatic enteropancreatic NET (with a Ki-67 index up to 10%) [2, 4, 5]. For metastatic midgut NET (jejunoileal and right colon), the phase III trial PROMID (n = 85) established the disease-stabilizing effects of octreotide long-acting release (octreotide LAR; median progression-free survival [PFS] of 14.3 versus [vs] 6 months with placebo), but it did not show tumor size reduction (overall response rate [ORR] of 2% in both arms, consistent with the mode of action) or a benefit in overall survival (median OS of 84.7 vs 83.7 months, potentially impacted by crossover) [10, 11]; an overview of clinical trials on gepNET is shown in Table 1 and Fig. 1. CLARINET was a larger study (n = 204) that allocated enteropancreatic NET patients without hormonal symptoms to lanreotide or placebo and demonstrated a prolongation of PFS (not reached [NR] vs 18.0 months, hazard ratio [HR] 0.47, 95% confidence interval [CI] 0.30–0.73) but again no significant difference in OS [12]. Both studies required positive SSTR expression, and CLARINET limited enrollment to NET G1/G2 patients with a Ki-67 index < 10%. Resultantly, guidelines recommend other options such as peptide receptor radionuclide therapy, chemotherapy, or targeted agents for the front-line treatment of more rapidly progressing gepNET (Ki-67 index > 10%) [4, 5]. Notwithstanding this, SSA could potentially be considered upfront in individual NET G2 cases with a Ki-67 index > 10% [13]. Interestingly, CLARINET FORTE (phase II study) showed a median PFS of 8.3 months (midgut NET, n = 51) and 5.6 months (pancreatic NET, n = 48) for lanreotide every 14 days (reduced dosing interval) after progression on standard-regimen lanreotide (every 28 days) [14], which is nearly identical to the 8.4 months seen in the control arm of NETTER‑1 (high-dose octreotide after progression on octreotide), but also to the 8.5 months observed in the control arm of NETTER‑2 (high-dose octreotide in NET patients with a Ki-67 index of 10–55%) [15, 16]. However, the added benefit of retreatment with above-label SSA is also currently unclear. There are advances in drug development that are currently in clinical evaluation in NET, e.g., octreotide subcutaneous depot (SORENTO, NCT05050942) and the oral SSTR2 agonist paltusotine (NCT05361668).

Peptide receptor radionuclide therapy (PRRT) is a SSTR-targeted nuclear medicine approach employing a radioactive compound such as lutetium-177-DOTATATE to destroy tumor cells [16]. In the 2020 ESMO guidelines, PRRT is positioned in later therapy lines for progressive SSTR-positive NET (after SSA in midgut and after approved treatments in pancreatic NET) [5]. This sequencing can be partly attributed to the phase III trial NETTER‑1, in which 229 midgut NET patients, all of whom were progressing on standard-dose octreotide, were randomized to PRRT or high-dose octreotide [16]. NETTER‑1 found a substantial reduction in the risk of progression or death with PRRT (HR 0.21, 95% CI 0.13–0.33; median PFS NR vs 8.4 months in the control arm) [16]. This study clearly established PRRT as a highly effective treatment for this patient group and resulted in its regulatory approval. [5] With the conceptual successor, the recently published (2024) phase III study NETTER‑2, there are now also favorable data for PRRT (median PFS of 22.8 vs 8.5 months with high-dose octreotide, ORR of 43% vs 9.3%) as a first-line treatment for gepNET G2/G3 with a Ki-67 index of 10–55% [15]. While these data are somewhat difficult to interpret due to the non-approved control group with high-dose SSA as first-line therapy, PRRT is now endorsed by the ENETS guidance paper as a potential first-line option for small intestinal NET (siNET) G2 with fast growth or high tumor burden (and for selected siNET G3) [4]. For pancreatic NET, the phase II OCLURANDOM trial (2022 abstract) reported data on 84 patients randomly assigned to PRRT or sunitinib (functioned as an internal control of this study), showing a median PFS of 20.7 months for PRRT and of 11 months for sunitinib [17]. Generally, PRRT is well tolerated, however, long-term hematological sequelae, i.e., myelodysplastic syndrome and leukemia, occur in about 3–4% of patients [5]. PRRT rechallenge is recommended for selected siNET or panNET patients by the ENETS guidance papers, but prospective evidence is lacking [2, 4]. Two phase II studies, ReLUTH (NCT04954820) and NET RETREAT (NCT05773274), are currently evaluating PRRT retreatment in siNET. Since SSA and PRRT generally require SSTR expression (the biological rationale), different treatment approaches are needed in SSTR-negative patients.

Everolimus, a mammalian target of rapamycin (mTOR) inhibitor, is a commonly used treatment in the first line and endorsed by current guidelines in the case of more rapid growth or SSTR negativity [5]. RADIANT‑3 (pancreatic NET) and RADIANT‑4 (nonfunctioning lung and gastrointestinal NET) were two phase III trials that compared everolimus to placebo and showed a benefit for everolimus treatment in terms of PFS (11.0 vs 4.6 months and 11.0 vs 3.9 months, respectively) [18, 19]. Notably, tolerability of everolimus can be poor (e.g., grade 3–4 stomatitis in 9%, grade 3–4 diarrhea and infections in 7% each, and grade 3–4 anemia, fatigue, and hyperglycemia in 4% each), and toxicity-related treatment discontinuation was observed in 12–13% [18, 19]. The value of everolimus in NET G3 is unclear [20]. According to the ESMO guidelines, everolimus is a third-line option in siNET (after SSA and PRRT) but is positioned before PRRT in panNET [5]. A recent (2024) retrospective study, SeqEveRIV, compared everolimus and PRRT (and the two respective treatment sequences), identifying PRRT as potentially more effective and better tolerable than everolimus in different NET primary localizations [21]. This needs to be confirmed in a prospective study. The phase III study COMPOSE (NCT04919226) is currently investigating PRRT in a high G2 and G3 gepNET collective in the first-/second-line setting (177Lu-DOTATOC vs everolimus or chemotherapy with folinic acid plus 5‑fluorouracil plus oxaliplatin [FOLFOX] or with capecitabine and temozolomide [CAPTEM]). Likewise, COMPETE (NCT03049189) is comparing 177Lu-DOTATOC with everolimus in progressive gepNET.

A first-line consideration for fast-growing panNET or NET G3 is chemotherapy, while the activity in siNET is generally low [4, 5, 22]. The phase II study ECOG-ACRIN E2211 randomized 144 panNET G1/G2 patients to temozolomide or capecitabine and temozolomide (CAPTEM), showing extended PFS with the combination treatment (median PFS of 22.7 vs 11.4 months, p = 0.022) [23]. A phase III study in panNET called SEQTOR compared streptozotocin/5-fluorouracil with everolimus, each followed by the other treatment, and found similar efficacy (23.8 vs 21.5 months) but superior activity with chemotherapy (ORR of 30% vs 11%) [24]. The positioning of chemotherapy in relation to everolimus or sunitinib is currently unclear in panNET, and ESMO listed all of those as potential options [5]. The 2023 ENETS panNET treatment algorithm proposed chemotherapy as a first-line treatment in case of symptomatic or fast-growing NET, followed by targeted therapies, with reverse sequencing for slow-growing disease (in case of SSTR-negativity) [2]. Regarding siNET, the chemotherapy regimens CAPTEM and FOLFOX can be endorsed only for highly selected patients [4]. Regarding current research, the phase II study MeTe (planned n = 46) is evaluating the use of metronomic temozolomide in frail patients (NCT05554003).

The registration trial of sunitinib, a multitargeted tyrosine kinase inhibitor (TKI) with antiangiogenic activity, showed an improvement in PFS (11.4 vs 5.5 months, HR 0.42, 95% CI 0.26–0.66), OS (HR 0.41, 95% CI 0.19–0.89), and ORR (9.3% vs 0%) in progressive panNET compared to placebo [25]—siNET were not included, so there is no indication for these patients. CABINET, a more recent placebo-controlled phase III trial of cabozantinib (published 2024), was positive regarding the primary endpoint PFS (8.4 vs 3.9 months with placebo in extrapancreatic and 13.8 vs 4.4 months in pancreatic NET) [26]. Cabozantinib could also be an effective option for the treatment of NET G3 and low-proliferative NEC (Ki-67 index 20–60%), as indicated by the phase II CABONEN trial (disease control rate after 6 months of 64.1% in the interim analysis) [27]. The addition of the immune checkpoint inhibitor atezolizumab to cabozantinib showed an ORR of 16.7% in gepNET in the phase II study CABATEN, but the added benefit of this anti-PD-L1 antibody is unclear [28]—overall, there is currently no relevant evidence for the efficacy of immunotherapy in NET [2, 4]. In the phase II study TALENT (2021), lenvatinib was associated with an ORR of 29.9% and a median PFS of 15.7 months in gepNET, which is promising (particularly the high activity) and deserves further study [29]. Moreover, the TKI surufatinib was tested in two phase III trials in a Chinese population and is approved for NET in China [6]. The updated ENETS guidance paper on siNET already includes novel TKI in the treatment algorithm as a potential option after failure of standard treatments [4]. A potential future basis for treatment choice might be the safety profile of these drugs. Use of everolimus might be less favored in case of pre-existing lung disease (adverse event pneumonitis of any grade in 16–17%) or diabetes (hyperglycemia in 10–13%) [18, 19], whereas sunitinib and cabozantinib seem less favorable in case of hypertension (hypertension grade 3+ in 10% and 13–22%, respectively), cardiovascular disease (single cases of heart failure observed with both), or bleeding [25, 26, 30].

This overview of the clinical trial landscape in neuroendocrine tumors (NET) illustrates the steady progress that is being made in the treatment of advanced or metastatic gastroenteropancreatic NET (gepNET). Overall, the effects of state-of-the-art NET therapies are the suppression of tumor growth (somatostatin analogs [SSA], everolimus, and sunitinib) or even a temporary reduction in tumor volume (peptide receptor radionuclide therapy [PRRT] and chemotherapy). Importantly, the functional activity of NET also plays a role in the management of this disease and can influence treatment decisions. While current clinical guidelines provide a clear treatment sequence for certain scenarios, they can only offer a range of treatment options to choose from in certain other situations.