We thank Helge Waldum for his correspondence on our Primer (Lamberti, G. et al. Gastric neuroendocrine neoplasms. Nat. Rev. Dis. Primers 10, 25 (2024))1, which draws attention to the role of hypergastrinaemia in gastric tumorigenesis (Waldum, H. The need for a better classification system for gastric neoplasms. Nat. Rev. Dis. Primers https://doi.org/10.1038/s41572-024-00590-2 (2025))2. The Primer focuses on gastric neuroendocrine neoplasms (gNENs), thus gastric carcinoma, including a suggested role of hypergastrinaemia in adenocarcinoma carcinogenesis, was not discussed. Available data have clearly demonstrated that hypergastrinaemia, secondary to drug-induced hypochlorhydria and/or achlorhydria (including by proton-pump inhibitors (PPIs)), led to hyperplasia of enterochromaffin-like cells and, eventually, gastric neuroendocrine tumours (gNETs) in mice3, rather than gastric adenocarcinoma transformation. This is in keeping with the worldwide decrease in gastric adenocarcinoma incidence despite the widespread use of PPIs4, which are, therefore, not listed as gastric carcinogens by the International Agency for Research on Cancer (IARC)5.

With respect to gastric neuroendocrine tumorigenesis, hypergastrinaemia alone cannot induce tumour transformation, as it serves solely as a promoter (inducing proliferation) and lacks initiator properties (inducing transformation). Considering the differing incidence of gNETs in patients with chronic atrophic gastritis (0.4–0.7% per year)6,7 and long-term users of PPIs (isolated cases)8, a simultaneous gastric-specific, yet unknown, initiator is needed for gNET development, as the result of a two-hit strike (for example, long-term chronic inflammation and atrophy for type I gNETs).

In addition, in the presence of ectopic paraneoplastic gastrin production by gastrinoma, gNETs can be observed in patients with multiple endocrine neoplasia type I (MEN1) syndrome (type II gNETs), whereas gNETs occur very rarely in patients with sporadic gastrinoma. This difference is due to the loss of oncosuppressor activity in patients with MEN1 in contrast to the later onset and more aggressive behaviour of sporadic gastrinoma, which results in shorter exposure to hypergastrinaemia9. Furthermore, localized type II gNETs resolve rapidly if hypergastrinaemia is blocked by radical excision or somatostatin analogue treatment10. This suggests that, in vivo, hypergastrinaemia and MEN1 are both needed, but neither are sufficient, to drive type II gNET tumorigenesis.

Taken together, we respectfully disagree that we underestimated the role of gastrin in gNET pathogenesis. In the Primer, gNETs were grouped into ‘gastrin-dependent’ and ‘gastrin-independent’ gNETs, according to the role of hypergastrinaemia in the development of different gNET types1, as defined by the WHO clinical classification11, which is the only approved classification11. We did not include PPI-induced gNETs (sometimes termed type IV gNETs), as evidence of a causative relationship is lacking and they are excluded from the current WHO classification11.

Neuroendocrine carcinoma (NEC) is described as a separate entity, as the latest insights into its biology suggest separate origins for NETs and NECs, rather than a malignant progression from well-differentiated to poorly differentiated morphology in the absence of a pre-existing mixed morphology12.

Overall, the classification system of gNENs presented in the Primer considers known biological features (such as the context in which gNETs develop and genomic differences between gNETs and gastric NEC) and unknown factors, which may contribute to the different clinical behaviour of different gNEN types. This classification may change when new reliable evidence becomes available.