Response to “Methodological and inferential limits before abandoning follow‐up in node‐negative typical lung carcinoids”

We thank Bertolaccini and colleagues for their interest in our recent study ¹ and for contributing to the important discussion regarding follow‐up strategies for patients with typical carcinoids.

Ideally, decisions regarding the implementation, modification, or discontinuation of surveillance or follow‐up programs should be informed by randomized controlled trials (RCTs). However, this is rarely feasible, particularly in low‐incidence diseases such as typical carcinoids. Current neuroendocrine tumor surveillance guidelines are largely based on retrospective studies and expert opinion, with limited prospective evidence available. This also applies to the current ENETS guidelines regarding follow‐up of typical carcinoids, for which the underlying evidence is primarily derived from retrospective studies with shorter follow‐up periods and smaller sample sizes than those included in the present study.

It is well recognized that established surveillance and follow‐up programs can be difficult or indeed impossible to modify or discontinue, even when research has failed to document significant benefits. ² Conversely, surveillance programs are often introduced based on a comparable level of evidence. We acknowledge the concerns raised by Bertolaccini and colleagues. At the same time, we emphasize that if RCT evidence is considered a prerequisite for modifying or discontinuing established follow‐up programs, meaningful change may become exceedingly difficult, if not impossible, to achieve.

In areas where evidence is necessarily more limited than that provided by RCTs, it is reasonable that clinicians and researchers may arrive at different interpretations and conclusions based on the available literature. Constructive discussion regarding the strengths and limitations of individual studies is therefore both expected and valuable. For these reasons, we believe that transparent and evidence‐based dialogue is essential, and we greatly appreciate the thoughtful and critical comments raised by Bertolaccini and colleagues. At the same time, we recognize the important role of ENETS and other stakeholders in encouraging the continued generation of stronger evidence to support clinical decision‐making.

We are aware that data comes from two highly specialized centers. Although the concerns raised may point toward a more favorable biology, we still believe that recurrence rates are low in general and that multi‐year CT scans should be considered very carefully. We acknowledge that healthcare systems and the degree of centralization of NEN care vary considerably across countries, and that a model with highly centralized management may not be directly representative of all settings. However, the clinical practice underlying the present study is based on and closely adheres to current international and European guideline recommendations for the diagnosis, surgical management, and follow‐up of patients with pulmonary carcinoid tumors. As such, while the organizational structure of care may differ, the diagnostic criteria, treatment principles, and follow‐up strategies applied at both our centers are aligned with those used internationally. Importantly, NET care in Denmark is highly centralized at three specialized centers that manage patients with pulmonary carcinoids irrespective of disease stage. This likely makes the cohort representative of the Danish pulmonary carcinoid population, with limited selection bias, thereby strengthening the generalizability of our findings to other populations regardless of how NET care is organized.

Differential nodal assessment may introduce a risk of misclassification of N status among patients undergoing wedge resection, as lymph node evaluation in this group was performed at the surgeon's discretion rather than through systematic dissection. If present, such misclassification would most likely result in occult node‐positive disease being classified as N0, thereby biasing the results toward an underestimation of the true difference in recurrence risk between groups. Consequently, this potential bias would not invalidate our conclusions; on the contrary, it would be expected to attenuate observed associations and thus, if anything, reinforce the robustness of our conclusions rather than weaken them. We fully agree that the definition of N0 disease is critically important, particularly when proposing that the distinction between N0 and N1 disease may be central to determining whether patients should be enrolled in follow‐up programs. In this context, ENETS also plays a crucial role in defining standards for lymph node sampling and assessment across different surgical procedures.

We acknowledge that uncertainty remains regarding the magnitude of the risk associated with radiation‐induced malignancy, just as uncertainty persists regarding recurrence risk and the extent to which patients benefit from the early detection of asymptomatic recurrence. These considerations underscore the complexity of surveillance decisions and further emphasize the need for additional studies with long‐term follow‐up and, where feasible, randomized designs to better inform these challenging clinical questions.

We acknowledge that we have not defined a clear clinical threshold for acceptable recurrence risk which is useful when surveillance is tailored based on prognostic factors and the evidence for exact risk of radiation‐induced malignancy in different age groups is lacking. As stated in the comment, the risk is lower in an elderly population but so is the risk of dying from recurrence of a slow‐growing disease. Selecting which patients should receive enhanced surveillance may require moving beyond risk stratification toward benefit stratification, that is, toward basing surveillance recommendations not only on the patient's risk of developing recurrence but also on the probability that surveillance enables earlier‐stage detection, receipt of effective treatment, and reduced mortality. However, the risk of recurrence was very low and the risk of having a recurrence that led to resection and thereby curative treatment even lower (1/290). Overall, given the very low number of recurrences and disease‐related deaths, we do believe that multi‐year CT scans should be considered very carefully in all centers and avoided in a centralized NEN setting.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

DATA AVAILABILITY STATEMENT

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.