IASLC pathologists propose a grade 3 category for lung carcinoids with more than 10 mitoses

A category the WHO classification currently assigns to LCNEC

The current WHO thoracic classification grades well-differentiated lung NETs as typical carcinoids (grade 1) and atypical carcinoids (grade 2). A tumour that keeps carcinoid morphology but exceeds 10 mitoses per 2 mm² falls outside the atypical range, and under the present rules it is diagnosed as LCNEC, a high-grade neuroendocrine carcinoma. The authors describe this approach as “a matter of growing debate”.

The question has practical weight because LCNEC is usually grouped with the poorly differentiated neuroendocrine carcinomas, which differ from carcinoids in their molecular drivers. The tumours studied here are rare: the committee members assembled 28 resected cases across nine institutions, which the authors describe as the largest cohort of such tumours to date.

Wild-type RB1 and TP53 with a carcinoid-type genome

Immunohistochemistry showed a wild-type pattern for both Rb and p53 in all 28 tumours. Loss of Rb and aberrant p53 are typical of small cell carcinoma and of a large share of LCNEC, so their absence is consistent with a carcinoid.

Next-generation sequencing was available for 9 of the 28 cases. All 9 carried wild-type RB1 and TP53, a tumour mutational burden between 1.3 and 3.3 mutations/Mb, and genomic alterations the authors describe as characteristic of lung carcinoids. Sequencing in only a third of the cohort limits how far the genomic finding can be generalised to the whole group.

Proliferation varied widely within each tumour

Mitotic counts across the cohort averaged 15 per 2 mm², with a range of 11 to 30. The mean hot-spot Ki-67 rate was 32%, ranging from 13% to 54% between tumours.

Within every individual tumour, Ki-67 was heterogeneous. Each case contained cold-spot zones with a Ki-67 rate of under 1% to 15%. This suggests that a small biopsy taken from a cold-spot zone could underestimate the grade of the tumour, a point that bears on how these cases are recognised before resection.

Shorter survival than typical and atypical carcinoids

The 28 tumours had significantly worse overall survival and recurrence-free survival than control groups of 117 typical and 52 atypical carcinoids, with P<0.0001 for every comparison. The abstract reports comparisons with lower-grade carcinoids only, and gives no comparison with conventional LCNEC or small cell carcinoma.

The cases were contributed by committee members and were not collected consecutively, which may introduce selection bias. All cases were surgical resections, so the findings describe operable disease.

“By assembling the largest cohort to-date of the study cases, we propose that they should be classified as grade 3 carcinoids/NETs rather than LCNEC to enable clinical management more aligned with their biology.”Rekhtman et al., Journal of Thoracic Oncology (2026)

A companion proposal to add Ki-67 to the grading criteria

The authors link their proposal to a concurrent IASLC proposal to include Ki-67 as an additional criterion for grade 3 carcinoids. Ki-67 would give pathologists a second criterion alongside the mitotic count, which matters for tumours whose Ki-67 varies as widely as it did in this cohort.

The authors present the work as addressing a major unresolved issue for the WHO classification. In their view, patients whose tumours are currently reported as LCNEC on the basis of mitotic count alone could then receive management more closely matched to the biology of their disease.