Natasha Rekhtman, Matthew G Hanna, Paul Hofman, Mari Mino-Kenudson, Philippe Joubert, Mauro Papotti, Lynette M Sholl, Hyo Sup Shim, Yuchen Han, Véronique Hofman, Yin P Hung, Marina K Baine, Rania G Aly, Deepali Jain, Teodora Radonic, Anjali Saqi, Mary Beth Beasley, Jin-Haeng Chung, Wendy Cooper, Fred R Hirsch, Keith Kerr, Sylvie Lantuejoul, Daisuke Matsubara, Noriko Motoi, Andrew G Nicholson, William D Travis, Jan von der Thüsen, Yasushi Yatabe, Anja C Roden
In this international multi-institutional study, we address a major unresolved issue for WHO classification. 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. The concurrent IASLC proposal to include Ki-67 as additional criterion for grade 3 carcinoids should further enhance this approach.
INTRODUCTION: Lung carcinoids are well-differentiated neuroendocrine tumors (NETs), designated in the current thoracic WHO classification as typical carcinoids (grade 1) and atypical carcinoids (grade 2), whereas rare tumors with carcinoid/NET morphology and >10 mitoses/2mm2 are designated as large cell neuroendocrine carcinoma (LCNEC). This approach, however, has been a matter of growing debate.
METHODS: Lung resections of tumors with carcinoid/NET morphology and >10 mitoses/2 mm2 were contributed by the members of IASLC Pathology Committee. Digitized slides and detailed clinicopathologic, immunohistochemical, and molecular data were collected and analyzed.
RESULTS: In total, 28 cases meeting the study criteria were collected from 9 institutions in 6 countries. Tumors exhibited mean mitotic counts of 15/2mm2 (range, 11-30/2mm2) and mean hot-spot Ki-67 rate of 32% (range 13-54%). In all cases, Ki-67 was heterogeneous, containing cold-spot zones with Ki-67 of <1%-15%. In all cases, Rb and p53 showed wild-type pattern by immunohistochemistry. Next-generation sequencing (n=9) revealed wild-type RB1 and TP53, low TMB (1.3-3.3 mutations/Mb), and genomic alterations characteristic of lung carcinoids. The study cohort exhibited significantly worse OS and RFS than control groups of 117 typical and 52 atypical carcinoids (P<0.0001 for all comparisons).
CONCLUSION: In this international multi-institutional study, we address a major unresolved issue for WHO classification. 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. The concurrent IASLC proposal to include Ki-67 as additional criterion for grade 3 carcinoids should further enhance this approach.