Open AccessMicroenvironment and tumor inflammatory features improve prognostic prediction in gastro‐entero‐pancreatic neuroendocrine neoplasms
Author(s) -
Milione Massimo,
Miceli Rosalba,
Barretta Francesco,
Pellegrinelli Alessio,
Spaggiari Paola,
Tagliabue Giovanna,
Centonze Giovanni,
Paolino Cinzia,
Mangogna Alessandro,
Kankava Ketevani,
Pusceddu Sara,
Giacomelli Luca,
Corti Ambra,
Cotsoglou Christian,
Mazzaferro Vincenzo,
Sozzi Gabriella,
Braud Filippo,
Pruneri Giancarlo,
Anichini Andrea
Publication year - 2019
Publication title -
the journal of pathology: clinical research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.849
H-Index - 21
ISSN - 2056-4538
DOI - 10.1002/cjp2.135
Subject(s) - hazard ratio , medicine , immunohistochemistry , cd8 , stromal cell , nomogram , immune system , proportional hazards model , oncology , tumor microenvironment , pathology , gastroenterology , confidence interval , immunology
Microenvironment‐related immune and inflammatory markers, when combined with established Ki‐67 and morphology parameters, can improve prognostic prediction in gastro‐entero‐pancreatic neuroendocrine neoplasms (GEP‐NENs). Therefore, we evaluated the prognostic value of microenvironment and tumor inflammatory features (MoTIFs) in GEP‐NENs. For this purpose, formalin‐fixed paraffin‐embedded tissue sections from 350 patients were profiled by immunohistochemistry for immune, inflammatory, angiogenesis, proliferation, NEN‐, and fibroblast‐related markers. A total of 314 patients were used to generate overall survival (OS) and disease‐free survival (DFS) MoTIFs prognostic indices (PIs). PIs and additional variables were assessed using Cox models to generate nomograms for predicting 5‐year OS and DFS. A total of 36 patients were used for external validation of PIs and nomograms' prognostic segregations. From our analysis, G1/G2 versus G3 GEP‐NENs showed phenotypic divergence with immune‐inflammatory markers. HLA, CD3, CD8, and PD‐1/PD‐L1 IHC expression separated G3 into two sub‐categories with high versus low adaptive immunity‐related features. MoTIFs PI for OS based on COX‐2 Tumor(T) > 4, PD‐1 Stromal(S) > 0, CD8 S < 1, and HLA‐I S < 1 was associated with worst survival (hazard ratio [HR] 2.50; 95% confidence interval [CI], 2.12–2.96; p < 0.0001). MoTIFs PI for DFS was based on COX‐2 T > 4, PD‐1 S > 4, HLA‐I S < 1, HLA‐I T < 2, HLA‐DR S < 6 (HR 1.77; 95% CI, 1.58–1.99; p < 0.0001). Two nomograms were developed including morphology (HR 4.83; 95% CI, 2.30–10.15; p < 0.001) and Ki‐67 (HR 11.32; 95% CI, 5.28–24.24; p < 0.001) for OS, and morphology (PI = 0: HR 10.23; 95% CI, 5.67–18.47; PI = 5: HR 2.87; 95% CI, 1.21–6.81; p < 0.001) and MoTIFs PI for DFS in well‐differentiated GEP‐NENs (HR 6.21; 95% CI, 2.52–13.31; p < 0.001). We conclude that G1/G2 to G3 transition is associated with immune‐inflammatory profile changes; in fact, MoTIFs combined with morphology and Ki‐67 improve 5‐year DFS prediction in GEP‐NENs. The immune context of a subset of G3 poorly differentiated tumors is consistent with activation of adaptive immunity, suggesting a potential for responsiveness to immunotherapy targeting immune checkpoints.