Myeloid and T-Cell Microenvironment Immune Features Identify Two Prognostic Sub-Groups in High-Grade Gastroenteropancreatic Neuroendocrine Neoplasms

Abstract

High-grade Gastroenteropancreatic Neuroendocrine neoplasms (H-NENs) comprehend well-differentiated tumors (NET G3) and poorly differentiated carcinomas (NEC) with proliferative activity indexes as mitotic count (MC) >20 mitoses/10 HPF and Ki-67 >20%. At present, no specific therapy for H-NENs exists and the several evidences of microenvironment involvement in their pathogenesis pave the way for tailored therapies. Forty-five consecutive cases, with available information about T-cell, immune, and non-immune markers, from surgical pathology and clinical databases of 2 Italian institutions were immunostained for Arginase, CD33, CD163 and CD66 myeloid markers. The association between features was assessed by Spearman’s correlation coefficient. A unsupervised K-means algorithm was used to identify clusters of patients according to inputs of microenvironment features and the relationship between clusters and clinicopathological features, including cancer-specific survival (CSS), was analyzed. The H-NEN population was composed of 6 (13.3%) NET G3 and 39 (86.7%) NEC. Overall, significant positive associations were found between myeloid (CD33, CD163 and Arginase) and T/immune markers (CD3, CD4, CD8, PD-1 and HLA-I). Myeloid and T-cell markers CD3 and CD8 identified two clusters of patients from unsupervised K-means analysis. Cases grouped in cluster 1 with more myeloid infiltrates, T cell, HLA and expression of inhibitory receptors and ligands in the stroma (PD-1, PD-L1) had significantly better CSS than patients in cluster 2. Multivariable analysis showed that Ki-67 (>55 vs. <55, HR 8.60, CI 95% 2.61–28.33, p < 0.0001) and cluster (1 vs. 2, HR 0.43, CI 95% 0.20–0.93, p = 0.03) were significantly associated with survival. High grade gastroenteropancreatic neuroendocrine neoplasms can be further classified into two prognostic sub-populations of tumors driven by different tumor microenvironments and immune features able to generate the framework for evaluating new therapeutic strategies.

Biliary Neuroendocrine Neoplasms: Analysis of Prognostic Factors and Development and Validation of a Nomogram

“…Our study also has limitations. A major constraint is that our nomogram was created using just six clinicopathological factors, lacking other additional variables such as Ki-67 ( 33 , 34 ), and the high Ki-67 index is associated with portal venous tumor invasion which is a prognostic factor for patients with pancreatic NENs ( 35 ). Besides, lactate dehydrogenase (LDH) has been found to increase in patients with NEC ( 36 ).…”

Biomarker Landscape in Neuroendocrine Tumors With High-Grade Features: Current Knowledge and Future Perspective

“…Moreover, a subset of NECs has microenvironment features consistent with spontaneous activation of adaptive immunity (co-expression of CD3, CD4, CD8, PD-1, and PD-L1). Recently, we further evaluated the tumor microenvironment of high-grade NENs, by expanding the immune profiling to myeloid markers and identifying two prognostic subpopulations of tumors likely compatible with the "hot/cold tumor" idea: highgrade NENs characterized by a prevalent immune infiltrate cells had better survival (67). According to this, it was suggested that microenvironment-related immune and inflammatory markers can improve prognostic prediction in GEP-NENs when combined with the known prognostic factors, and they may predict potential responsiveness to immunotherapy of GEP NECs (66,67).…”

“…Recently, we further evaluated the tumor microenvironment of high-grade NENs, by expanding the immune profiling to myeloid markers and identifying two prognostic subpopulations of tumors likely compatible with the "hot/cold tumor" idea: highgrade NENs characterized by a prevalent immune infiltrate cells had better survival (67). According to this, it was suggested that microenvironment-related immune and inflammatory markers can improve prognostic prediction in GEP-NENs when combined with the known prognostic factors, and they may predict potential responsiveness to immunotherapy of GEP NECs (66,67). Furthermore, Bosch et al demonstrated that high TILs and PD-1 expression are significantly associated with shorter survival and higher grading in GEP NENs.…”

Elevated Flt3L Predicts Long-Term Survival in Patients with High-Grade Gastroenteropancreatic Neuroendocrine Neoplasms

“…A negative impact on the clinical course of panNET was also proposed for tumor-associated macrophages (TAM) based on their positive correlation with poor disease-specific and disease-free survival [13,14] in G1 and G2 tumors. Recent immunohistological profiling of tumor microenvironment and tumor inflammatory features pointed towards an even more prominent role of immune-related features in high grade NEN [15,16]. In a NEN cohort featuring 40% G3 specimens the authors reported a notable shift of immune tumor microenvironment (TME) marker profiles in G3 compared to G1 and G2 samples, which could be harnessed for improved stratification of the prognostic patient subgroups.…”

“…In a NEN cohort featuring 40% G3 specimens the authors reported a notable shift of immune tumor microenvironment (TME) marker profiles in G3 compared to G1 and G2 samples, which could be harnessed for improved stratification of the prognostic patient subgroups. In a follow-up study with an extended panel of myeloid and lymphoid markers, the marker-based clustering of samples separated further prognostic subgroups within G3 NEN, with better representation of lymphoid and myeloid markers in the favorable "hot" cluster when compared to the "cold" cluster [16]. Overall, a decisive contribution of adaptive and innate immune responses in more aggressive and advanced disease situations emerges as a common theme, despite the vastly different methodological approaches and NEN cohort compositions, providing a growing rationale for immunedirected therapies.…”