Neuroendocrine tumors : the quality of prognostic and predictive markers

Neuroendocrine neoplasms (NENs), rare tumors derived from disseminated neuroendocrine (NE) cells, can develop in different organs. NENs are sometimes diagnosed from metastatic tissue. The primary tumor location and grade affect both prognosis and the management of the disease. Histomorphologically, NENs show similarities independent of tumor origin. Immunohistochemistry (IHC) plays a key role in NEN diagnostics. Chromogranin A and synaptophysin expression confirms the NE nature of tumor cells, while tissue-specific IHC markers may indicate the primary tumor location. The proliferation marker Ki-67 determines the tumor grade of an NEN. Careful verification and validation of diagnostic, prognostic, and predictive markers, as well as reliable and reproducible staining assessment are necessary in NEN diagnostics. This thesis primarily aimed to identify new IHC markers for NENs and evaluate scoring methods for Ki-67 stainings using digital pathology and image analysis. The studies relied on a NE tumor (NET) cohort consisting of 91 NENs from 12 different primary sites, including primary–metastasis pairs. Tissue microarray (TMA) blocks were constructed from archived tumor specimens. First, the specificity of the primary antibodies studied was confirmed through technical validation, after which each marker’s expression was evaluated. The Ki-67 proliferation index determining the NET grade was examined using the image analysis software ImmunoRatio and conventional eyeballing. All NENs expressed neuropeptide S receptor 1 (NPSR1) and the neuropeptide S (NPS) ligand except for pheochromocytomas (PHEOs). The expression of proprotein convertase subtilisin/kexin type 2 (PCSK2) was identified in midgut NENs, PHEOs, paragangliomas, and pulmonary carcinoids. NENs from other primary locations were PCSK2-negative. The staining profiles of NPSR1, NPS, and PCSK2 were similar in primary tumors compared with metastatic samples. Validation of somatostatin receptor (SSTR) primary antibody clones revealed that only a few localized the receptor to the correct cell components. UMB clones exhibited the best cell membrane positivity. Different NENs showed varying SSTR 1-5 profiles depending on the primary location and tumor grade. In addition, assessment of the Ki-67 index using eyeballing revealed more variation than computer-assisted scoring. Thus, grading using image analysis proved more reliable than traditional analysis. To conclude, NPSR1 and PCSK2 represent potential markers to identify NETs and the possible primary tumor origin. Standardized IHC methods and the choice of primary antibody clone remain crucial for reliably analyzing diagnostic markers, while computer-aided proliferation index assessment yields a more objective classification of NENs.