Is radiomics a useful addition to magnetic resonance imaging in the preoperative classification of PitNETs?

BACKGROUND

The WHO 2021 introduced the term pituitary neuroendocrine tumours (PitNETs) for pituitary adenomas and incorporated transcription factors for subtyping, prompting the need for fresh diagnostic methods. Current biomarkers struggle to distinguish between high- and low-risk non-functioning PitNETs. We explored if radiomics can enhance preoperative decision-making.

METHODS

Pre-treatment magnetic resonance (MR) images of patients who underwent surgery between 2015 and 2019 with available WHO 2021 classification were used. The tumours were manually segmented on the T1w, T1-contrast enhanced, and T2w images using 3D Slicer. One hundred Pyradiomic features were extracted from each MR sequence. Models were built to classify (1) somatotroph and gonadotroph PitNETs and (2) high- and low-risk subtypes of non-functioning PitNETs. Feature were selected independently from the MR sequences and multi-sequence (combining data from more than one MR sequence) using Boruta and Pearson correlation. Support vector machine (SVM), logistic regression (LR), random forest (RF), and multi-layer perceptron (MLP) were the classifiers used. Data imbalance was addressed using the Synthetic Minority Oversampling TEchnique (SMOTE). Performance of the models were evaluated using area under the receiver operating curve (AUC), accuracy, sensitivity, and specificity.

RESULTS

A total of 222 PitNET patients (train, n = 149; test, n = 73) were enrolled in this retrospective study. Multi-sequence-based LR model discriminated best between somatotroph and gonadotroph PitNETs, with a test AUC of 0.84, accuracy of 0.74, specificity of 0.81, and sensitivity of 0.70. Multi-sequence-based MLP model perfomed best for the high- and low-risk non-functioning PitNETs, achieving a test AUC of 0.76, accuracy of 0.67, specificity of 0.72, and sensitivity of 0.66.

CONCLUSIONS

Utilizing pre-treatment MRI and radiomics holds promise for distinguishing high-risk from low-risk non-functioning PitNETs based on the latest WHO classification. This could assist neurosurgeons in making critical decisions regarding surgery or alternative management strategies for PitNETs after further clinical validation.

Transcriptome of GH-producing pituitary neuroendocrine tumours and models are significantly affected by somatostatin analogues

Pituitary neuroendocrine tumours (PitNETs) are neoplasms of the pituitary that overproduce hormones or cause unspecific symptoms due to mass effect. Growth hormone overproducing GH-producing PitNETs cause acromegaly leading to connective tissue, metabolic or oncologic disorders. The medical treatment of acromegaly is somatostatin analogues (SSA) in specific cases combined with dopamine agonists (DA), but almost half of patients display partial or full SSA resistance and potential causes of this are unknown. In this study we investigated transcriptomic landscape of GH-producing PitNETs on several levels and functional models-tumour tissue of patients with and without SSA preoperative treatment, tumour derived pituispheres and GH3 cell line incubated with SSA to study effect of medication on gene expression. MGI sequencing platform was used to sequence total RNA from PitNET tissue, pituispheres, mesenchymal stromal stem-like cells (MSC), and GH3 cell cultures, and data were analysed with Salmon-DeSeq2 pipeline. We observed that the GH-producing PitNETs have distinct changes in growth hormone related pathways related to its functional status alongside inner cell signalling, ion transport, cell adhesion and extracellular matrix characteristic patterns. In pituispheres model, treatment regimens (octreotide and cabergoline) affect specific cell proliferation (MKI67) and core functionality pathways (RYR2, COL8A2, HLA-G, ARFGAP1, TGFBR2). In GH3 cells we observed that medication did not have transcriptomic effects similar to preoperative treatment in PitNET tissue or pituisphere model. This study highlights the importance of correct model system selection for cell transcriptomic profiling and data interpretation that could be achieved in future by incorporating NGS methods and detailed cell omics profiling in PitNET model research.

Distribution of E- and N-cadherin in subgroups of non-functioning pituitary neuroendocrine tumours

PURPOSE

Clinically non-functioning pituitary neuroendocrine tumours (NF-PitNETs) present a varying degree of aggressiveness, and reliable prognostic markers are lacking. We aimed to characterise the distribution of E- and N-cadherin in corticotroph, PIT1 and null-cell NF-PitNETs, and link it to the course of the tumours.

METHODS

The distribution of E- and N-cadherin was investigated by immunohistochemistry in a retrospective cohort of 30 tumours of the less common NF-PitNETs (corticotroph (N = 18), PIT1 (N = 8) and null-cell PitNETs (N = 4)). Immunoreactive scores (IRS) were compared to previously presented cohorts of gonadotroph NF-PitNETs (N = 105) and corticotroph functioning PitNETs (N = 17).

RESULTS

We found a low IRS for the extra-cellular domain of E-cadherin (median 0 (IQR 0-0, N = 135)), a medium to high IRS for the intra-cellular domain of E-cadherin (median 6 (IQR 4-9)) and a high IRS for N-cadherin (median 12 (IQR 10.5-12)) throughout the cohort of NF-PitNETs. The corticotroph NF-PitNETs presented a higher IRS for both the extra- and intra-cellular domain of E-cadherin (median 0 (IQR 0-1) and median 9 (IQR 6-12), respectively) than the gonadotroph NF-PitNETs (p < 0.001 for both comparisons). Presence of nuclear E-cadherin was associated with a weaker staining for the intra-cellular domain of E-cadherin (median 4 (IQR 0.5-6) and median 9 (IQR 9-12), for tumours with and without nuclear E-cadherin, respectively), and with a lower rate of re-intervention (p = 0.03).

CONCLUSIONS

Considering our results and the benign course of NF-PitNETs, we suggest that a high N-cadherin and downregulation of membranous E-cadherin are not associated with a more aggressive tumour behaviour in these subgroups of NF-PitNETs.

Stem Cell Transcription Factor Sox2 Is Expressed in a Subset of Folliculo-stellate Cells of Growth Hormone-Producing Pituitary Neuroendocrine Tumours and Its Expression Shows No Association with Tumour Size or IGF1 Levels: a Clinicopathological Study of 109 Cases

Sox2 is one of the transcription factors responsible for the maintenance of stem cell phenotype. It has been implicated as a marker of stem cells in normal pituitaries and pituitary neuroendocrine tumours. To explore the clinical significance of Sox2 expression in histological sections, we performed immunohistochemical detection of Sox2 in 113 pituitary neuroendocrine tumours from 109 patients with acromegaly. In 11 tumours, we performed double immunostaining for Sox2, annexin A1 and S100 protein. Tumours were characterised using the WHO classification system. Proliferative activity and invasion were assessed. The amount of immunoreactive cells was evaluated and correlated with tumour size and biochemical features (levels of IGF1, GH, prolactin, βTSH). Sox2⁺ cells were identified in 35/38 normal pituitaries adjacent to the tumours. In 36 tumours (33%), ≥ 1% of the cells expressed Sox2, in 24 cases (22%), Sox2⁺ cells comprised < 1% and 49 cases (45%) were negative. We found no significant differences between Sox2⁺ and Sox2^- groups with respect to the age, initial levels of GH, IGF1, prolactin, βTSH, tumour size, invasion, proliferative activity or histological features. We observed a positive correlation between Sox2⁺ cell count and βTSH immunoreactive cells (r = 0.459, p < 0.001) that was further verified by multivariate analysis. Using double stain, the majority of Sox2⁺ cells coexpressed annexin A1 (average 89%) and S100 protein (average 76.2%) and showed morphological features of folliculo-stellate cells. Sox2⁺ cells are thus commonly present in growth hormone-producing tumours and normal pituitaries, and their amount does not have any prognostic significance. Most of these cells represent a subpopulation of folliculo-stellate cells, pointing out to their role as a possible stem cell population.

A standardised diagnostic approach to pituitary neuroendocrine tumours (PitNETs): a European Pituitary Pathology Group (EPPG) proposal

The 2017 World Health Organization (WHO) classification proposes to type and subtype primary adenohypophyseal tumours according to their cell lineages with the aim to establish more uniform tumour groups. The definition of atypical adenoma was removed in favour of high-risk adenoma, and the assessment of proliferative activity and invasion was recommended to diagnose aggressive tumours. Recently, the International Pituitary Pathology Club proposed to replace adenoma with the term of pituitary neuroendocrine tumour (PitNET) to better reflect the similarities between adenohypophyseal and neuroendocrine tumours of other organs. The European Pituitary Pathology Group (EPPG) endorses this terminology and develops practical recommendations for standardised reports of PitNETs that are addressed to histo- and neuropathologists. This brief report presents the results of EPPG's consensus for the reporting of PitNETs and proposes a diagnostic algorithm.