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

MRI Non-functioning Pituitary neuroendocrine tumours Preoperative Radiomics, WHO Classification

Journal

Acta neurochirurgica
ISSN: 0942-0940
Titre abrégé: Acta Neurochir (Wien)
Pays: Austria
ID NLM: 0151000

Informations de publication

Date de publication:
20 Feb 2024
Historique:
received: 26 10 2023
accepted: 18 01 2024
medline: 20 2 2024
pubmed: 20 2 2024
entrez: 20 2 2024
Statut: epublish

Résumé

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. 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. 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. 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.

Sections du résumé

BACKGROUND 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 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 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 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.

Identifiants

pubmed: 38376544
doi: 10.1007/s00701-024-05977-4
pii: 10.1007/s00701-024-05977-4
doi:

Types de publication

Journal Article

Langues

eng

Sous-ensembles de citation

IM

Pagination

91

Informations de copyright

© 2024. The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature.

Références

Ahmadi J, North CM, Segall HD, Zee CS, Weiss MH (1986) Cavernous sinus invasion by pituitary adenomas. AJR Am J Roentgenol 146(2):257–262
doi: 10.2214/ajr.146.2.257 pubmed: 3484572
Almeida JP, Stephens CC, Eschbacher JM et al (2019) Clinical, pathologic, and imaging characteristics of pituitary null cell adenomas as defined according to the 2017 World Health Organization criteria: a case series from two pituitary centers. Pituitary 22(5):514–519
doi: 10.1007/s11102-019-00981-9 pubmed: 31401793
Asa SL, Mete O, Perry A, Osamura RY (2022) Overview of the 2022 WHO Classification of Pituitary Tumors. Endocr Pathol 33(1):6–26
doi: 10.1007/s12022-022-09703-7 pubmed: 35291028
Asa SL, Mete O, Riddle ND, Perry A (2023) Multilineage pituitary neuroendocrine tumors (PitNETs) Expressing PIT1 and SF1. Endocr Pathol 34(3):273–278
doi: 10.1007/s12022-023-09777-x pubmed: 37268858
Asha MJ, Takami H, Velasquez C, Oswari S, Almeida JP, Zadeh G, Gentili F (2019) Long-term outcomes of transsphenoidal surgery for management of growth hormone-secreting adenomas: single-center results. J Neurosurg 11:1–11. https://doi.org/10.3171/2019.6.JNS191187
doi: 10.3171/2019.6.JNS191187
Cazabat L, Dupuy M, Boulin A, Bernier M, Baussart B, Foubert L, Raffin-Sanson M-L, Caron P, Bertherat J, Gaillard S (2014) Silent, but not unseen: multimicrocystic aspect on T2-weighted MRI in silent corticotroph adenomas. Clin Endocrinol (Oxf) 81(4):566–572
doi: 10.1111/cen.12443 pubmed: 24601912
Cordeiro D, Xu Z, Mehta GU et al (2018) Hypopituitarism after gamma knife radiosurgery for pituitary adenomas: a multicenter, international study. J Neurosurg 131(4):1188–1196
doi: 10.3171/2018.5.JNS18509
Fan Y, Jiang S, Hua M, Feng S, Feng M, Wang R (2019) Machine learning-based radiomics predicts radiotherapeutic response in patients with acromegaly. Front Endocrinol 10:588
doi: 10.3389/fendo.2019.00588
Goyal-Honavar A, Sarkar S, Asha HS, Kapoor N, Thomas R, Balakrishnan R, Chacko G, Chacko AG (2021) Impact of experience on outcomes after endoscopic transsphenoidal surgery for acromegaly. World Neurosurgery 151:e1007–e1015
doi: 10.1016/j.wneu.2021.05.030 pubmed: 34023462
Goyal-Honavar A, Sarkar S, Hesarghatta A, Kapoor N, Balakrishnan R, Vanjare H, Chacko G, Chacko A (2021) A clinicoradiological analysis of silent corticotroph adenomas after the introduction of pituitary-specific transcription factors. Acta Neurochir. https://doi.org/10.1007/s00701-021-04911-2
doi: 10.1007/s00701-021-04911-2 pubmed: 34181087
Haddad AF, Young JS, Oh T et al (2020) Clinical characteristics and outcomes of null-cell versus silent gonadotroph adenomas in a series of 1166 pituitary adenomas from a single institution. Neurosurg Focus 48(6):E13
doi: 10.3171/2020.3.FOCUS20114 pubmed: 32480370
Jahangiri A, Wagner JR, Pekmezci M, Hiniker A, Chang EF, Kunwar S, Blevins L, Aghi MK (2013) A comprehensive long-term retrospective analysis of silent corticotrophic adenomas vs hormone-negative adenomas. Neurosurgery 73(1):8–18
doi: 10.1227/01.neu.0000429858.96652.1e pubmed: 23685641
J Jia L Meng G Song S Sun C Li J Tian Y Zhang 2020 Prediction of response to stereotactic radiotherapy for nonfunctioning pituitary adenoma using radiomic feature https://doi.org/10.21203/rs.2.21209/v1
Kiseljak-Vassiliades K, Carlson NE, Borges MT, Kleinschmidt-DeMasters BK, Lillehei KO, Kerr JM, Wierman ME (2015) Growth hormone tumor histological subtypes predict response to surgical and medical therapy. Endocrine 49(1):231–241
doi: 10.1007/s12020-014-0383-y pubmed: 25129651
Langlois F, Lim DST, Yedinak CG, Cetas I, McCartney S, Cetas J, Dogan A, Fleseriu M (2018) Predictors of silent corticotroph adenoma recurrence; a large retrospective single center study and systematic literature review. Pituitary 21(1):32–40
doi: 10.1007/s11102-017-0844-4 pubmed: 29032459
Louis DN, Perry A, Wesseling P et al (2021) The 2021 WHO Classification of Tumors of the Central Nervous System: a summary. Neuro Oncol 23(8):1231–1251
doi: 10.1093/neuonc/noab106 pubmed: 34185076 pmcid: 8328013
MacFarlane J, Gillett D, Koulouri O, Bashari W, Casey R, Gurnell M (2022) Radiomics as a tool for risk stratification of non-functioning pituitary adenomas following primary surgery. Endocr Abstr. https://doi.org/10.1530/endoabs.86.OC3.5
doi: 10.1530/endoabs.86.OC3.5
Machado LF, Elias PCL, Moreira AC, Dos Santos AC, Murta Junior LO (2020) MRI radiomics for the prediction of recurrence in patients with clinically non-functioning pituitary macroadenomas. Comput Biol Med 124:103966
doi: 10.1016/j.compbiomed.2020.103966 pubmed: 32860977
Mendi BAR, Batur H, Çay N, Çakır BT (2023) Radiomic analysis of preoperative magnetic resonance imaging for the prediction of pituitary adenoma consistency. Acta Radiol 64(8):2470–2478
doi: 10.1177/02841851231174462 pubmed: 37170546
Mete O, Lopes MB (2017) Overview of the 2017 WHO Classification of Pituitary Tumors. Endocr Pathol 28(3):228–243
doi: 10.1007/s12022-017-9498-z pubmed: 28766057
Micko ASG, Wöhrer A, Wolfsberger S, Knosp E (2015) Invasion of the cavernous sinus space in pituitary adenomas: endoscopic verification and its correlation with an MRI-based classification. JNS 122(4):803–811
doi: 10.3171/2014.12.JNS141083
Osborn AG, Louis DN, Poussaint TY, Linscott LL, Salzman KL (2022) The 2021 World Health Organization Classification of Tumors of the Central Nervous System: what neuroradiologists need to know. Am J Neuroradiol. https://doi.org/10.3174/ajnr.A7462
doi: 10.3174/ajnr.A7462 pubmed: 35710121 pmcid: 9262075
Peng A, Dai H, Duan H, Chen Y, Huang J, Zhou L, Chen L (2020) A machine learning model to precisely immunohistochemically classify pituitary adenoma subtypes with radiomics based on preoperative magnetic resonance imaging. Eur J Radiol 125:108892
doi: 10.1016/j.ejrad.2020.108892 pubmed: 32087466
Rui W, Qiao N, Wu Y, Zhang Y, Aili A, Zhang Z, Ye H, Wang Y, Zhao Y, Yao Z (2022) Radiomics analysis allows for precise prediction of silent corticotroph adenoma among non-functioning pituitary adenomas. Eur Radiol 32(3):1570–1578
doi: 10.1007/s00330-021-08361-3 pubmed: 34837512
Wang H, Chang J, Zhang W et al (2023) Radiomics model and clinical scale for the preoperative diagnosis of silent corticotroph adenomas. J Endocrinol Invest 46(9):1843–1854
doi: 10.1007/s40618-023-02042-2 pubmed: 37020103
Won SY, Lee N, Park YW, Ahn SS, Ku CR, Kim EH, Lee S-K (2022) Quality reporting of radiomics analysis in pituitary adenomas: promoting clinical translation. Br J Radiol 95(1139):20220401
doi: 10.1259/bjr.20220401 pubmed: 36018049 pmcid: 9793472
Zhang Y, Luo Y, Kong X, Wan T, Long Y, Ma J (2022) A preoperative MRI-based radiomics-clinicopathological classifier to predict the recurrence of pituitary macroadenoma within 5 years. Front Neurol 12:780628
doi: 10.3389/fneur.2021.780628 pubmed: 35069413 pmcid: 8767054
Zhang S, Song G, Zang Y, Jia J, Wang C, Li C, Tian J, Dong D, Zhang Y (2018) Non-invasive radiomics approach potentially predicts non-functioning pituitary adenomas subtypes before surgery. Eur Radiol 28(9):3692–3701. https://doi.org/10.1007/s00330-017-5180-6
doi: 10.1007/s00330-017-5180-6 pubmed: 29572634

Auteurs

Quantitative Imaging Research and Artificial Intelligence Lab, Department of Radiation Oncology Unit II, Ida B Scudder Cancer Centre, Christian Medical College, Vellore, India.

Abhijit Goyal-Honavar (A)

Department of Neurosurgery, Christian Medical College, Vellore, India.

Ari G Chacko (AG)

Department of Neurosurgery, Christian Medical College, Vellore, India.

Anitha Jasper (A)

Department of Radiodiagnosis, Christian Medical College, Vellore, India.

Geeta Chacko (G)

Department of General Pathology, Christian Medical College, Vellore, India.

Devadhas Devakumar (D)

Department of Nuclear Medicine, Christian Medical College, Vellore, India.

Joshua Anand Seelam (JA)

Department of Radiodiagnosis, Christian Medical College, Vellore, India.

Balu Krishna Sasidharan (BK)

Quantitative Imaging Research and Artificial Intelligence Lab, Department of Radiation Oncology Unit II, Ida B Scudder Cancer Centre, Christian Medical College, Vellore, India.

Simon P Pavamani (SP)

Quantitative Imaging Research and Artificial Intelligence Lab, Department of Radiation Oncology Unit II, Ida B Scudder Cancer Centre, Christian Medical College, Vellore, India.

Hannah Mary T Thomas (HMT)

Quantitative Imaging Research and Artificial Intelligence Lab, Department of Radiation Oncology Unit II, Ida B Scudder Cancer Centre, Christian Medical College, Vellore, India. hannah.thomas@cmcvellore.ac.in.

Classifications MeSH