Classifying Pituitary Adenoma Invasiveness Based on Radiological, Surgical and Histological Features: A Retrospective Assessment of 903 Cases

Abstract

Invasiveness is a major predictor of surgical outcome and long-term prognosis in patients with pituitary adenomas (PAs). We assessed PA invasiveness via radiological, surgical and histological perspectives to establish a classification scheme for predicting invasive behavior and poor prognosis. We retrospectively analyzed 903 patients who underwent transnasal-transsphenoidal surgery between January 2013 and December 2019. Radiological (hazard ratio (HR) 5.11, 95% confidence interval (CI): 3.98−6.57, p < 0.001) and surgical (HR 6.40, 95% CI: 5.09−8.06, p < 0.001) invasiveness better predicted gross-total resection (GTR) and recurrence/progression-free survival (RPFS) rates than did histological invasiveness (HR 1.44, 95% CI: 1.14−1.81, p = 0.003). Knosp grades 2 (HR 4.63, 95% CI: 2.13−10.06, p < 0.001) and 3 (HR 2.23, 95% CI: 1.39−3.59, p = 0.011) with surgical invasiveness were better predictors of prognosis than corresponding Knosp grades without surgical invasiveness. Classifications 1 and 2 were established based on radiological, surgical and histological invasiveness, and Knosp classification and surgical invasiveness, respectively. Classification 2 predicted RPFS better than Knosp classification and Classification 1. Overall, radiological and surgical invasiveness were clinically valuable as prognostic predictors. The convenience and good accuracy of Invasiveness in Classification 2 is useful for identifying invasive PAs and facilitating the development of treatment plans.

Keywords: cavernous sinus; classification; histology; invasiveness; pituitary adenoma; radiology; surgery.

Figure 1

Flowchart summarizing the enrollment strategy and design of the study. PAs: Pituitary Adenomas; TTS: Transnasal-Transsphenoidal Surgery.

Figure 2

Venn chart illustrating the number and interactive information of patients with radiological, surgical, and histological invasiveness.

Figure 3

Radar chart illustrating the proportional distribution of invasiveness for different clinical subtypes. PRL: Prolactin; GH: Growth Hormone; ACTH: Adrenocorticotropic Hormone.

Figure 4

Kaplan–Meier survival curves of different invasiveness groups. (A) Kaplan–Meier survival curves of invasiveness classification 1. (B) Kaplan–Meier survival curves of histological invasiveness. (C) Kaplan–Meier survival curves of surgical invasiveness. (D) Kaplan–Meier survival curves of radiological invasiveness. KM: Kaplan–Meier; HR: Hazard Ratio.

Figure 5

Bar chart illustrating the gross-total resection (GTR) rates of Knosp grade 2–3 with or without surgical invasiveness. SI: surgical invasiveness. ** p < 0.01.

Figure 6

Kaplan–Meier survival curves of Knosp grade 2–3 with or without surgical invasiveness. (A) Kaplan–Meier survival curves of Knosp grade 2 with or without surgical invasiveness. (B) Kaplan–Meier survival curves of Knosp grade 3 with or without surgical invasiveness. KM: Kaplan–Meier; HR, Hazard Ratio.

Figure 7

Kaplan–Meier survival analysis of Invasion Classification 2. (A) Kaplan–Meier survival curves of Invasion Classification 2. (B) Kaplan–Meier survival curves of two subgroups in grade 2. Non-invasiveness: Invasion Classification grade 0–1; Unilateral invasiveness: Invasion Classification grade 2; Invasiveness: Invasion Classification grade 3–4; 2A: Knosp grade 3 PAs without surgical invasiveness; 2B: Knosp grade 0–2 PAs with surgical invasiveness. KM: Kaplan–Meier; HR, Hazard Ratio.

Figure 8

Receiver operator characteristic (ROC) analysis of gross-total resection (GTR) and recurrence/progression (RP). ROC curves of predicting GTR in revised-Knosp classification, Invasiveness Classification 1, and invasiveness classification 2 (A). ROC curves of predicting RP for revised-Knosp classification, Invasiveness Classification 1, and Invasiveness Classification 2 (B). CI: Confidence Interval; FPR: False Positive Rate; TPR: True Positive Rate.