Comparative analysis of pituitary adenoma with and without apoplexy in pediatric and adolescent patients: a clinical series of 80 patients

Pituitary adenomas (PAs) have a low incidence in pediatric and adolescent patients, and their clinical characteristics remain unclear. As a severe complication of PA, apoplexy was investigated in young patients in the present study. Eighty patients younger than 20 years with PAs who underwent surgery were included and divided into an apoplexy group and non-apoplexy group. The clinical data of these two groups were statistically analyzed and compared. The study included 33 boys and 47 girls, with a mean age of 16.9 years. There were six (7.5%) adrenocorticotropic hormone–secreting, 13 (16.3%) growth hormone–secreting, 47 (58.7%) prolactin-secreting, and 14 (17.5%) non-functioning PAs. There were 34 (42.5%) patients in the apoplexy group and 46 (57.5%) patients in the non-apoplexy group. Pre-operatively, patients in the apoplexy group were more likely to have visual impairment (hazard ratio: 2.841, 95% confidence interval: 1.073–7.519; P = 0.033) and had poorer visual impairment scores than those in the non-apoplexy group (P = 0.027). Furthermore, a longer duration of symptoms before surgery was significantly correlated with a poorer visual outcome in the apoplexy group (R =  − 1.204; P = 0.035). However, apoplexy was not associated with tumor type, tumor size, resection rate, or tumor recurrence. Tumor apoplexy is common in pediatric and adolescent patients with PAs and is associated with more severe preoperative visual deficits. Hence, the appropriate timing of surgical treatment may be important for rescuing visual function in young PA patients.


Introduction
Although pituitary adenoma (PA) is the third most common intracranial tumor in adults, its incidence in pediatric and adolescent patients (age < 20 years) is low, accounting for approximately 2.6 to 6.1% of all pituitary tumors [5,16] and 1% of all intracranial tumors [30] in pediatric and adolescent patients. Nevertheless, previous studies [7,16] have reported that young patients with PAs may have unique clinical and pathological characteristics that require distinct analyses and considerations.
Apoplexy is a major complication of pituitary tumors and may result in sudden headaches, visual deficits, vomiting, and other symptoms [13]. However, the specific clinical features of pituitary adenoma apoplexy (PAA) remain unclear in pediatric and adolescent patients due to the limited number of reported cases. Moreover, conflicting results have been found in the literature concerning the clinical and prognostic implications of PAA in young patients [15,33].
In the present study, we analyzed clinical data from 80 pediatric and adolescent patients with PAs. We divided these patients into an apoplexy group and non-apoplexy group and compared their clinical characteristics.

Patients
Between January 2011 and February 2019, 3137 patients were diagnosed with PA and underwent surgery in the Department of Neurosurgery at The First Hospital of China Medical University. Of these, 80 (2.6%) were patients under the age of 20 who were included in the present study. All patients underwent trans-sphenoidal surgery. This study was approved by the institutional review board at The First Hospital of China Medical University, and written informed consent was obtained from each patient and their parent(s) for the use of clinical data for future research.

Clinical examinations
All patients routinely underwent pre-and post-operative hormonal tests, magnetic resonance imaging (MRI) of the sellar region, and visual examinations. The degree of tumor invasion to the supra-and parasellar regions was graded according to the Hardy-Wilson classification and Knosp classification. Tumor apoplexy was determined through preoperative MRI (Fig. 1) and intraoperative observation. Visual acuity and visual fields were evaluated, and the visual impairment score (VIS) of the German Ophthalmological Society was recorded, as previously in the literature [8]. Tumor size was calculated using the ellipsoid volume formula V = πabc/6, in which a, b, and c denote the anteroposterior diameter, transverse diameter, and axial diameter, respectively. The resection rate was determined by the comparison of preand post-operative MRI scans. Near-total resection (NTR) was defined as more than 90% tumor resection, while the subtotal resection (STR) was defined as less than 90% tumor resection.

Statistical analysis
Logistic regression, chi-squared tests, and t-tests were used to evaluate risk factors of apoplexy and to conduct comparisons between the two groups. Linear regression analysis was performed to examine the association between the duration of symptoms of apoplexy before surgery and postoperative visual outcome in the apoplexy group. All statistical analyses were performed using SPSS v. 25.0 (IBM Corporation, Armonk, NY, USA). A P-value of less than 0.05 was defined as statistically significant.
During the follow-up period, 10 (12.5%) patients were confirmed to exhibit tumor recurrence. The median progression-free survival (PFS) was 24 months (range: with hypopituitarism received long-term hormone replacement therapy. Among the 10 patients with tumor recurrence, five were treated with secondary surgery; of these, three with PRL-secreting tumors were treated with bromocriptine, while the remaining two patients were followed up with due to their asymptomatic recurrence.

Comparison of tumors with and without apoplexy
The apoplexy group included 34 (42.5%) patients and the non-apoplexy group included 46 (57.5%) patients. Patients' sex, age, preoperative body mass index (BMI), visual function, tumor type, hypopituitarism, tumor size, supra-and parasellar invasion, surgical approach, extent of resection, and duration of surgery were compared between the two groups. Results are presented in Tables 1 and 2. A total of 15 and 10 patients suffered visual deficits during the preoperative period in the apoplexy group and non-apoplexy group, respectively. More patients in the apoplexy group showed preoperative visual abnormalities [hazard ratio (HR): 2.841, 95% confidence interval (CI): 1.073-7.519; P = 0.033] (Fig. 2a). Patients in the apoplexy group had significantly worse preoperative VIS values than those in the non-apoplexy group (P = 0.027) (Fig. 2b). Although surgery improved visual function in both groups, the postoperative VIS values of the apoplexy group remained worse than those of the non-apoplexy group (P = 0.034) ( Table 1). Among the 15 patients in the apoplexy group with preoperative visual deficits, the duration of symptoms before surgery ranged from four days to 23 days (mean: 13.9 ± 6.2 days) and linear regression analysis showed that a longer duration of symptoms before surgery significantly correlated with poorer visual outcome (R = − 1.204; P = 0.035) (Fig. 2c). PAA was not associated with sex (P = 0.364), age (P = 0.383), or pre-operative hypopituitarism (P = 0.655). Post-operatively, six patients (17.6%) in the apoplexy group and 14 patients (30.4%) in the non-apoplexy group with hypopituitarism received long-term hormone replacement therapy (P = 0.192). There were also no significant differences between the two groups in terms of tumor size (P = 0.848) or invasiveness (Hardy-Wilson classification, P = 0.142; Knosp classification, P = 0.117). PAA also did not significantly affect operative time (P = 0.192) or resection rate (P = 0.572). The difference in Ki-67 proliferation index between the apoplexy group and non-apoplexy group did not reach statistical significance (P = 0.070) (Fig. 2d). As shown in Table 3, tumor apoplexy was not associated with postoperative tumor recurrence (HR: 1.414, 95% CI: 0.375-5.319; P = 0.608) (Fig. 2e). Cox multivariate regression analysis demonstrated that only STR was independently correlated with postoperative tumor recurrence (HR: 4.556, 95% CI: 1.276-16.260; P = 0.019), although the Knosp classification (HR: 3.533, 95% CI: 1.004-12.563; P = 0.043) (Fig. 2f) and STR (HR: 4.000, 95% CI 0.956-16.667; P = 0.045) (Fig. 2g) were both significantly associated with recurrence in univariate regression.

Discussion
PAA was first reported in 1898 and cases were first summarized in 1950 [1]. As a severe complication, in adult PA patients, the overall incidence of apoplexy is 0.6% to 25% [13,17,23,26], and the condition is more likely to occur in those aged 30 to 50 years [32]. Hypotheses concerning the pathogenesis of PAA include the following: hypertension [13,21] or acute hypotension [20,27]; history of pelvic, cardiac, or vascular surgery [11,13,31]; history of antiplatelet and/or anticoagulant medication [11,13,17,18,27]; history of gonadotropin-releasing hormone medication [21] or dopamine-agonist medication [4]; trauma [21]; presence of aneurysm [22]; and radiotherapy [21,31]. However, none of the pediatric and adolescent patients in the present study had any of the above history or clinical features. However, a large tumor size [18], imbalance in high local metabolism, and poor nutritional status [20] may also lead to local vascular occlusion and rupture and may result in the development of apoplexy. Moreover, molecular factors (e.g., vascular endothelial growth factor [12], high-mobility group box1 [19], estrogen [14], and AIP mutation [28]) may also contribute to the development of apoplexy by altering the microvascular permeability, promoting local subacute intratumoral inflammatory responses, and inducing the faster growth of tumor cells. However, PAA has not been thoroughly described in pediatric and adolescent patients due to the limited number of published cases.
In the present study, apoplexy was found in 42.5% of pediatric and adolescent PAs. This result is consistent with those of previous reports, where the prevalence of PAA in pediatric and adolescent patients was higher than that in adult patients (approximately 0.6-25%) [13,17,23,26]. Moreover, consistent with in previous studies, although apoplexy is more frequently found in non-functional tumors in adults [33], our present study corroborated the idea that more PAs with apoplexy are functional tumors in pediatric and adolescent patients [6,16]. Although the precise reason for this phenomenon remains unclear, some plausible suggestions have been made. In adult patients, functional PAs are often detected by specific hormone-secreting symptoms, while non-functional PAs are often detected by apoplexy triggering a sudden increase in tumor size and an obvious occupying effect. In contrast, the majority of PAs in pediatric and adolescent patients are functional [3,9,10], and non-functioning PA is extremely rare in young patients [29]. Moreover, hormonal symptoms cannot be easily identified in the growing stages of each organ; therefore, apoplexy-related symptoms are frequently found in functional PAs in younger patients.
Some authors have reported that, in pediatric and adolescent patients, PAA may be associated with tumor size, tumor invasiveness, and PRL-secretory type [33]. However, other studies have not found such associations [15]. Similarly, in our present study, apoplexy did not relate to tumor size, Knosp/Hardy-Wilson classifications, or PRL tumors. Nevertheless, our results indicate that, as is true in adult patients [13], PAA is significantly associated with preoperative visual loss and/or deficits in pediatric and adolescent patients. Although surgery improved visual function in both groups, the post-operative VIS values of the apoplexy group were still worse than those of the non-apoplexy group. Moreover, in the apoplexy group, the duration of symptoms before surgery negatively correlated with the postoperative visual outcome. Thus, early surgery to decompress the optic nerves is critical. It has been reported that it is optimal to operate within one week or even 48 h, which can reduce the rate of disability [2,24]. Therefore, the appropriate timing of surgical management is essential for pediatric and adolescent patients with PAA to rescue visual function. It has been reported that apoplexy-inducing pituitary adenomas may have a tendency for rapid growth; if so, this would induce an insufficient blood supply and an ischemic-hypoxic tumor microenvironment, leading to apoplexy [25]. However, in the present study, the difference in the Ki-67 proliferation index between the apoplexy group and the non-apoplexy group did not reach statistical significance (P = 0.070). A power analysis using Minitab version 17.0 showed that a sample size of more than 153 participants might be required to detect a difference in the magnitude if it were, in fact, significant at the P < 0.05 level. We also investigated the effects of apoplexy on tumor recurrence in pediatric and adolescent patients and found that the presence of tumor apoplexy did not affect tumor resection nor was it associated with tumor recurrence. Furthermore, using multivariate analysis, only STR was found to be independently associated with recurrence, and the Knosp classification correlated with the resection rate.

Limitations
Our study was a retrospective study with a limited sample size, and bias may have existed. Thus, more convincing clinical evidence is still needed from subsequent large-sample prospective studies.

Conclusions
Tumor apoplexy is common in pediatric and adolescent patients with PA and is associated with more severe preoperative visual deficits. Therefore, the appropriate timing of surgical treatment may be important for rescuing visual function.