Risk Factor and Replacement Therapy Analysis of Pre- and Postoperative Endocrine Deficiencies for Craniopharyngioma

Simple Summary Although craniopharyngiomas are pathologically defined as benign tumors, their lesions could involve the hypothalamic–pituitary axis and adjacent to several important anatomical structures. Treatment strategies, including surgery and radiation, have a higher risk of damaging these important structures, which leads to excess recurrence and disability rates. Endocrine deficiencies caused by hypothalamic–pituitary axis damage are the most common symptoms and postoperative complications of craniopharyngioma patients, which severely affect perioperative and long-term treatment effects. However, limited studies have focused on the endocrine deficiencies of craniopharyngioma patients. In this study, the influencing factors and replacement therapies for pre- and postoperative pituitary hormone deficiency were retrospectively analyzed in 126 craniopharyngioma patients, which provides evidence for surgical strategy selection and long-term management during craniopharyngioma treatments. Abstract Background: Pituitary hormone deficiency (PHD) is one of the most common symptoms and postoperative complications of craniopharyngiomas (CPs). However, the risk factors for PHD in CPs are little known. The purpose of this study was to analyze the risk factors of pre- and postoperative PHD and to investigate replacement therapy for CP patients. Methods: A retrospective study of 126 patients diagnosed with CP was performed. Univariate analysis was performed using Pearson’s chi-squared test or Fisher’s exact test, and a multiple logistic binary regression model was used to identify the influencing factors of pre- and postoperative PHD in craniopharyngioma. Results: Children and patients with hypothalamic involvement were more likely to have preoperative PHD. Patients with suprasellar lesions had a high risk of postoperative PHD, and preoperative PHD was a risk factor for postoperative PHD. Conclusion: Children have a high incidence of preoperative PHD. Preoperative PHD can serve as an independent risk factor for postoperative PHD. Preoperative panhypopituitarism can serve as an indication of pituitary stalk sacrifice during surgery. The management of replacement therapy for long-term postoperative endocrine hormone deficiency in patients with craniopharyngioma should be enhanced.


Introduction
Craniopharyngiomas (CPs) originate from remnants of the craniopharyngeal duct epithelium and account for 1.2-4.6% of all brain tumor [1]. Patients with craniopharyngiomas have a bimodal age distribution, with peak incidence rates in children aged 5-14 years and in adults aged 50-74 years. Adamantinomatous craniopharyngioma (ACP) and papillary craniopharyngioma (PCP) are two histological subtypes of craniopharyngiomas. As CP is pathologically defined as a benign tumor (WHO I), it is curable when gross total resection is achieved. The 10-year survival rate of patients with CP ranges from 65% to 100%; however, In total, 485 CP patients who underwent surgical treatment from 2010 to 2019 were retrospectively analyzed. All patients were diagnosed with CP based on a pathological examination. The inclusion criteria were as follows: (1) the pre-and postoperative clinical data, laboratory, and imaging examinations were complete; (2) the survival time more than 1 year; (3) without other endocrine diseases; (4) with physical condition can finish follow-up. Finally, 126 patients were included, and all patients were accurately followed up. Patient information, including age, sex, main symptoms, number of previous surgeries, intraoperative conditions, and pathological classification. Endocrine hormone and neuroimaging evaluations were performed in all patients pre-and post-surgery, re-examinations were conducted during follow-up, and the replacement therapies of each patient were recorded.
Permanent DI was defined as persistent DI symptoms for more than 1 year, requiring medication to control the amount of urine.

Neuroradiological Evaluation
All patients underwent preoperative brain CT and MRI examinations. The location, size, texture, and involvement of the hypothalamus in the lesions were analyzed. Brain MRI was reexamined within 24 h and at 3 months, 6 months, and 1 year post-operation. Tumor size was calculated according to the maximum diameter based on MRI, and the patients were divided into ≤3 cm and >3 cm groups. According to the location of the lesion, we grouped the patients into intrasellar, suprasellar-extraventricular, and suprasellarintraventricular types. Based on the tumor texture, patients were classified into three groups: solid, cystic-solid, and cystic. The relationship between the tumor and the hypothalamus of each patient was also classified into three grades according to the Puget grade: Grade 0, no hypothalamic involvement; Grade 1, the tumor abutting or displacing the hypothalamus; Grade 2, hypothalamic involvement (the hypothalamus is no longer identifiable) [11].

Treatment
Patients with preoperative thyroid hormone and cortisol deficiencies were routinely supplemented. In patients with acute hydrocephalus, external ventricular drainage should be performed first, if necessary. The surgical strategy depends on a preoperative imaging examination. There are two mainstream approaches: transsphenoidal approaches (TSA) and transcranial approaches (TCA). The intrasellar type is mainly resected via transsphenoidal endoscopy or microscopy. The pituitary stalk preservation standard means that the pituitary stalk can be identified and is intact during surgery. Complete resection was defined as the complete removal of the lesion under a microscope with no residual tumor or calcification on reimaging [12].

Follow-Up
The patients were followed up at 1, 3, and 6 months and every year postoperatively. Endocrine hormone and imaging examinations were performed at each follow-up visit. Pituitary hormone status, replacement therapy of patients, and tumor recurrence, if found, were recorded. Recurrence was defined as the growth of the residual tumor or a new lesion on follow-up MR imaging [3]. The follow-up time ranged from 12 to 111 months, with a mean of 54.1 ± 33 months.

Statistical Methods
Data were processed using IBM SPSS22.0 software (IBM Corp, Armonk, NY, USA). Mean values between groups were analyzed using the Student's t-test. Pearson's Chi-Squared (χ 2 ) test or Fisher's exact test was used for univariate analysis to describe the frequency and ratio of categorical variables. Variables with univariate analysis results p < 0.1 were included in the multivariate logistic regression model. A multivariate logistic binary regression model was used to determine independent risk factors that affect pituitary hormone deficiency in patients with CP pre-and post-surgery. A two-sided level of p < 0.05 was considered statistically significant.

Patient Characteristics and Symptoms
The basic information and clinical characteristics of the 126 patients are shown in Table 1. In this study, 47 patients were female (37.3%). Patients' ages ranged from 3 to 74 years, with an average of 35.4 ± 19.3 years, including 29 children (<18 years old, 23%) and 97 adults (≥18 years old, 77%). Approximately half of the patients showed symptoms related to intracranial hypertension (50.8%), such as headaches, nausea, and vomiting. Vision loss and visual field defects were observed in 59 (46.9%) patients. Twenty patients (15.9%) showed symptoms of hormone deficiency, including abnormal development, sexual dysfunction, amenorrhea, and diabetes insipidus. Twelve patients (9.6%) exhibited abnormal hypothalamic function, including abnormal sleep, weight gain, and personality changes. Another nine patients were incidentally discovered by regular check-ups, three patients for epilepsy, two patients for high fever, and one patient for a sudden loss of consciousness. Seventeen patients had a history of CP surgery, while the other 109 cases (86.5%) underwent surgery for the first time.

Endocrinological Function
The pre-and postoperative pituitary hormone levels and replacement treatments are shown in Figure   Only two patients with anterior pituitary hormone deficiency recovered after surgery in our cohort. One of them had a lower level of preoperative ACTH. Another patient had panhypopituitarism before surgery, and preoperative MR images revealed an intrasellar CP with a stroke. The pituitary stalk was not invaded by the tumor and was completely retained, and the endocrine function of this patient returned to normal after transsphenoidal resection (shown in Figure 4).  Replacement treatment was implemented in 65.6% of patients with HPTD, 46% with cortisol deficiency, and 23.5% and 20.4% with GD and GHD, respectively; only patients with DI received replacement treatment.

Surgical Strategies and Outcomes
Craniotomy was performed in 104 patients (82.5%), and resection through transsphenoidal endoscopy or microscopy was performed in 22 patients (17.5%). The pituitary stalk was completely retained in 74 patients (58.7%). Postoperative imaging confirmed that 80 patients underwent total resection (63.5%). According to postoperative pathological analysis, 73.8% of the patients were adamantinomatous, and 26.2% were papillary. Thirty-one patients (24.9%) relapsed during the postoperative follow-up period. Only two patients with anterior pituitary hormone deficiency recovered after surgery in our cohort. One of them had a lower level of preoperative ACTH. Another patient had panhypopituitarism before surgery, and preoperative MR images revealed an intrasellar CP with a stroke. The pituitary stalk was not invaded by the tumor and was completely retained, and the endocrine function of this patient returned to normal after transsphenoidal resection (shown in Figure 4).

The Difference in Clinical Presentation and Results between Children and Adults
Compared to the children-onset (CO) group, the adult-onset (AO) group more frequently presented with the symptom of visual impairment (52.6% vs. 24.1%), while the ratio of tumor size more than 3 cm was higher in the CO group (55.2% vs. 30.9%), suprasellar extension reaching the ventricle was also more frequently present in the CO group, preand postoperative PHD were significantly more prevalent in the CO group, especially CH (86.2% vs. 61.9%), as well as the postoperative transient DI. The results of operation showed that the AO group more easily got GTR and STR than the CO group (p < 0.01). Replacement treatment was implemented in 65.6% of patients with HPTD, 46% with cortisol deficiency, and 23.5% and 20.4% with GD and GHD, respectively; only patients with DI received replacement treatment.

Surgical Strategies and Outcomes
Craniotomy was performed in 104 patients (82.5%), and resection through transsphenoidal endoscopy or microscopy was performed in 22 patients (17.5%). The pituitary stalk was completely retained in 74 patients (58.7%). Postoperative imaging confirmed that 80 patients underwent total resection (63.5%). According to postoperative pathological analysis, 73.8% of the patients were adamantinomatous, and 26.2% were papillary. Thirty-one patients (24.9%) relapsed during the postoperative follow-up period.

The Difference in Clinical Presentation and Results between Children and Adults
Compared to the children-onset (CO) group, the adult-onset (AO) group more frequently presented with the symptom of visual impairment (52.6% vs. 24.1%), while the ratio of tumor size more than 3 cm was higher in the CO group (55.2% vs. 30.9%), suprasellar extension reaching the ventricle was also more frequently present in the CO group,

Risk Factors for Endocrine Dysfunction in Patients
The univariate analysis of preoperative pituitary hormone levels is shown in Table 2. There were differences between children and adults (p = 0.008); patients with hormone deficiency-related symptoms (p = 0.000) and hypothalamic-associated symptoms (p = 0.016) had a higher ratio of hormone deficiency. Imaging data analysis showed that factors influencing preoperative hormone levels included tumor size, location, and degree of hypothalamic involvement. Multiple logistic regression analysis showed that children were more susceptible to preoperative endocrine hormone deficiency (OR = 3.024, p = 0.024; 95% CI = 1.21-13.44), and patients with hypothalamic involvement had a high risk of preoperative hormone deficiency (Puget grade 1 vs. grade 0, OR = 4.974, p = 0.014; 95% CI = 1.39-17.81) (Puget grade 2 vs. grade 0, OR = 11.452, p = 0.000; 95% CI = 2.98-43.97) ( Table 3).

Discussion
Most patients with CP already have pituitary hormone deficiencies at the time of diagnosis; it has been reported that 40-87% of patients had at least one endocrine hormone deficit, and 17-27% of patients had central DI at the time of diagnosis [13,14]. Postoperatively, endocrine deficiencies of patients with CP are usually aggravated; more than 50% of patients with normal endocrine hormones develop a new deficit of the hypothalamicpituitary axis after surgery, 80-90% of patients with panhypopituitarism [15,16]. A 10-year follow-up study of patients with CP after surgery showed that the ratios of GH, FSH/LH, ACTH, and TSH deficiency were 88%, 90%, 86%, and 80%, respectively, and DI was about 65% [17]. Consistent with previous reports, in this study, 65.9% of the patients had preoperative anterior pituitary hormone deficiency, and 28.6% had panhypopituitarism. Postoperatively, the percentage of patients with hormone deficiency increased to 84.1%, and 67.5% presented panhypopituitarism. Of the patients in this study, 11.1% had preoperative central DI, while postoperative transient central DI was as high as 93.6%. Although treatment methods and surgical safety are improving daily, patients with CPs still have a high proportion of long-term endocrine defects, which leads to a poor prognosis and is a challenge for CP treatment.
In this study, we analyzed the factors influencing pre-and postoperative hormonal status in 126 patients with CP. Univariate analysis of preoperative hormone status indicated that patients with hypothalamus-related endocrine symptoms had a higher proportion of preoperative pituitary hormone deficiency. Imaging analysis showed that patients with extrasellar lesions, lesions > 3 cm in size, and lesions with hypothalamic involvement (Puget grades 1 and 2) had a higher ratio of preoperative endocrine defects. All patients with preoperative endocrine-related symptoms had different degrees of pituitary hormone deficiency, as confirmed by laboratory examinations. Capatina et al. performed a retrospective analysis of 107 CP cases (72 adults, 35 children) and found that children are more likely to have endocrine hormone deficiency at diagnosis than adults (GHD (68.8% vs. 17.1%), DI (28.5% vs. 8.3%)) [18]. Gautier et al. found that 10-18-year-old children were more prone to hormone deficiency by analyzing 171 patients with CP [19]. Consistent with these previous reports, we also revealed that children had a higher risk of endocrine hormone deficiency at the time of diagnosis. Most patients with hypothalamic involvement had hypothalamic-pituitary axis invasion and destruction, especially those with Puget grade 2. In our group, 82% (41/50) of Puget grade 2 patients had preoperative endocrine hormone deficiency; studies have also confirmed that hypothalamic involvement is associated with panhypopituitarism [20].
Univariate analysis of patients' postoperative hormonal status in our group demonstrated that location, size, degree of hypothalamic involvement, surgical methods, and preoperative hormonal status were the main influencing factors. Multivariate regression analysis indicated that preoperative hormonal status and lesion location were the main factors influencing postoperative hormonal status. It has been widely reported that endocrine defects of CPs are extremely rare to recover, which is distinct from those of pituitary tumors [17,[21][22][23]. Different to pituitary tumors, CP has large lesions at diagnosis, which severely destroy the hypothalamus-pituitary system and lead to permanent endocrine disorders. Surgical resection of the tumor cannot recover the endocrine function of patients. Only two patients with preoperative hormone deficiency recovered after surgery in our cohort; the pituitary stalk was not invaded by the tumor and was completely retained. We believe that the preoperative hormone deficiency was transient and was due to acute pituitary ischemia caused by acute compression. After surgery, the compression was relieved, and the pituitary function recovered. The preoperative hormone levels of the other 18 patients with normal postoperative hormone levels were also normal. These results indicate that preoperative hormone status is one of the main factors influencing postoperative hormone levels, and that preoperative hormone deficiency in patients with CP rarely leads to postoperative recovery. With the development of endoscopic technology, the endonasal endoscopic approach (EEA) has been widely used in CP surgery. The EEA has no limitations of visualization, which can be extended to the transplanum and transtuberculum approaches that provide direct midline access to sellar, supra-sellar, retrochiasmatic, and third ventricular tumors [24]. In recent years, the EEA has been used for not only purely sellar CP but also suprasellar CP, even with intra-ventricular extension [25]. The detailed surgical view of the surrounding nerve and vascular structures provided by endoscopy makes it easy for total resection of the tumor. Dho reported a group of patients undergoing EEA for craniopharyngioma; the rate of GTR was up to 91.1% (62/68) [26]. EEA also has advantages in endocrine function preservation, as it has been reported that TCA results in increased new hypopituitarism versus EEA (57.1% vs. 28.9%) [27]. Based on an analysis of 112 cases of surgically treated CPs, Mortini reported that patients who underwent transsphenoidal surgery had a lower risk of de novo endocrine hormone deficiencies [4]. In addition, studies have shown that EEA preserves hypothalamic function regardless of preoperative hypothalamic invasion [28]. The univariate analysis in this group also confirmed that, compared to craniotomy, the TSA had a low ratio of hormone deficiency. Multivariate analysis revealed that intrasellar lesions were associated with a lower risk of hormone deficiency. We usually choose the TSA for intrasellar lesions because it is less involved in the hypothalamic-pituitary axis (mostly Puget Grade 0), and the pituitary stalk is easier to retain during surgery, which has less effect on endocrine function. In the future, EEA should be considered the most important choice for CP to achieve better outcomes and fewer pituitary hormonal disorders [29].
Achieving tumor control and function preservation simultaneously is a challenge for CP treatment. Residual lesions are the main risk factors for CP recurrence. Some scholars believe that the pituitary stalk should be resected to avoid residual lesions and to prevent recurrence accompanied by hormone replacement therapy [4,23,30]. However, other researchers believe that the pituitary stalk should be retained as much as possible to preserve endocrine function, especially when the tumor does not invade the pituitary stalk [31]. Li et al. reported that pituitary stalk preservation reduced endocrine changes in 420 surgically treated CPs without increasing tumor recurrence [32]. In this study, the pituitary stalk was sacrificed during surgery in 52 patients (41.3%), all of whom experienced panhypopituitarism after the operation. However, whether the pituitary stalk should be retained during CP surgery remains controversial. According to previous reports and this study, surgery cannot reverse existing hormone deficiencies in patients with CP. Therefore, we believe that preoperative panhypopituitarism can serve as an indication for pituitary stalk scarification, but for patients with normal preoperative hormones, intrasellar lesions, and without hypothalamic-pituitary axis involvement, the pituitary stalk should be retained as much as possible to preserve endocrine function.
Postoperative hypothalamic damage is a common complication for CP, mainly related to increased body weight, altered circadian rhythms, behavioral changes, sleep irregularities, and so on; the rate is up to 65-80% in CP patients after GTR [33]. The Puget Classification was chosen to help surgeons classify tumors and avoid hypothalamic injury. Patients with no hypothalamic involvement (Grade 0) should attempt GTR, while those with clear hypothalamic invasion (Grade 2) should undergo subtotal re-section, and combined with adjuvant radiation therapy [27]. Many studies have also revealed that the risk of recurrence after GTR vs. STR + radiation therapy did not reach significance [34]. However, endocrinopathy is the most common late toxicity for radiation therapy. Rachel reported a group of pediatric patients treated with surgery and proton therapy and found that the surgery increased the rates of endocrine deficiency from 40% to 87%, complete hormone deficiency increased from 30% to 36%, while, after a median follow-up of 4.8 years from the initiation of RT treatment, up to 94% of patients had some form of endocrinopathy, and 56% had panhypopituitarism [35]. Another study also revealed that radiation therapy is effective in reducing the recurrence rates of STR-treated CPs but increases the rates of endocrine dysfunction [14]. In this study, most patients had endocrinopathies, and RT was less frequently recommended, but the potential adverse effects of RT on endocrine cannot be neglected.
Hormone deficiency in patients with CP is mostly life-long, and CP is known as a lifechanging tumor [36]. Long-term hormone deficiency severely influences the development and mental health of patients, particularly children. In this study, we also revealed that endocrinopathy was more common in children, especially panhypopituitarism. Childhood and adolescence are important stages in the physical development and maturation of secondary sexual characteristics, which require a variety of hormones to play a synergistic role in achieving natural physical development. Normal pituitary hormone replacement is critical to decreasing the morbidity associated with the endocrine and metabolic consequences of CP [37]. Therefore, we must pay more attention to hormone deficiency in children with CP. In addition, studies have shown that GH deficiency could increase the risk of cardiovascular disease and that sex hormone deficiency could increase the risk of osteoporosis [38,39]. Although there is still controversy about replacement therapy for endocrine hormone deficiency after brain tumor surgery, at present, replacement therapy remains the treatment option for endocrine deficiency [40]. Supplementation with hormones, including GH, thyroid hormones, and sex hormones, can improve insulin resistance, control body weight, and improve quality of life without increasing the risk of tumor recurrence [13,17,41,42]. The long-term follow-up of the patients in this study indicated that the proportion of patients who accomplished standardized and continuous endocrine replacement therapy was low; the replacement therapy of sex hormones and GH was only approximately 20%.
It remains a challenging tumor for CP to manage. Hypothalamic-pituitary axis related to tumor control, body's growth, development, metabolism, reproduction, and psychology. With increasing recognition of the impacts of both tumors and the consequences of their management, we should focus on treatments that not only improve survival but also reduce the adverse impact of hypothalamic morbidity. Regardless of treatment choice, life-long multidisciplinary care is essential to surveil for recurrence, as well as to manage and offer rehabilitation to those who are left with endocrine, metabolic, visual, neurocognitive, and psychosocial sequelae [40,43]. Replacement therapy for long-term endocrine hormone deficiency in patients with CP should be guided by physicians from neurosurgery, endocrinology, pediatrics, radiology, neurooncology, and psychology. They consisted of a specialized multidisciplinary team and made individualized replacement-treatment strategies for each patient, as this can improve the quality of life and long-term prognosis of CP patients.

Conclusions
Children and patients with hypothalamus-involved tumors are more likely to have hormone deficiency at diagnosis. Surgical resection cannot correct preoperative hormone deficiency, and preoperative hormone status is the main factor influencing postoperative endocrine function. Preservation of the pituitary stalk helps prevent pituitary hormone deficiency, but in patients with preoperative panhypopituitarism, the pituitary stalk can be sacrificed to achieve total resection. Simultaneously, multidisciplinary cooperation should be strengthened to guide replacement therapy for patients with CP and long-term postoperative endocrine hormone deficiency. A clinical algorithm for managing hormone deficiency in patients with craniopharyngioma is shown in Figure 5.