Comparison of different predominant lesions in intracranial bifocal germ cell tumors to predict neuroendocrine outcomes

Introduction

Intracranial germ cell tumors (iGCTs) are rare malignancies of the central nervous system that probably arise from primordial germ cells (1, 2, 3, 4) or transform from endogenous brain cells (5, 6), which usually develop along the midline of the brain as well as in the bifocal regions. Bifocal lesions are considered the special image feature of iGCT (7), commonly involving the sellar and pineal regions, rarely as the sellar and basal ganglia lesions (8). Neuroendocrine disorders may probably be due to the tumor mass effect in sellar area as well as to the related treatment. Neuroendocrine disorders affect the initial presentations, the safety during the period of neurosurgery and chemoradiotherapy, and complications and sequelae. However, it is unclear what factors affect or predict neuroendocrine outcomes.

The tumor mass effect manifests as a crucial factor affecting endocrinopathies in iGCT patients. Tumors located in the sellar area may involve adjacent neuroendocrine organs that affect the regulation of hormone synthesis and secretion, as well as in the diffuse neuroendocrine system. Pre-treatment image, to some extent, reflects tumor mass effect and the involvement with neuroendocrine organs. However, for those bifocal patients, it has not been discussed whether the different predominant lesion possesses different neuroendocrine characteristics. It is unclear whether the sellar-predominant bifocal iGCT patients are similar to sellar solitary ones, while non-sellar-predominant bifocal iGCT patients are not. Furthermore, with the advantage of the neurosurgery and chemoradiotherapy techniques, iGCTs are still considered curable diseases with a 5-year event-free survival of more than 90% in localized germinomas and more than 70% in non-germinomatous germ cell tumors (9). It is also unclear whether identifying predominant lesions in bifocal iGCT can predict neuroendocrine outcomes and guide long-term neuroendocrine management of patients.

Previous studies have described the images in the sellar region of patients with bifocal lesions as five categories and six types (10). However, the role of this complex image classification on neuroendocrine outcomes is unclear. The distinction of the predominant lesion in bifocal iGCT patients has a wide applicability. We considered that a giant pineal mass or a basal ganglia mass with pituitary stalk involvement was a typical non-sellar-predominant type. On the contrary, iGCTs with hypothalamus and/or pituitary involvement with a limited lesion in the pineal gland or the basal ganglia mass were considered to be the predominant type. To explore the characteristics and neuroendocrine outcomes of bifocal GCTs with different predominant lesions, this article was performed. Further comparison between different subtypes of sellar-predominant bifocal iGCT patients would also be performed.

Subject and methods

This study included 38 patients who were diagnosed with iGCTs between January 2015 and December 2022 at the Beijing Tiantan Hospital. The inclusion criteria were: (i) the defined diagnosis based on human chorionic gonadotropin (hCG) and alpha-fetoprotein (AFP) in serum and cerebrospinal fluid (CSF), histological features and patterns, and the response to chemotherapy or radiotherapy; (ii) the pre-treatment brain images indicating a sellar lesion; and (iii) the completion of the multidisciplinary treatment for iGCT. The exclusion criteria were (i) a history of chromosomal abnormalities or other malignant tumors; (ii) loss of physical examinations due to mental disorders or severe physical disabilities; and (iii) incomplete revaluation of neuroendocrine function.

Medical records were compiled, including general characteristics, clinical manifestations, neuroimaging features, laboratory examination, tumor markers, diagnostic methods, therapeutic strategies, and outcomes. The symptom interval was defined as the time span from its occurrence to its diagnosis. Tumor markers were measured using an electrochemiluminescence immunoassay. Hormone levels were measured by an immunochemiluminometric assay. Common biochemical parameters were measured using an autoanalyzer. Pre-treatment-enhanced magnetic resonance imaging (MRI) was divided into sellar predominance and non-sellar predominance. Sellar predominance was divided into two subtypes: subtype 1 means tumors located in the intrasellar site or with a pituitary stalk involvement (Fig. 1A) and subtype 2 means tumors located in the third ventricle or with an extension to the pituitary stalk or intrasellar site (Fig. 1B). Non-sellar predominance means the isolated lesion was located on the floor of the third ventricle or pituitary stalk (Fig. 1C). Therapeutic strategies for iGCTs were platinum-based induction chemotherapy (ifosfamide, 1.5 g/m², d1–3; etoposide, 70 mg/m²; d1–3; and cisplatin, 30 mg/m²; d1–3) every 4 weeks for two cycles following radiotherapy (whole-brain for metastasis-free lesions or craniospinal irradiation for metastasis lesions) and then an additional two cycles of chemotherapy. Alternative neurosurgery included a ventriculoperitoneal shunt (VPS) operation for increased pressure hydrocephalus, transcranial or transsphenoidal biopsy for diagnosis, and transcranial or transsphenoidal surgery for removing residual after chemoradiotherapy.

View Full Size

Different imaging grades for patients with bifocal intracranial germ cell tumors. Each lesion was marked by white dotted lines: (A) the neuroimaging features of sellar-predominant subtype 1, (B) neuroimaging features of sellar-predominant subtype 2, and (C) neuroimaging features of non-sellar-predominant patients.

Citation: Endocrine Connections 12, 8; 10.1530/EC-23-0168

• Download Figure
• Download figure as PowerPoint slide

Neuroendocrine disorders were evaluated before treatment, after tumor elimination, and at the last follow-up visit. Central diabetes insipidus (CDI): presentation of polyuria (>50 mL/kg d or 3.5 L/day) with elevated plasma osmolality (300 mOsm/kg H₂O) and decreased urine osmotic pressure (<600 mOsmol/L) mOsm/kg H₂O) and negative urine glucose. Impairment to hypothalamic–pituitary–adrenal axis (HPA): basal cortisol level <3 μg/dL indicated impaired HPA function while the basal cortisol level >15 μg/dL excluded it. If basal cortisol levels ranged from 3 to 15 μg/dL, the circadian rhythm of adrenocorticotropic hormone and cortisol (08:00, 16:00, and 24:00 h) was monitored. If the circadian rhythm of the adrenocorticotropic hormone and cortisol disappeared, the level of the adrenocorticotropic hormone was low or normal, indicating that HPA was potentially damaged. Impairment to hypothalamic–pituitary–thyroid axis (HPT): the basal free thyroxine level was below the reference values (7.64–16.3 pmol/L), while the thyroid-stimulating hormone level was lower, normal, or slightly higher than the reference values (0.49–4.91uIU/mL). Impairment to growth hormone–insulin-like growth factor-1 (GH/IGF1): children with growth retardation or adults with obvious hypometabolic syndrome had IGF1 levels below 2 s.d. than their peers and had additional pituitary hormone deficiencies. Impairment to hypothalamic–pituitary-–gonad axis (HPG) axis: adolescents (males aged 14–18 years, females aged 13–18 years) had follicle-stimulating hormone, luteinizing hormone levels, and sex steroid hormones below the Tanner stage 2 reference values or bone age was below the actual age of 2 years. Adults (over 18 years) were below the adult reference values. Hyperprolactinemia (HPRL): the basal prolactin levels were greater than 17 ng/mL in males or 20 ng/mL in females. Slightly elevated basal prolactin levels were repeatedly evaluated on another day.

The results of the continuation variable were presented by mean and standard deviation (mean ± s.d.) and median and range (interquartile range). The independent sample t-test and the Wilcoxon rank sum test were performed. Qualitative variables were presented by number and percentage (n, %). Fisher’s exact test was performed. A P-value less than 0.05 was considered statistically significant. SPSS 22.0 software (IBM) was used for statistical analysis.

This study protocol was reviewed and approved by IRB of Beijing Tiantan Hospital, Capital Medical University, approval number (KY2022-024-01). Informed consent was obtained from all participants (or their parents/legal guardian/next of kin) to participate in the study.

Results

Discussion

This study focuses on the potential relationship between image features and neuroendocrine outcomes of bifocal iGCT patients. Based on the characteristics of patients with different predominant lesions, this study aimed at optimizing neuroendocrine management during the period of multidisciplinary therapy and the following survival time. The comparison of the neuroendocrine outcomes between patients with different predominant lesions before multidisciplinary therapy, after multidisciplinary therapy, and at the last follow-up visit indicated that (i) bifocal iGCT patients initially had neuroendocrine disorders regardless of the predominant lesion but (ii) non-sellar bifocal patients may have better neuroendocrine outcomes after multidisciplinary therapy. These two results remain after excluding the three patients who underwent surgical resection. The differences in neuroendocrine outcomes between patients with different predominant lesions seem less related to the multidisciplinary strategy.

Neuroendocrine disorders are commonly seen in iGCT patients with sellar involvement, including those non-sellar-predominant patients. In this study, the ratio of sellar-predominant bifocal patients and non-sellar-predominant ones is nearly 5:4, indicating that the evaluation of neuroendocrine function and relevant hormone replacement for non-sellar-predominant bifocal patients deserves more concern. Before multidisciplinary therapy for iGCT, patients with both the sellar-predominant and non-sellar-predominant lesion presented with hypothalamic adenohypophysis and neurohypophysis disorders. Both sellar predominant and non-sellar predominant manifested as polydipsia and polyuria, nausea, and fatigue, without statistical significance. Differences in HPA impairment, HPT impairment, GH/IGF1 impairment, HPG impairment, HPRL, and CDI were not significant between different predominant groups. According to the similarity of initial neuroendocrine function, the multidisciplinary therapy team, including the oncology, radiology, and neurosurgery departments, should pay more attention to the evaluation of the neuroendocrine function of those non-sellar-predominant patients. The proper exogenous hormone replacement treatments are needed for all iGCT patients with sellar involvement to ensure the safety of tumor treatment, regardless of the predominant lesion.

Neuroendocrine function varied during multidisciplinary therapy. After treatment and during follow-up, partial adenohypophysis disorders might spontaneously relieve in the non-sellar-predominant patients, leaving constant neurohypophysis disorders. During survival time, sellar-predominant patients had a higher incidence of HPA impairment, HPT impairment, GH/IGF1 impairment, HPG impairment, HPRL, and CDI than those of non-sellar-predominant patients. The differences in HPA impairment, HPT impairment, and HPG impairment were significant. These differences suggested that sellar-predominant patients tend to have similar neuroendocrine outcomes to isolated sellar lesions, while the non-sellar-predominant patients do not. Based on the differences in long-term neuroendocrine outcomes, the multidisciplinary therapy team should consider constant exogenous hormone replacement treatment for sellar-predominant lesions, and reduced hormone replacement dose for non-sellar-predominant patients. To explore therapeutic influence, the comparison of neuroendocrine outcomes between the sellar-predominant group and the non-sellar-predominant group was repeated excluding the three patients who received surgery resection. The results also indicated that non-sellar-predominant patients have a greater potential for spontaneous improvement. In fact, iGCTs generally do not need to be removed by neurosurgery, which reduces surgery-related endocrinopathy (9). In addition, the high sensitivity of iGCTs to radiation makes it possible for the preservation of the neuroendocrine function during the therapy for tumors (11, 12, 13, 14). However, previous studies which discussed original therapeutic strategies in bifocal iGCT rarely focused on neuroendocrine outcomes.

The distinction of the predominant lesion in patients with bifocal iGCT plays a valuable role in predicting neuroendocrine outcomes during the survival time. However, further comparison between different subtypes of sellar-predominant patients did not indicate a significant difference. It is suggested that the current image subtype is insufficient to predict long-term neuroendocrine outcomes. In fact, previous studies have discussed the image classification of sellar GCT; however, the image classification of sellar iGCT has not achieved a consensus. A previous study involving 38 patients with solitary sellar GCT divided the image into three categories mainly based on the location of sellar turcica and the third ventricle (15). Tumors located on the floor of the third ventricle to the third ventricle were classified as type 1, those located in the pituitary stalk were classified as type 2, and those located in the sellar turcica were classified as type 3. Type 2 means easy to delayed diagnosis or misdiagnose due to the indistinct pituitary stalk and type 3 means difficult to differentiate from craniopharyngioma in the early stage. This study described the typical morphology of the sellar GCT with obvious limitations as follows. Firstly, type 1 and 3 images of this study are not the commonly seen iGCT images. Lastly, this study did not compare the influence of tumor mass on endocrine glands and hormone levels. However, another study involving 22 solitary sellar iGCT and 20 bifocal iGCT patients classified the morphology of GCT in the sellar into five categories (10), based mainly on the relationship between the direction of tumor growth and the floor of the third ventricle and the turcica sellar. Type 1: a localized lesion at the floor of the third ventricle, type 1a without pituitary stalk involvement while 1b with. Type 2: a localized lesion at the floor of the third ventricle with growing toward the third ventricle and pituitary stalk. Type 3: a lesion located at the floor of the third ventricle with growing toward the sellar turcica. Type 4: a lesion located at the floor of the third ventricle, limited to the upper end of the pituitary stalk. Type 5: an isolated pituitary stalk lesion. This study suggested that sellar lesions located between the floor of the third ventricular and the upper part of the pituitary stalk were the most common image classification, while isolated pituitary stalk lesions are rare. Besides, this study indicated that the prognosis of bifocal iGCTs is worse than that of solitary iGCTs. This study described more systematical morphological features of the sellar GCTs but also had some limitations. First, the image classification of type 6 is relatively complex and subjective. For example, both type 1a and type 4 images were located on the floor of the third ventricle with an identifiable pituitary stalk. The difference depends only on approximate tumor mass rather than a precision volume Secondly, the study proposed the controversial point that multidisciplinary treatment for bifocal tumors should refer to metastasis iGCTs. However, most research on iGCTs still considers that the dose and field of radiotherapy and relevant chemotherapy depend on tumor metastasis or not. Finally, this study did not compare the neuroendocrine outcomes among patients with different image type.

We focused on the structure of the hypothalamus, pituitary stalk, and pituitary gland to describe the potential damage of tumors on neuroendocrine organs rather than directly comparing lesions’ volume. The subtypes in this study were mainly based on the extension of tumor growth, dividing into the involvement of the pituitary gland area (subtype 1) or both the pituitary gland and the hypothalamus area (subtype 2). This type scheme would be more rationale for revealing the damage of neuroendocrine organ damage. In the present study, although the differences in neuroendocrine outcomes are not significant between subtypes 1 and 2, it could be predicted that patients with the subtype 1 and 2 images would have better neuroendocrine outcomes than those with non-sellar patients. Combined with the previous study on pituitary stalk GCT (16), this study indicated that a limited sellar iGCT located in the pituitary and the floor of the third ventricle, including solitary and bifocal lesions, has the potential to spontaneously improve neuroendocrine outcomes. Hence, it is valuable evidence to support the role of sellar lesion image in predicting neuroendocrine outcomes and in administrating individual hormone replacement treatment. In the future, if more research concerns on the image features, a new image classification might be discussed. Sellar iGCT image classification may be comparable to the image classification of craniopharyngioma, involving tumor growth, extent, and the relationship to surrounding brain structure. However, craniopharyngioma is a type of tumor that is primarily dependent on neurosurgery, while iGCT is dependent on radiotherapy and chemotherapy. The potential role of image classification might be different between these two sellar diseases.

IGCTs have a high overall survival rate regardless of the controversy about synchronous (17) or metastatic occurrence (10, 18, 19, 20) of bifocal lesions. Hence, the command of improving the quality of life of those survivors deserves more attention. This study, primarily based on hypothalamic neurohypophysis and adenohypophysis, indicates that identifying predominant lesions in bifocal iGCT can predict the neuroendocrine outcomes of patients, and guide their long-term neuroendocrine management to some extent. In future prospective studies, neuroendocrine-related disorders need to be further explored, not only concentrating on pituitary target hormone levels, but also on hypothalamic function, optic and oculomotor nerve function, cavernous sinus and other factors affecting quality of life.

This study has certain limitations. First, this study only included 32 patients who may not reflect some potential risk factors for neuroendocrine function. Regression analysis of multiple related factors prefers if future patients are sufficient. Next, although irrelevant to the predominant lesion in bifocal iGCTs, the new subtype of the sellar iGCT image should be considered to predict the neuroendocrine outcomes of iGCT patients with sellar involvement. Finally, as a retrospective design, this study did not mention the final height, puberty development, and the quality of life. All these long-term outcomes are valuable for survivors and require future prospective research to explore.

Conclusion

Patients with bifocal iGCTs with sellar involvement initially present with neuroendocrine disorders regardless of the predominant lesion. However, non-sellar-predominant patients would have better neuroendocrine outcomes than those sellar-predominant ones after tumor treatment. The distinction of the predominant lesion in patients with bifocal iGCTs plays a valuable role in predicting neuroendocrine outcomes, as well as in optimizing long-term neuroendocrine management during survival time.

Supplementary materials

This is linked to the online version of the paper at https://doi.org/10.1530/EC-23-0168.

Declaration of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Funding

This study did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.