2022 Volume 69 Issue 3 Pages 313-318
The pathogenesis of gonadotropin-independent precocious puberty (PP) includes both congenital and acquired forms, the latter of which may be associated with neoplasms, such as sex-steroid hormone-producing tumors. Beta-human chorionic gonadotropin (β-hCG)-producing tumors also cause gonadotropin-independent PP by stimulating the production of testosterone in Leydig cells. Germ cell tumors and hepatoblastoma both produce β-hCG; however, there is limited evidence to show that gonadotropin-independent PP is caused by other β-hCG-producing tumors. We herein report the first case of β-hCG-producing neuroblastoma associated with the development of gonadotropin-independent PP. A 2-year-old boy presented with an increased penile length, enlargement of the testes, pigmentation of the external genitalia, and growth acceleration. Imaging, blood, and urinary examinations revealed the presence of neuroblastoma in the right adrenal region. Decreased levels of luteinizing hormone and follicle-stimulating hormone with an increased testosterone level were indicative of gonadotropin-independent PP. Since serum β-hCG was elevated, β-hCG-producing neuroblastoma was suspected. Histological findings of the resected tumor were compatible with neuroblastoma. An immunohistochemical analysis using serial sections revealed staining for β-hCG in synaptophysin-positive cells. Furthermore, immunofluorescence showed the co-staining of β-hCG with neuron-specific enolase. These results suggested that β-hCG was produced by tumor cells. Surgical removal of the tumor promptly normalized serum β-hCG and testosterone levels. In conclusion, we propose the addition of neuroblastoma to the list of differential diagnoses of gonadotropin-independent PP with β-hCG positivity in serum that includes germ cell tumors and hepatoblastoma.
PUBERTY is a biologically complex process during which individuals acquire secondary sexual characteristics and reproductive capacity [1, 2]. Precocious puberty (PP) is defined as the earlier progression of this process in both boys and girls at a chronological age earlier than the lower limit of the reference age in the general population to which they belong [1, 2], and it is divided into two types: gonadotropin-dependent and -independent PP [2]. Gonadotropin-dependent PP is caused by the premature maturation of the hypothalamus-pituitary-gonadal axis and is responsible for the majority of PP cases, particularly in girls [1, 2].
In contrast to gonadotropin-dependent PP, gonadotropin-independent PP, in which sex steroids are excessively produced independent of gonadotropin secretion, is less common and may be associated with tumors including androgen-producing adrenal tumors and Leydig cell tumors. Beta-human chorionic gonadotropin (β-hCG)-producing tumors have also been implicated in gonadotropin-independent PP due to their stimulation of testosterone production by Leydig cells [2]. Although germ cell tumors are the predominant cause of β-hCG-producing tumors, hepatoblastoma has also been shown to induce gonadotropin-independent PP by secreting β-hCG [3-5]; however, there is limited evidence to show that gonadotropin-independent PP is caused by β-hCG-producing tumors other than germ cell tumors or hepatoblastoma. We herein report the first case of gonadotropin-independent PP associated with β-hCG-producing neuroblastoma.
Written informed consent for the publication of this case was received from the patient’s parents.
ImmunohistochemistryTumor samples were fixed in 10% buffered formalin and paraffin-embedded samples were prepared. After deparaffinization and rehydration, antigen retrieval was performed using citrate buffer (pH 6.0 for anti-β-hCG and neurofilament antibodies, pH 9.0 for an anti-synaptophysin antibody) at 95°C for 60 min. Antigen retrieval was not performed for neuron-specific enolase (NSE). Endogenous peroxidase activity was quenched. After blocking, sections were incubated with an anti-synaptophysin antibody (ready-to-use, 713831, Nichirei Biosciences Inc.), anti-neurofilament antibody (1:100, M0762, DAKO), anti-NSE antibody (ready-to-use, 722081, Nichirei Biosciences Inc.), or anti-β-hCG antibody (ready-to-use, 422321, Nichirei Biosciences Inc.) at room temperature for 30 minutes. Sections were then incubated with a horseradish peroxidase-labeled polymer (414131, Histofine® Simple Stain MAX PO (MULTI), Nichirei Biosciences Inc.), followed by visualization with 3,3'-diaminobenzidine.
ImmunofluorescenceSections were prepared as described above and incubated with an anti-β-hCG antibody (ready-to-use, 422321, Nichirei Biosciences Inc.) and anti-NSE antibody (ready-to-use, 722081, Nichirei Biosciences Inc.). Antigen retrieval was performed using citrate buffer (pH 6.0). The expression of β-hCG and NSE was visualized using an Alexa Fluor® 488 (A11001, Invitrogen, Thermo Fisher Scientific) or Alexa Fluor® 555-conjugated secondary antibody (A21429, Invitrogen, Thermo Fisher Scientific), respectively. 4',6-Diamidino-2-phenylindole (340-07971, DOJINDO) was used to stain the nucleus. Images were captured by confocal microscopy (TCS SP8, Leica Microsystems).
A 2-year-old boy presented with an increased penile length, enlargement of the testes, and pigmentation of the external genitalia. He was born to non-consanguineous Japanese parents at 38 weeks of gestation without any complications during pregnancy and delivery. His birth weight and length were 3,265 g (0.6 SD) and 51.3 cm (1.5 SD), respectively. His family history was unremarkable. A physical examination revealed a height SD score (SDS) and weight SDS of +2.8 and +3.5, respectively (Fig. 1a). Height velocity SDS was +10.7. He showed Tanner stage 4 penile and scrotum development and Tanner stage 2 pubic hair. The volume of each testis was 6 mL. Radiographs of the hands and wrists revealed an accelerated bone age of 3.8 years evaluated using the TW2-RUS method standardized for Japanese children.
Growth chart and imaging findings of the patient.
a. Growth chart of the patient. Accelerated growth was observed at 1 year and six months old and resolved after resection of the tumor. b. T2-weighted image of abdominal MRI in the transverse plane showing a right adrenal mass (63 × 52 × 45 mm) as indicated by the red arrow. c. Coronal plane of I123 MIBG (meta-iodobenzylguanidine) scintigraphy showing a hyperactive area in the right adrenal region with higher intensity than that in the liver. The red arrow indicates the tumor.
A blood examination showed an increased serum testosterone level (5.55 ng/mL) associated with decreases in the levels of luteinizing hormone (LH)(<0.2 mIU/mL) and follicle-stimulating hormone (FSH)(<1.0 mIU/mL) (Table 1). Serum levels of dehydroepiandrosterone sulfate (DHEA-S) and androstenedione were not elevated, indicating that an androgen-producing adrenocortical tumor was not the underlying cause (Table 1). Serum levels of β-hCG and alpha-fetoprotein were 3.93 ng/mL (reference range: less than 0.1) and 2.8 ng/mL (reference range: 0.9–7.6), respectively (Table 1). Based on the elevated serum level of β-hCG, a β-hCG-producing tumor was suspected.
| WBC | 10,900/μL | BUN | 8.0 mg/dL | LH | <0.2 mIU/mL |
| Neut | 59.20% | Cre | 0.34 mg/dL | FSH | <1.0 mIU/mL |
| Lymph | 29.80% | LDH | 252 U/L | testosterone | 5.55 ng/mL |
| Hb | 12.5 mg/dL | ALP | 447 U/L | IGF-1 | 201 ng/mL |
| PLT | 29.3 × 104/μL | CK | 129 U/L | ACTH | 87.1 pg/mL |
| CRP | 0.42 mg/dL | T-Bil | 0.8 mg/dL | cortisol | 9 μg/dL |
| TP | 6.4 g/dL | 17-OHP | 0.98 ng/mL | ||
| ALB | 4.5 g/dL | α-FP | 2.8 ng/mL | DHEA-S | 167 ng/mL |
| AST | 26 U/L | β-hCG | 3.9 mIU/mLa | androstenedione | <0.10 mg/mL |
| ALT | 10 U/L | NSE | 25.7 ng/mLb | plasma renin activity | 2.3 ng/mL/hr |
| Na | 139 mEq/L | aldosterone | 10.0 pg/mL | ||
| K | 4.2 mEq/L | Urinary VMA/Cre | 31.16 μg/mg Crec | ||
| Cl | 107 mEq/L | Urinary HVA/Cre | 73.04 μg/mg Cred |
α-FP, alpha fetoprotein; β-hCG, beta-human chorionic gonadotropin; NSE, neuron-specific enolase; VMA, vanillylmandelic acid; HVA, 3-methoxy-4-hydroxyphenyl acetic acid; Cre, creatinine; LH, luteinizing hormone; FSH, follicle-stimulating hormone; IGF-1, insulin-like growth factor-1; ACTH, adrenocorticotropic hormone; DHEA-S, dehydroepiandrosterone sulfate
Reference range: a (less than 0.1), b (5.0–12.3), c (less than 16), d (less than 35)
An imaging analysis by computed tomography (CT) and magnetic resonance imaging (MRI) revealed a large mass at the right adrenal gland of 66 × 42 × 45 mm with internal calcifications (Fig. 1b). Multiple calcified lymph nodes were detected in the para-aortic region. An intracranial tumor was not observed on head MRI. The serum level of NSE was elevated (25.7 mg/mL, reference range: 5.0–12.3) (Table 1). Urinary tests showed increased levels of creatinine (Cre)-corrected vanillylmandelic acid and homovanillic acid of 31.16 μg/mg Cre (reference range: less than 16) and 73.04 μg/mg Cre (reference range: less than 35), respectively (Table 1). Iodine-123 MIBG (meta-iodobenzylguanidine) scintigraphy revealed the accumulation of the tracer in the right adrenal region, which corresponded to the site of the tumor detected by CT and MRI (Fig. 1c). Metastasis was not detected. Based on these results, β-hCG-producing neuroblastoma was suspected and surgical resection of the tumor was performed.
Hematoxylin & eosin staining revealed that the tumor predominantly consisted of neuroblasts with variable degrees of neuroblastic maturation that were positive for markers of neuronal cells, such as synaptophysin, neurofilament, and NSE (Fig. 2a). The percentage of mature ganglion-like cells in neoplastic cells is used to assess the level of differentiation and when it exceeds 5%, the tumor is classified as a differentiating type. In addition, the percentage of background Schwann-like stroma is important for differentiating neuroblastoma from ganglioneuroblastoma and ganglioneuroma, the former of which is comprised of less than 50% Schwann-like stroma. To clarify the characteristics of the tumor, we performed histological analyses, including an immunohistochemical analysis using S-100 antibodies that stain the myelin sheath of Schwann cells, which revealed that the tumor was compatible with neuroblastoma, differentiating type. N-MYC gene amplification, which is associated with a poor prognosis, was not detected. Immunostaining for β-hCG revealed positive staining in tumor cells (Fig. 2a). To establish whether β-hCG was produced from tumor cells, we performed immunostaining for β-hCG and synaptophysin using serial sections and observed staining for β-hCG in synaptophysin-positive cells (Fig. 2b). We also conducted an immunofluorescence analysis to further confirm that β-hCG was expressed in tumor cells and detected co-staining for β-hCG in NSE-positive cells (Fig. 3). However, staining for β-hCG was not found in all tumor cells (Fig. 2b and 3). Serum β-hCG and testosterone levels promptly normalized after surgery. Chemotherapy was initiated following surgical resection of the tumor. No signs of gonadotropin-dependent PP were identified in three years of follow-ups after surgery.
Histological findings of the patient.
a. Hematoxylin & eosin (H&E) staining (scale bar: 50 μm) and an immunohistochemical analysis of synaptophysin, neurofilament, neuron-specific enolase (NSE), and β-hCG (Scale bar: 100 μm) are shown. The arrowhead indicates tumor cells. b. An immunohistochemical analysis of synaptophysin and β-hCG was performed using serial sections (Scale bar: 50 μm). The lower panel shows a magnified image of the indicated area of the upper panel. The arrow indicates the same cell.
Co-immunofluorescence analysis of β-hCG and NSE.
Sections were incubated with anti-β-hCG and neuron-specific enolase (NSE) antibodies and their expression was visualized by an Alexa Fluor® 488 or Alexa Fluor® 555-conjugated secondary antibody. Control IgG was used as a negative control. 4',6-Diamidino-2-phenylindole (DAPI) was used to stain the nucleus.
Neoplasms cause symptoms independent of their mass effect to the neighboring tissues and/or the development of metastasis, and this includes paraneoplastic syndrome, in which neoplasms induce characteristic symptoms through the secretin of bioactive substances or the induction of the immune system [6]. Endocrine paraneoplastic syndrome is defined as a disorder in which neoplasm-related symptoms are caused by secretory factors, such as hormones, peptides, or cytokines, that are not anticipated to be secreted from the underlying neoplasm [6].
During childhood, neoplasms may contribute to the development of PP, and the underlying causes include increases in physiological gonadotropin production, excessive sex steroid production, or paraneoplastic syndrome mostly caused by the excessive production of β-hCG. In sex-steroid producing tumors, sex steroids are generally produced by tumor cells that are pathologically linked to the generation of these hormones; therefore, PP caused by sex-steroid producing tumors may be distinct from paraneoplastic syndrome. Wendt et al. previously reported the neoplastic cause of abnormal puberty using this classification and found a very low rate of paraneoplastic PP. They identified 24 out of 13,615 children and adolescents who presented with PP at the time of a tumor diagnosis, and 9 out of the 24 subjects showed paraneoplastic PP, including 5 with germ cell tumors and 4 with hepatoblastoma [3].
One of the clinical characteristics of a β-hCG-producing tumor is mild testicular enlargement despite the progressive maturation of sexual characteristics. β-hCG possesses similar biological properties to LH and binds to LH/CG receptors in Leydig cells in order to stimulate the production of testosterone; therefore, the secretion of FSH is suppressed in patients with β-hCG-producing tumors. Since the effects of FSH are also important for increases in testicular volume, the lack of an elevation in FSH levels causes a discrepancy between testicular volume and the extent of the maturation of sexual characteristics. The present case showed Tanner stage 4 penile size, indicating inconsistent maturation with the testicular volume of 6 mL. Another clinical characteristic of gonadotropin-independent PP is the potential occurrence of gonadotropin-dependent PP after tumor resection. Although the underlying mechanisms have not yet been elucidated, exposure to a high concentration of testosterone may accelerate the maturation of the hypothalamic gonadotropin-releasing hormone pulse generator, which results in the occurrence of gonadotropin-dependent PP.
Since the decreases observed in LH and FSH levels with the increase in the testosterone level were not indicative of gonadotropin-dependent PP, we performed differential diagnoses of gonadotropin-independent PP in the present case. In this process, we found that serum and urine markers for neuroblastoma and Iodine-123 MIBG scintigraphy were highly suggestive of the presence of neuroblastoma. However, positivity for β-hCG complicated our understanding of the pathogenesis of gonadotropin-independent PP because there is currently no evidence to show that neuroblastoma produces β-hCG. Therefore, we investigated the serum levels of adrenal hormones that may be associated with the development of gonadotropin-independent PP, but failed to detect elevations in the levels of adrenal androgens in this case. Although the level of DHEA-S was slightly elevated for a 2-year-old boy, this may have been due to an increase in the production of DHEA-S from the testis because 5% of circulating DHEA-S was previously shown to be derived from the testes in men [7]. Although difficulties are associated with differentiating adrenal gland- and testis-derived DHEA-S, the stimulation of LH/CG receptors by β-hCG may have caused a mild increase in DHEA-S production by the testis in the present case.
Neuroblastoma is the most common solid tumor in children, accounting for 8% of childhood cancers [8]. Although paraneoplastic manifestations, including neurological symptoms, are sometimes observed in children with neuroblastoma [9], limited information is currently available on endocrine manifestations. Gonadotropin-independent PP has not been reported as a paraneoplastic manifestation of neuroblastoma. However, we considered gonadotropin-independent PP in the present case to be caused by β-hCG-producing neuroblastoma based on the following results. An immunohistochemical analysis of the tumor showed abundant staining for β-hCG in the tumor. However, this did not necessarily indicate that β-hCG was produced by tumor cells; therefore, we performed immunohistochemistry with serial sections and a co-immunostaining analysis and confirmed that β-hCG was expressed in tumor cells. Staining for β-hCG was not detected in all tumor cells, indicating that tumor cells acquired the capacity to produce β-hCG during the process of tumorigenesis; however, the underlying mechanisms currently remain unclear [6]. Furthermore, surgical resection of the tumor promptly normalized β-hCG and testosterone levels. Based on these results, we concluded that PP in the present case was a paraneoplastic manifestation of β-hCG-producing neuroblastoma.
In summary, we herein described an unrecognized case of β-hCG-producing neuroblastoma that was associated with the development of gonadotropin-independent PP. Although extremely rare, we propose the addition of neuroblastoma to the list of differential diagnoses of gonadotropin-independent PP with β-hCG positivity in the serum that includes germ cell tumors and hepatoblastoma.
T.M. and M.K. wrote the manuscript. C.I., M.T., and M.K. performed histological analyses. All authors analyzed the data. All authors read and approved the submission of the manuscript.
The authors have no conflicts of interest to disclose.
The manuscript has not been published and is not being considered for publication in another journal.
