Intratumoural haemorrhage in intracranial germ cell tumours: A review of literature with an illustrative case

1. Introduction

Intratumoural haemorrhage may be the first sign of an intracranial tumour and occurs in 5–10% of cases.¹ When haemorrhage occurs within the tumour, it is associated with a poor outcome.² Intratumoural haemorrhage occurs commonly in high grade gliomas and oligodendrogliomas but they have been also reported to occur in pituitary adenomas, meningiomas, pilocytic astrocytomas, hemangiomas and in cerebral metastasis.3, 4, 5, 6, 7 Bleeding within the tumour occurs most likely due to tumour vascularization, neovascularization with microvascular proliferation and hyperplasia of endothelial cells along with poorly formed dilated vessels predisposing to haemorrhage within the tumour.⁸ Haemorrhage within intracranial germ cell tumours (iGCTs) are exceedingly rare, reported mostly in cases of mixed germ cell tumours and choriocarcinomas.² Intracranial germ cell tumours are extragonadal germ cell tumours usually occurring in the suprasellar and pineal regions. They commonly occur in children and adolescents and are broadly divided into germinoma and non-germinomatous germ cell tumours. In the literature, the incidence of iGCTs are more in East Asian countries, specially Korea and Japan.⁹

The prognosis of iGCTs depends on the histological type of tumours with purely germinomas having an excellent outcome whereas yolk sac tumours, embryonal carcinomas and choriocarcinomas are associated with a poor outcome.¹⁰ Intratumoural haemorrhage may cause acute onset cerebral oedema and raised intracranial pressure. Choriocarcinomas and mixed germ cell tumours are at an increased risk of haemorrhage and poor outcomes.¹¹ We report a case of a suprasellar germinoma in a child presenting with haemorrhage and the results of a systematic review of intracranial germ cell tumours with haemorrhage, their clinical features, management and reported outcomes.

2. Methods

The present study was performed according to the guidelines recommended by the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA). Publicly available studies with deidentified patient data were included in the study. We included both case reports and case series in the study along with any cohort studies if available owing to the rarity of the presentation under review.

3. Data sources and searches

A comprehensive search of the literature was carried out in PUBMED interface. Keyword searches using the search terms “intracranial germ cell tumours”, “central nervous system germ cell tumours”, “haemorrhage” and “apoplexy”, used in combination, was performed in PUBMED and further relevant studies were identified by citation search. Details of the search strategy is available in Supplement 1.

4. Study selection

Studies which reported the presence of haemorrhage in intracranial germ cell tumours or central nervous system germ cell tumours, were included in the present study. Inclusion criteria were - 1. Studies reporting spontaneous haemorrhage in intracranial germ cell tumours, Exclusion criteria - 1. Haemorrhage after surgery/biopsy in intracranial germ cell tumours, 2. Diagnosis has not been confirmed, 3. Full text of the article not available, 4. Article not in English language. Owing to the rarity of the condition under investigation, case reports and case series were included. Conference proceedings or abstracts only were not included.

5. Data extraction

Initial search results were assessed by a single investigator. Review of title and abstract were done, and studies were shortlisted for further assessment. The shortlisted articles were further reviewed by two investigators independently for inclusion as per inclusion criteria. Differences were resolved by discussion. After a study met the inclusion criteria, data extraction was done using pre-designed template. The following data were extracted - Study name and citation, year of publication, age of the patient, sex, clinical symptoms and signs, location of the tumour, histological type, type of management, mortality.

6. Definitions

Haemorrhage was considered to be present when it was described radiologically or confirmed intraoperatively or at autopsy. The type of tumour was taken as described histopathologically or diagnosed with serum/cerebrospinal fluid (CSF) markers. Mortality was considered to be present when it was reported as such.

7. Data analysis

Statistical analysis was done using SPSS software version 25. For continuous variables, mean and standard deviations were reported. For categorical variables, frequencies and proportions were reported. Owing to the small number of cases, tests for significance were not carried out.

8. Results

783 studies were identified from PUBMED search and an additional 16 studies were identified from citation search. After removal of duplicates, abstract screening of 527 studies were done to see if they meet the inclusion criteria. 26 studies met the inclusion criteria and were sought for retrieval. Out of these 26 studies, 5 studies were excluded as they were either not in English language or the haemorrhage had occurred after surgery/biopsy. 21 studies were included for final analysis. Out of these 21 studies, 18 were case reports and 3 were case series. The PRISMA flow diagram for the identification, screening and inclusion process is given in Fig. 1. The details of the included and excluded studies are available in supplement 2 and 3 respectively. Critical appraisal of the included studies were done by two authors using the Joanna Briggs Institute (JBI) critical appraisal checklist for case reports and case series (Supplement 4) .

Fig. 1.

– Prisma flowchart.

52 patients were identified from the literature review, 19 from case reports and 33 from case series (Table 1). The mean age of the patients in the case report group was 15.24 years (SD 6.73). Mean age of patients with haemorrhage was reported in 1 case series (11.1 years, SD 4.3), two other case series did not report the mean age or sex ratio. Of the patients identified from case reports, 78.9 % were males and 21.1% were females. In the series reported by Chen et al, 70.6% were males and 29.4% were females.² The symptoms reported can be grouped into two major types – symptoms due to raised intracranial pressure and symptoms due to endocrine dysfunction. Headache, vomiting and blurred vision were the most common reported symptoms suggestive of raised intracranial pressure. Among the endocrine symptoms, polydipsia and polyuria suggestive of diabetes insipidus was the most common followed by precocious puberty. Two of the reported cases had neurodeficit in the form of hemiparesis and cranial nerve palsies and two patients had history of coma or loss of consciousness.

Table 1.

Literature review of intratumoural haemorrhage in intracranial germ cell tumours.

Srl No	Author	Year	No. of patients	Age (Yrs)	Sex	Clinical features	Pathology	Location	Chemotherapy	Radiotherapy	Surgery	Mortality
Case Reports
1	Hirano T12	1976	1	8	Female	Headache, vomiting, polyuria, polydipsia, diplopia	Teratoma with choriocarcinoma	Third ventricle	No	No	No	Yes
2	Rao CV13	1979	1	8	Male	Precoccious puberty, loss of vision, polyuria, polydipsia	Choriocarcinoma	Suprasellar region	Yes	No	No	Yes
3	Kawakami Y14	1980	1	11	Male	Precoccious puberty, headache, double vision, vomiting	Choriocarcinoma	Pineal region	Yes	Yes	Yes	No
4	Fujii T15	1981	1	11	Male	Headache, vomiting	Choriocarcinoma	Pineal region	No	No	No	Yes
5	Naganuma H16	1984	1	14	Male	Headache, vomiting, hemiparesis, horner's syndrome	Endodermal sinus tumour	Pineal region	Yes	Yes	No	Yes
6	Chan HS17	1984	1	9.5	Male	Headache, retroorbital pain, vomiting, blurring of vision	Choriocarcinoma	Pineal region	No	No	Yes	Yes
7	Bjornsson J18	1986	1	10	Female	Not described	Choriocarcinoma	Third ventricle	No	No	Yes	Yes
8	Page R19	1986	2	12	Female	Loss of vision, headache, altered sensorium	Choriocarcinoma	Sellar region with extension into bilateral cavernous sinus	No	No	Yes	Yes
				17	Male	Headache, blurring of vision, episodic drowsiness	Choriocarcinoma	Pineal region	Yes	No	Yes	No
9	Graziano SL20	1987	1	15	Male	Headache, vomiting, blurred vision	Mixed germ cell tumour	Pineal region	Yes	Yes	Yes	No
10	Tsunoda S21	1993	1	29	Male	Altered sensorium, raised ICP	Embryonal carcinoma	Suprasellar region with extension into third and lateral ventricle	Yes	Yes	Yes	Yes
11	Shinoda J22	2004	1	10	Male	Polydipsia, polyuria, precoccious puberty	Choriocarcinoma	Suprasellar region with extension into midbrain	No	Yes	No	Yes
12	Kamitani H23	2006	1	21	Male	Loss of consciousness	Mixed germ cell tumour	Medial temporal lobe with brainstem extension	Yes	Yes	Yes	Yes
13	Huang MN24	2006	1	22	Male	Proximal muscle weakness, hypernatremia	Mixed germ cell tumor	Hypothalamic	No	No	Yes	Yes
14	Koh EJ25	2009	1	14	Male	Hemiparesis, cranial nerve palsies	Mixed germ cell tumour	Cerebral peduncle	Yes	Yes	Yes	No
15	Ichikawa T26	2011	1	22	Male	Headache, vomiting	Mixed germ cell tumour with areas of hemangioblastoma	Cerebellum	Yes	Yes	Yes	No
16	Kim M27	2012	1	19	Male	Diplopia, headache, nausea, vomiting	Choriocarcinoma	Pineal region	Yes	Yes	Yes	No
17	Keenan C28	2021	1	8	Female	Headache, vomiting, hemiparesis, cranial nerve palsy	Mixed germ cell tumour	Suprasellar region	Yes	Yes	No	No
18	Yang X29	2022	1	29	Male	Headache, coma	Yolk sac tumor	Suprasellar, pineal and left temporal	No	Yes	Yes	Yes
Case series
1	Chen JT2	2018	17	11.1 ± 4.3 (4–19)	70.6:29.4	Headache, vomiting, blurred vision, limb weakness, DI	Germinoma - 7, Yolk sac tumour - 1, Choriocarcinoma-1, Mixed GCT - 8	Pineal - 6, Suprasellar - 3, Pineal and suprasellar - 3, Basal ganglia - 4, Other −1	14	17	14	6
2	Li W30	2021	13	NR	NR	NR	Germinoma - 2, Mixed GCT - 11	NR	NR	NR	NR	NR
3	Liang L31	2002	3	NR	NR	NR	Mixed germ cell tumour - 3	NR	NR	NR	NR	NR

NR – not reported.

Mixed germ cell tumours were the most common histological type that presented with haemorrhage (29/52). The next most common histological type was choriocarcinoma (10/52) followed by germinoma (9/52). The other types were two cases of Yolk sac tumour and one each of endodermal sinus tumour and embryonal cell carcinoma. Pineal region was the most common location that presented with haemorrhage followed by suprasellar region. Four cases had synchronous pineal and suprasellar tumours. Type of management was reported in 36 cases, 28 out of 36 reported cases had received radiotherapy and 25 out of 36 received chemotherapy. 27 out of 36 had undergone some form of surgical treatment. Data about mortality was available in 36 patients. 18 patients out of 36 had died. From the case reports in which patient level data was available, patients were sub-grouped according to mortality (Table 2). Raised intracranial pressure was the most common symptom among patients who died. Choriocarcinoma was the most common histological type in patients who died. Pineal and suprasellar regions were the most common locations although ¾ patients with suprasellar germ cell tumours with haemorrhage died compared to 3/7 of pineal region tumours.

Table 2.

Demography, clinical symptoms and management of patients from case reports.

		Mortality
		No	Yes
Age (Years) (Mean ± SD)		15.14 (4.74)	15.29 (7.86)
Sex	Female	1	3
	Male	6	9
Symptoms	Raised ICP	4	8
	Endocrine	0	3
	Endocrine + Raised ICP	1	0
	Neurodeficit	2	0
	Not reported (1)		1
Pathology	Choriocarcinoma	3	6
	Mixed GCT	4	3
	Embryonal cell carcinoma	0	1
	Endodermal sinus tumour	0	1
	Yolk sac tumour	0	1
Location	Pineal	4	3
	Suprasellar	1	3
	Sella	0	1
	3rd Ventricle	0	2
	Synchronous Suprasellar and pineal	0	1
	Other	2	2
Management	Radiotherapy	6	5
	Chemotherapy	7	4
	Surgery	6	7

9. Illustrative case

An 11 year old girl presented with history of painless, progressive loss of vision in her right eye for 3 days along with intermittent episodes of headache for 1 week. There was no history of vomiting, convulsion, increased thirst or urination, unexplained weight gain or loss, any loss of smell or any behavioural changes. She had normal developmental history and was attending school regularly. On examination, she had only perception of light in her right eye with relative afferent pupillary defect, left sided vision was intact. There were no other cranial nerve palsies. Her motor, sensory and higher mental functions were intact. Laboratory investigations showed presence of secondary hypothyroidism. Her serum sodium, cortisol and prolactin levels were normal. Non-contrast computed tomography (NCCT) scan of head showed presence of a hyperdense space occupying lesion (SOL) in the suprasellar region without any hydrocephalus. Magnetic resonance Imaging (MRI) with contrast showed an avidly contrast enhancing SOL in the suprasellar region with blooming in SWI images compressing the right optic nerve (Fig. 2). She was taken up for urgent surgery. Right pterional craniotomy was done and trans-sylvian approach taken. A reddish tumour was present in the suprasellar region, soft in consistency with blood clots present in the tumour cavity. The tumour was decompressed through the interoptic, optico-carotid and carotid-occulomotor triangles followed by gross total excision. The pituitary stalk was identified and preserved along with the basal perforators arising from the internal carotid artery. She developed transient diabetes insipidus in the post-operative period which was managed medically. Histopathological examination of the excised tumour showed the presence of germinoma. She had improvement in her vision and on Visual Evoked potential (VEP) in the post-operative period. Post-operative serum alpha fetoprotein (AFP) and lactate dehydrogenase (LDH) were within normal limits, however, serum beta human chorionic gonadotropin (HCG) was mildly elevated. She was discharged to home with advice to follow up for chemo-radiation.

Fig. 2.

A- Axial section of non-contrast CT head showing a hyperdense SOL in suprasellar region.

B – Axial section of T1 weighted MRI showing an isodense SOL in suprasellar region.

c – Axial section of T2 weighted MRI showing areas of hypointesities within the SOL in suprasellar region

D – Axial section of post-contrast T1 weighted images showing intense contrast enhancement in the SOL.

E − Sagittal section of T2 weighted image showing an SOL with areas of hypointensity in the suprasellar region. Pituitary gland can be seen separately.

F – Sagittal section of post-contrast T1 weighted image showing heterogenous contrast uptake within the SOL.

G – SWI sequence showing blooming within the SOL suggestive of haemorrhage.

H – Post-operative non-contrast CT of head. Ommaya reservoir in situ.

10. Discussion

Intracranial germ cell tumours (iGCTs) encompass 0.3–0.5% of all primary intracranial neoplasms and approximately 3% of primary malignant paediatric brain tumours according to Western literature.^32,33 However, series from Japan and Taiwan indicate that these tumours are far more common in Asia, where iGCTs make up 2–5% of all primary intracranial neoplasms and account for up to 15% of primary paediatric intracranial neoplasms.^32,33 Ninety percent of patients with iGCTs present with symptoms before the age of 20 years. Sixty-five percent of tumours occur in the second decade of life (11–20 years). The peak incidence is around 10–12 years of age. Intracranial germinomas appear mainly in the pineal and suprasellar regions. Less common locations are the basal ganglia, ventricles, thalamus, medulla oblongata or the cerebral hemispheres.³²

Intratumoural haemorrhage is a rare presentation of iGCTs. A case series of 17 paediatric patients of iGCTs over a 14 year period was reported by Chen et al²They had attempted to identify the prognostic factors of iGCTs with haemorrhage by examining the clinical and radiological findings of these patients. They found that 6 of 17 patients had died (35.3%) and 52.9% had a poor outcome according to the modified Rankin scale. Among the prognostic factors, they identified haematoma volume and haematoma/tumour volume ratio to be significant predictors of poor outcome. However, only haematoma/tumour volume ratio was a significant predictor of mortality. The mean haematoma/tumour volume ratio was 15.7 ± 16.1% in the survivors and 46.0 ± 31.5% in the mortality group. The underlying pathologic type of the tumour (germinoma vs non-germinomatous germ cell tumours) and the location of the tumour didn't have a significant effect either on the outcome or on the mortality. The hematoma/tumour volume ratio is an interesting metric and warrants further exploration as a prognostic indicator in intratumoural haemorrhage. Unfortunately, we were not able to analyse its utility as a prognostic indicator as it was not reported in most of the studies that met the inclusion criteria for the review.

Similar effect of tumoural haemorrhage on survival was reported by Shinoda et al.²² Patients with tumoural haemorrhage had a significantly worse survival compared to patients who did not had haemorrhage. Apart from tumoural haemorrhage, they also found less than subtotal resection, pineal region location and female sex to negatively affect the survival whereas patients who received chemotherapy and radiotherapy had a significantly better outcome. This negative effect of intratumoural haemorrhage on survival was however contradicted by Jiang et al, who had conducted a literature review to identify the prognostic factors associated with survival in intracranial choriocarcinomas.³⁴ They had identified 51 cases of primary intracranial choriocarcinoma meeting their inclusion criteria. On multivariate analysis, they found only the combination of chemotherapy along with surgery to be a significant predictor of survival. Presence of intratumoural haemorrhage didn't affect the survival.

In a retrospective review of intracranial germinomas and mixed germ cell tumours, Li et al had found that the incidence of intratumoural haemorrhage was significantly more in mixed germ cell tumours than in germinomas (68.75% vs 11.76%).³⁰ Liang et al also reported presence of intratumoural haemorrhage only in mixed germ cell tumours.³¹ However, Chen et al noted presence of intracranial haemorrhage in 7 cases of germinomas and in 10 cases of non-germinomatous germ cell tumours.²

Clinical features suggestive of raised intracranial pressure was the most common clinical feature in our review followed by endocrine symptoms and focal neurodeficit. This is similar to the findings reported by Chen et al who reported headache, vomiting and blurred vision to be the most common symptoms in their cohort of patients.² In the present review, we found that pineal region tumours followed by suprasellar tumours most commonly bled however there were some unusual locations reported like hypothalamus, cerebellum and cerebral peduncles. This is similar to the cohort reported by Chen et al where pineal region was also the most common location followed by suprasellar region.²

The mechanism of intratumoural haemorrhage is speculative and not known exactly. Tumours contain two main types of vessels, neoangiogenic vessels formed by angiogenic factors and co-opted vessels which are pre-existing vessels co-opted by the growing tumour.³⁵ Both these types of tumour vasculature are immature and fenestrated, lacking in tight junctions and blood–brain barrier.³⁶ These factors predispose to spontaneous vessel rupture and tumour bleed secondary to tumour necrosis. Vascular endothelial growth factor (VEGF) and matrix-metalloproteinases (MMPs) have been shown to be involved in tumoural haemorrhage associated with metastatic brain tumours.³⁷

There are several drawbacks of the present study. Publication bias is a significant drawback, all of the cases of iGCTs with haemorrhage may not have been reported in the literature. Two of the case series did not report mortality, demographic or management data. Only English language studies were included in the review. Due to the relatively small number of reported cases, statistical analysis was not performed. It would be interesting to note if there is any association between location, histology, demographic factors, type of management and the risk of mortality.

11. Conclusion

Intratumoural haemorrhage in iGCTs is a rare but intriguing presentation of iGCTs. Depending on the haematoma volume and the acuity of presentation, patients may require urgent management to achieve a favourable outcome. Literature suggests that intratumoural haemorrhage, particularly large volume haematomas, increases the risk of mortality. Early detection and prompt management should be instituted for this presentation. However, the optimum mode of treatment of this presentation of iGCTs have not been elucidated in the literature.

CRediT authorship contribution statement

Debajyoti Datta: Writing – review & editing, Writing – original draft, Validation, Software, Methodology, Formal analysis, Data curation, Conceptualization. Partha Ghosh: Supervision, Conceptualization. Sutirtha Hazra: Validation, Methodology. Soutrik Das: Validation. Debajyoti Pathak: Writing – review & editing, Writing – original draft, Supervision, Methodology, Formal analysis, Data curation, Conceptualization.

Declaration of competing 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.

Abbreviations

AFP –

Alpha feto protein

CSF–

Cerebrospinal fluid

HCG –

Human chorionic gonadotropin

iGCT –

Intracranial germ cell tumour

JBI–

Joanna Briggs Institute

LDH –

Lactate dehydrogenase

MMP –

Matrix metalloproteinases

MRI –

Magnetic resonance imaging

NCCT

Non-contrast computed tomography

PRISMA

Preferred Reporting Items for Systematic Review and Meta-Analysis

SD –

Standard deviation

SOL-

space occupying lesion

SWI –

Susceptibility weighted imaging

VEGF –

Vascular endothelial growth factor

Appendix A. Supplementary data

The following is/are the supplementary data to this article.