Purpose: Giant cystic craniopharyngioma extending into the posterior cranial fossa (PFGCCP) is rare. The purpose of this paper is to investigate the clinical characteristics of this tumor and the efficacy of interstitial radiotherapy.
METHODS: A retrospective analysis of 20 cases of PFGCCP admitted to the Naval General Hospital and treated with radioisotope (P-32) intratumoral radiotherapy between 2000 and 2011 was conducted to summarize their clinical characteristics, follow-up treatment results, and a comprehensive review of relevant literature reports.
RESULTS: M/F = 11/9, age 13 months to 14 years (mean 86.7 months). There were 4 cases of physical growth retardation; 16 cases of intracranial hypertension symptoms; 15 cases of optic nerve damage and 12 cases of other cranial nerve damage; and 3 cases of polydipsia and polyuria. The tumor volume was 65-215 ml, with a mean of 137.3 ml. 20 patients underwent a total of 21 surgical operations to treat 27 cysts. Follow-up ranged from 7-138 months, with an average of 48.6 months. The tumors completely disappeared in 12 cases (14 tumors) and decreased in size by 75% in 8 cases (13 tumors). 1 patient died due to tumor progression. 7 de novo cysts appeared in 5 patients. 8-66 months follow-up after the same dose of interstitial radiotherapy. 4 tumors completely disappeared, 2 were significantly reduced, and 1 progressed. 24 eyes with impaired vision completely recovered 4, improved 12, unchanged 4, and aggravated 2. Other Of the 15 cranial nerves damaged 13 were completely recovered, 1 partially recovered, and 1 unchanged. No serious complications related to surgical manipulation and interstitial radiotherapy were found.
CONCLUSION: PFGCCP is a rare type of tumor, mostly seen in children, without gender differences. Increased intracranial pressure and cranial nerve damage are the main clinical manifestations. Despite the large size of the tumor, radioisotope intrastromal radiotherapy can achieve relatively satisfactory tumor control rates and clinical outcomes and quality of survival with a low complication rate.
Patients and data
From January 2000 to December 2013, a total of 782 patients with craniopharyngioma were admitted to the Naval General Hospital, including 26 cases of PFGCCP (3.32%).The diagnostic criteria for PFGCCP were that the tumor originated in the saddle area and extended into the posterior cranial fossa with a diameter greater than 150 px or a volume greater than 60 ml, and ectopic craniopharyngioma originating in the posterior cranial fossa or the pontocerebellar horn were not included in this group ( Figure 1).
In this group of patients, 14 were male and 12 were female, and their age at diagnosis ranged from 3 months to 14 years (mean 7.2 years). The medical history ranged from 0.5 to 12 months (mean 5.8 months). The manifestations were physical growth retardation in 6 cases; intracranial hypertension symptoms (headache, nausea, vomiting, drowsiness, etc.) in 21 cases; vision loss in 22 cases, hearing loss in 8 cases (2 bilateral and 6 unilateral), abducens nerve palsy in 5 cases (4 bilateral and 1 unilateral), unilateral facial nerve palsy in 3 cases (complete palsy in 1 case and incomplete palsy in 2 cases), and incomplete unilateral articulatory nerve palsy in 3 cases. There were 5 cases of polydipsia and polyuria; 2 cases of seizures. There were 7 cases of combined hydrocephalus. Previous surgical history: craniotomy in 4 cases and ventriculoperitoneal shunt in 3 cases.
All patients underwent head CT and MRI examinations. The results showed that the tumors originated in the saddle area and extended to the unilateral or bilateral posterior cranial fossa or pontocerebellar horn area, and some patients had cysts up to the level of occipital foramen or even over the occipital foramen into the spinal canal, including one case reaching the level of cervical 2 vertebrae. The tumors were purely cystic in 19 cases and mixed cystic and solid in 7 cases. CT manifestation: the cystic part showed simple hypodensity in 8 cases, simple high density in 5 cases, simple isodensity in 2 cases, mixed in 5 cases (isodensity + high density in 4 cases, high density + low density in 1 case); with calcification in 19 cases: the cystic wall showed typical eggshell-like calcification in 12 cases, and intra-tumor calcification in 7 cases. MRI manifestations: 9 cases with short T1 and long T2, 6 cases with long T1 and long T2, 1 case with equal T1 and long T2, and 4 cases with mixed signal: all showed short T1 and long T2 + long T1 and long T2. The tumor volume was 65-215 ml, with an average of 129.5 ml.
The diagnosis was based on medical history, clinical manifestations, typical imaging manifestations, intraoperative extraction of cystic fluid containing cholesterol crystals and pathological diagnosis from previous surgeries.
Methods
All patients in this group underwent stereotactic cyst fluid aspiration (drainage) followed by radioisotope P-32 intra-tumoral interstitial radiotherapy using framed or frameless stereotactic techniques with local anesthesia in most patients and general anesthesia in children under 5 years of age, head mounted Leksell-G head frame or four Mark dots affixed to the head, and enhanced MRI T1-weighted image axial and coronal 3 mm layer thickness scans. The scanned images are transmitted to the workstation of the stereotactic surgery planning system through the network to calculate the tumor volume, design the coordinates of the target site, the surgical path and the puncture entry point. The puncture path should avoid the cerebral sulcus, ventricles and any nerves and blood vessels. If the tumor volume is <100 ml< span="">, direct cyst puncture is chosen and 1/3 to 1/2 of the tumor volume is aspirated and radioisotope P-32 is injected; if the tumor volume is >100 ml, a microfine drainage tube is implanted in the tumor for continuous or intermittent drainage of cystic fluid, and after 3-5 days, the isotope is injected and the drainage tube is removed. The activity of injected radioisotope P-32 colloid was calculated according to the prescribed dose of 150 Gy for uniform distribution within the tumor and the cyst wall, and in practice the table of volumetric activity calculated by the following formula was consulted [7]: activity = 0.1365*(Dose in Gy)*vol(ml)/0.455. For mixed cystic solid tumors, solid tumors in the saddle area were treated simultaneously with Gamma knife treatment was performed with a prescribed dose of 14-16 Gy and a central dose of 28-32 Gy. Repeat treatment was performed on nine cysts with large tumor volumes, at intervals of 2.5 to 6 months from the first treatment, with a mean of 4.7 months. The calculated irradiated dose to the cyst wall was the same as the first treatment.
Results
A total of 27 surgical operations were performed in 26 patients, including 14 framed stereotactic operations and 13 unframed operations. A total of 33 cysts were treated: 21 for single cysts and 6 for 2 cysts (in one patient 4 cysts were punctured in two separate procedures, 2 cysts per procedure, with an interval of 3 months between the two procedures).
The extracted cyst fluid was light yellow, dark yellow, yellowish brown, dark brown and dark black. 24 cysts were drained by tube placement for 2-5 days with an average of 3.3 days.
All patients had relief of intracranial hypertension after surgery, and 7 cases with hydrocephalus showed normal ventricular size on postoperative imaging.
Follow-up: time 8-144 months, mean 52.5 months.
Survival: 1 patient died of tumor progression during the follow-up.
Tumor control rate: Of the 26 patients (33 cysts) treated initially, tumors disappeared completely in 16 cases (19 tumors) and tumors decreased in size by 75% in 10 cases (14 tumors). The time to tumor disappearance ranged from 6 months to 24 months after surgery, with a mean of 11.2 months. (Figure 2).
Figure 1: Male, 9 years old. He was diagnosed with craniopharyngioma by MRI for 6 months due to bilateral vision loss (A), and the tumor shrank after 1 year of stereotactic P-32 intra-tumoral interstitial radiotherapy (B), and the tumor was stable after 21 months (C), and the tumor recurred in the saddle area after 4 years with multiple cysts and protruded into the third ventricle (D), and 3 months after another interstitial radiotherapy, the suprasellar and third ventricle tumors were stable but the pontine anterior 6 months after the third interstitial radiotherapy, the tumor in the third ventricle disappeared, but the new tumor in the anterior pontine brain increased significantly and extended downward to the suboccipital foramen (F). 2 years after the fourth interstitial radiotherapy, the tumor extending to the posterior cranial fossa disappeared completely, but the tumor in the saddle area progressed significantly and was multicystic (G, H, I), and the patient gave up treatment and died 1 year later because of tumor progression.
Figure 2: Female, 8 years old, with intermittent headache and vomiting with bilateral hearing loss for 4 months; MRI scan showed a giant craniopharyngioma in the saddle area extending into the posterior cranial fossa into the cervical spinal canal (A, B); 1 year after stereotactic P-32 intra-tumor interstitial radiotherapy, the tumor largely disappeared (C); 1.5 years after treatment, a new cystic tumor grew in the right paracranial saddle (D, E, F), and again After 1.5 years of treatment, the tumor disappeared, but there was another new cystic tumor at the pontocerebellar horn (G, H), and a third interstitial radiotherapy was performed; 5, 6, and 9 years after the first treatment, CT scans showed that the tumor disappeared, and only calcifications remained in the saddle area (I, J, K).
De novo tumors: 7 de novo cysts with a volume of 3-11 ml (mean 6.4 ml) were seen in 5 patients during follow-up, including 2 de novo cysts in 2 patients (3 months, 3 years and 7 months, 3 years after the first treatment, respectively) and 1 de novo cyst in 3 patients (6, 9 and 15 months after the first treatment, respectively). The time of appearance ranged from 3 months to 3 years from the first treatment, with a mean of 13.8 months. These neoplastic cysts also received the same dose of intra-tumoral P-32 interstitial radiotherapy with a follow-up of 8-66 months, with 4 complete tumor disappearances, 2 significant shrinkages and 1 progression, for a control rate of 85.7% of neoplastic tumors.
Recovery of cranial nerves: 22 cases of decreased vision, 8 cases of complete recovery, 9 cases of improvement, 5 cases of no change; 8 cases of hearing loss all recovered completely; 5 cases of adductor nerve palsy, 4 cases recovered completely, 1 case of no change; 2 of 3 cases of facial nerve palsy recovered completely, 1 case of no change. 3 cases (1 case of complete palsy, 2 cases of incomplete), 2 of 2 cases of unilateral motoneural nerve palsy recovered completely, 1 case of no change. Polydipsia and polyuria in 5 cases; seizures in 2 cases.
Endocrine and others: Among the 6 patients with delayed physical development, 3 recovered and gradually became similar to their peers, 2 improved, and 1 had no change. 1 of the 5 cases of polyhydramnios recovered completely, 2 improved, and 2 had no change. 2 cases with seizures did not have further seizures while taking antiepileptic drugs.
Complications: Gamma radiography was performed on the first postoperative day in 6 patients and no isotope leakage was detected. No serious complications related to surgical manipulation and interstitial radiotherapy (e.g., internal carotid artery, suboptic thalamus injury, etc.) were also found in this group of patients.
Discussion
The present report is a summary of the largest number of cases of PFGCCP that we are aware of. Only one group of 7 cases has been reported in all previous literature reports [8], while the others are case reports of 1-3 cases [9-18]. This confirms the rare and low incidence of this tumor. Large groups of cases show that “giant cystic craniopharyngiomas” account for about 4-9% of all craniopharyngiomas, while giant cystic craniopharyngiomas that extend towards the posterior cranial fossa account for only about 4% and consist of multiple sacs [13,19]. Giant craniopharyngiomas in which the main body of the tumor is located suprasellarly and extends to the level of the foramen magnum and upper cervical spine are even rarer, with only a few cases reported in the literature [9,10,12,13,15,16]. Most authors consider tumors more than 4 cm or 5 cm in diameter as giant craniopharyngiomas. This type of craniopharyngioma accounted for 3.32% of all craniopharyngioma patients admitted to our group during the same period. From the literature and our clinical data, we found that almost all PFGCCP occurred in pediatric patients, with no significant gender differences.
The clinical manifestations of PFGCCP are closely related to the biological behavior of the tumor. Unlike the usual clinical manifestations of craniopharyngioma, the clinical features of PFGCCP are, firstly, an increase in intracranial pressure due to the large size of the tumor, and secondly, the extension of the cystic portion of the tumor along the brain pool and subarachnoid space may cause functional deficits of the affected cranial nerves. The most susceptible cranial nerve is still the optic nerve, other cranial nerves include the auditory nerve, facial nerve, abducens nerve and oculogyric nerve, from the anatomical point of view, during the expansion of the tumor into the posterior cranial fossa, the costochondral nerve and trigeminal nerve are also susceptible to be involved. This may be due to the buffering effect of the long-traveling nerve in the skull on the distending compression, and the tough cerebellar curtain that limits the compression of the nerve by the cystic tumor. The trigeminal nerve, on the other hand, is more tolerant to compression or the compression symptoms are easily ignored. Third, the degree of compression of the pituitary and inferior optic thalamus by the tumor determines whether there is clinical manifestation of pituitary function or inferior optic thalamus hypofunction. Fourth, due to the presence of a large natural subarachnoid space in the middle and posterior cranial fossa in children and the children’s own greater tolerance and compensability to increased intracranial pressure, the early stage of the lesion often has no typical symptoms and is easily overlooked.
The imaging features of PFGCCP are clearly bounded cystic or cystic solid occupational manifestations, with the suprasellar portion connected to the posterior cranial fossa portion, showing eggshell-like calcification of the cystic wall and irregular calcification of the solid portion on CT. The non-cystic portion on MRI mostly shows mild to moderate enhancement of the T1-weighted image (like the cystic wall), and the cystic component often shows T1 low signal T2 high signal. If there is hemorrhage or high concentration of cholinesterol in the cystic fluid, it shows high signal in both T1 and T2-weighted images. Among the 24 cysts recorded in our group, MRI showed 13 cysts with short T1, 10 cysts with long T1, and only 1 cyst with equal T1, all of which showed long T2 signal on T2-weighted images. It is not difficult to make a diagnosis based on the typical imaging presentation.
The treatment of giant craniopharyngiomas with a diameter of more than 4 CM remains a challenge for neurosurgery [21], and there are many controversies regarding the choice of treatment. This is especially true for PFGCCP.
Given the benign pathological nature of the tumor, total surgical resection should be the most desirable option and has been recently attempted by some authors with better recent results.PFGCCP involves more important intracranial structures than the usual saddle craniopharyngioma, although in most cases there is a clear boundary between the tumor capsule wall and normal brain tissue and it is necessary to achieve complete separation of the tumor from the surrounding tissue by pulling the capsule wall during resection. Although in most cases, there is a clear boundary between the tumor capsule wall and normal brain tissue, and complete separation of the tumor from the surrounding tissues can be achieved only by stretching the capsule wall during resection, there is definitely a risk of stretching the structures such as neurovascular. Therefore, total tumor resection has a high risk of mortality and disability [21-24]. Moreover, total tumor resection does not exclude the possibility of local tumor recurrence, even if no tumor remains on postoperative MRI [25-27]. The quality of life of some patients in long-term survivors after total resection is also unsatisfactory [28-32]. Therefore, some authors also recommend subtotal or partial resection with or without postoperative adjuvant combination therapy (including conventional radiotherapy or interstitial radiotherapy), especially in those with preoperative imaging confirming tumor invasion of hypothalamic structures [33-35].
There is also more controversy regarding the development of the surgical plan and the choice of the surgical approach, and some authors have reported the completion of one-stage subtotal sarcoid resection of suprasellar and posterior cranial fossa tumors via the Kawase approach [8,11,12,16,17]. However, some authors [13] advocated a staged procedure, with the suprasellar portion being removed first via a rock or bifrontal craniotomy, followed by a suboccipital approach a few weeks later to remove the posterior cranial fossa portion of the tumor, and the second procedure should be performed before the formation of postoperative adhesions.Gangemi et al [18] reported that an endoscopic cyst fistula should be performed first to drain the tumor, and then a total craniotomy should be performed after the tumor has reduced in size. It is believed that endoscopic fistula first can rapidly reduce intracranial hypertension and improve cerebrospinal fluid circulation, and the adhesions with the surrounding important structures are reduced after the tumor volume is reduced, which facilitates the next step of total resection. The selection of the surgical approach requires consideration of the extent of exposure to the lesion, whether there are important anatomical structures that affect the surgical operation and the experience of the surgeon, as well as the patient’s preoperative status and functional deficits (visual acuity, hormones, and suboptic thalamus).
Table 3 summarizes the results of surgical treatment and follow-up of PFGCCP reported in the literature in recent years. Among all 22 patients, a total of 35 surgical resections were undergone, with a mean of 1.59 resections/case. 16 cases had total resection of the tumor, 2 cases had subtotal resection, and 4 cases had partial resection. 2 cases died at 4 days and 12 days postoperatively, respectively, and 1 case underwent 2 surgical total resections of the tumor and died after 3 years. The number of cases followed up for more than 3 months was 13, and although none of them had tumor progression, 3 patients required hormone replacement therapy, and the overall treatment outcome was not satisfactory.
Table 3 Surgical and follow-up results of PFGCCP reported in the literature in recent years
Authors (year) Number of cases Number of operations Extent of resection Adjuvant therapy Follow-up time Prognosis
Baba (1978) [9] 1 4 times Partial resection — — —
Young (1987) [10] 2 2 partial resections — —
Shimada (1989) [11] 1 1 total resection — —
Sarigolu (1996) [12] 1 1 total resection — 3 months No progress
Connolly (1997) [13] 3 2, 4, 3 total excisions Hormone replacement (2) 47-77 months Return to school
Sener (1997) [14] 1 2 total resections — 12 days Death
Buhl (2000) [15] 2 1 sub-total resection — 12 months No progression
2 total resections Hormone replacement 12 months No progression
Goyal (2002) [16] 1 1 total resection External radiotherapy — —
Kiran (2008) [17] 2 2 total resections — 4 days Sudden death
3 months No progression
Gangemi (2009) [18] 1 2 total resections — 1 month Pituitary hypofunction
(endoscopy + craniotomy) Urogynosis
Zhou (2009) [8] 7 8 total resections (5) Gamma knife (2) 3-145 months 1 death
Sub-total resection (1) (mean 68.6 months)
Partial resection (1)
It has been well documented that P-32 intra-tumoral mesenchymal radiation therapy can treat cystic craniopharyngioma with satisfactory immediate and long-term results [7,36-40], but its optimal indication is for simple cystic tumors of 3-40 ML volume [41], and we have not found any reports about isotopic intra-tumoral radiation therapy for giant cystic craniopharyngioma. Among the 26 patients treated in our group, the smallest tumor volume was 65 ml and the largest was 215 ml. Among the 33 cysts treated, 19 cysts disappeared and 14 cysts were significantly reduced, showing ideal results. This indicates that the cyst wall tumor cells of giant cystic craniopharyngioma in children are highly sensitive to interstitial radiotherapy. However, this treatment did not completely prevent tumor recurrence and de novo tumor growth, and a total of 6 patients in this group developed de novo tumors during the follow-up period, and 1 of them died of de novo tumor.
Previously, the cumulative irradiation dose to the cyst wall reported in the literature was 50-1000 GY, and most of them reported 200-300 GY [36-41], and some authors considered that the efficacy of treatment doses below 200 GY were unsatisfactory and the recurrence rate was high. We chose a dose of 150 GY for the tumor wall, considering the reduction in tumor size after drainage of the cystic fluid, and we found no clear complications related to radiation therapy after treatment and subsequent follow-up, but it is unknown whether some of the nerve damage that did not fully recover was related to radiation therapy.
The most obvious clinical outcome after surgical treatment is the relief of intracranial hypertension, and this relief should be sustained unless the tumor recurs. Whether the symptoms of cranial nerve damage can be recovered or the degree of recovery depends largely on the reversibility of cranial nerve damage due to compression. Fortunately, the compression of cranial nerves is mostly slow-onset, thus the cranial nerves develop a certain tolerance to the tumor compression, and the function of cranial nerves can be partially or fully recovered when the compression is released. In this group, 25 cranial nerves were damaged except for the optic nerve, and all of them were completely restored except for one facial nerve, ophthalmic nerve and abducens nerve, which were not restored during the follow-up period. However, in the three cases of PFGCCP in children with unilateral or bilateral hearing loss reported by Connolly, 47-77 months (mean 5 years) of follow-up after staged surgical total excision of the tumor showed no recurrence of the tumor but no recovery of hearing.
Giant cystic tumors tend to grow in the direction of low pressure, with relatively less compression of the inferior optic thalamus and pituitary structures, and therefore patients have less endocrine disorders than usual craniopharyngioma, which is favorable for patient prognosis.
We believe that in deciding on treatment options for PFGCCP, the surgeon must balance the relationship between tumor resection or control, patient survival, and quality of life. Different approaches and technical means (microsurgery, endoscopy, radiosurgery, fractionated external radiotherapy, intra-tumoral isotope intra- or interstitial radiotherapy or chemotherapy, etc.) can be considered as long as they provide the best benefit and the best quality of life for the patient.
In conclusion: PFGCCP is a rare type of tumor that occurs in almost all patients in children, without gender differences. MRI shows a well-defined cystic or multicystic tumor with a long T2-weighted image and T1-weighted signal depending on the contents. The T1-weighted signal depends on the contents. The prognosis is better than that of craniopharyngioma in the saddle area if the treatment plan is properly chosen.