Factors of influence upon overall survival in the treatment of intracranial MPNSTs. Review of the literature and report of a case

Background Intracranial malignant peripheral nerve sheath tumors are rare entities that carry a poor prognosis. To date, there are no established therapeutic strategies for these tumors. Methods We review the present treatment modalities and present the current therapeutic dilemmas. We perform a statistical analysis to evaluate the prognostic factors for Overall Survival of these patients. Additionally, we present our experience with a 64-year-old man with a MPNST of the left cerebellopontine angle. Results To our best knowledge, forty three patients with intracranial MPNSTs, including our case, have been published in the international literature. Our analysis showed gross total resection, radiotherapy and female gender to be beneficial prognostic factors of survival in the univariate analysis. Gross total resection was recognized as the only independent predictor of prolonged Overall Survival. In our case, we performed a gross total resection followed for the first time by stereotactically guided radiotherapy. Conclusion Considering the results of the statistical analysis and the known advantages of the stereotaxy, we suggest aggressive surgery followed by stereotactically guided radiotherapy as therapy of choice.


Background
Malignant Peripheral Nerve Sheath Tumors (MPNST) usually arise de novo or from a malignant transformation of a neurofibroma. Rarely MPNSTs may arise from schwannoma, ganglioneuroma or phaeochromocytoma [1,2]. Incidence rates of MPNSTs are identified at less than 1/10 6 /year, with the majority of cases located in the brachial or lumbal plexus. Their intracranial occurrence is even more sporadic. To date, no generally accepted therapeutic strategies or prognostic factors of intracranial MPNSTs are established.
Furthermore, we present a case of a MPNST of the VIIIth nerve, and propose a novel therapeutic strategy consisting of aggressive surgical resection followed by stereotactically guided radiotherapy.

Methods
Twenty case reports and four retrospective clinical studies concerning intracranial MPNSTs were identified using the NCBI PubMed. No limitations regarding the language or time of publication were imposed on the search process. Two studies concerned MPNSTs as a whole, including tumors of the head and neck, without specifying whether the latter were extracranial or intracranial [14,15]. Thus, they were excluded from our review analysis. Similarly excluded were the cases of MPNSTs arising from extracranial trigeminal branches.
Overall survival (OS) was analyzed with the Kaplan-Meier method. Assessments of potential prognostic factors were carried out using log-rank tests. The multivariate analysis was performed using the Cox Regression Hazard Models-Backward Stepwise Procedure. P values ≤ 0.05 were regarded significant.

Results
A total of forty three patients with intracranial MPNSTs, including our case, were identified. The mean age was 37.6 ± 20.3 (3-69) years. A male predominance (30 males, 69.8%) was observed. 63.9% of the MPNSTs arised de novo; the rest derived from benign tumors. NF1 was present in 17.1% of the patients. Gross total resection (GTR) was achieved in 42.9% whereas 51.3% and 2.3% of the patients received postoperative adjuvant radiotherapy (RT) and chemotherapy, respectively (Table 1-[3 -5,7-10,12,13,16-26]). When administrated, radiotherapy was usually whole brain radiation with 60 Gy fractioned over 6 weeks. Median OS was 9 months. Progression free survival was not documented in the majority of the cases, and could not be evaluated.
In the univariate analysis, female gender (p = 0.048), GTR (p = 0.004) and RT (p = 0.010) were significant beneficial factors for OS ( Figure 1). Notably, younger age, malignant transformation of a former benign tumor and the presence of NF1 did not significantly influence outcome (p > 0.05) ( Table 2).
Some factors of potential influence upon OS, such as histological grade and tumour size, were not estimated due to the lack of reported data.
We included the significant factors above in a multivariate analysis, using the backward stepwise procedure. GTR was found to be an independent beneficial prognostic factor for OS (HR = 0.258, CI 95% 0.102-0.653, p = 0.004) ( Table 2).

Illustrative Case
A 64-year-old man presented with progressive headache, vertigo, nausea, hypogeusia and ataxia commencing 3 weeks prior to admission. A left hearing loss was known since three decades. A brain MRI approximately 10 years prior to admission revealed a small tumor localized at the left cerebellopontine angle. There were no history or clinical stigmata of Neurofibromatosis types 1 and 2.
Preoperative MRI and CT demonstrate a 3.5*4 cm measuring well delineated contrast-enhancing lesion in the left cerebellopontine angle with mass effect (Figure  2A, B). A thoracoabdominal CT as well as MRI of brachial and lumbal plexus performed ulteriorly excluded other manifestations of the MPNST.
A gross total tumor resection using neuromonitoring of the motor tract and facial nerve function was achieved. Postoperatively, a transient facial nerve palsy House-Brackmann grade III occurred as sole complication.
Histopathological examination revealed a highly cellular tumor with considerable cytologic atypia. (Figure 3). Immunohistochemical examinations revealed only focal immunoreactivity for antibodies against S-100-protein and p75. Tumors cells were strongly immunopositive for vimentin and variable immunoreative for CD99 and    Bcl-2. The tumor was classified as grade II according to FNCLCC grading system [27]. Four weeks after surgery, the patient underwent fractionated stereotactic and image guided radiotherapy using single isocentre dose delivery. A total of 60 Gy was delivered in 30 fractions. The treatment was performed using the Novalis(r) system with micro-multileaf-collimator and ExacTrac(r). The patient was immobilized using a relocatable stereotactic frame with an aquaplast mask (all components by BrainLAB(r), Germany). Because there was no detectable residual tumour on post operative MRI ( Figure 2C, D), the CTV (clinical target volume) was defined as the former tumour cavity which was delineated by fusing the pre-and post-op T1 MRI sequences with contrast enhancement. The safety margin was set to 2 mm receiving the PTV (planning target volume) of 19.026 cc, (Figure 4A, B). By using 8 non-coplanar conformal static beams the 90% isodose encompassing PTV with a conformity index of 1.52. All delivery parameters were according to the guidelines of RTOG ( Figure 4C, D, E, F).
The radiotherapy was well tolerated without acute toxicities. Clinical and MRI follow up at 12 months is without any hints of tumour recurrence.

Discussion
In contrast to their benign counterparts, neurofibromas or schwannomas, intracranial MPNSTs carry a poor prognosis with a median OS of 9 months, (range 1 to 66 months, present review). In combined series of intracranial and extracranial MPNSTs, Zou et al report a 5year survival rate of 38.7%, whereas Anghileri et al described a 5-year cause-specific mortality of 39.9%. When the influence of tumor site is considered, Anghileri reported an increased 5-year mortality of head and neck MPNSTs of 66.7%, as compared to 48.8% and 27.5% of trunk and extremities MPNSTs, respectively. The rarity of intracranial MPNSTs hampers the establishment of evidence based strategies for their optimal treatment. Thus, the management of the intracranial MPNSTs should also consider the experience gained from the treatment of extracranial MPNSTs.
Anghileri et al conducted a study of 205 patients with MPNSTs, of which 9 cases were head and neck tumors, and found that GTR, achieved in 62% of the patients, correlated significantly with longer OS, and inversely with local recurrence on multivariate analysis [14]. Zou et al carried out another study of 140 patients with MPNSTs, including 20 tumours of the head and neck, and showed that a complete surgical resection was inversely related to local recurrence on univariate analysis [15]. The results of the present review verify for intracranial MPNSTs the statistically significant influence of GTR upon OS in the univariate and multivariate analysis. Thus, a main goal in the treatment of the intracranial MPNSTs should be the complete surgical tumour resection with preservation of neurological function, whenever applicable.
The role of adjuvant radiotherapy remains controversial. Some studies suggest that radiation may be implicated in the pathogenesis of MPNSTs [8,28]. Foley et al suggested that ionizing radiation may cause chromosomal injury and induce proliferation as well as cytologic atypia in Schwann cells, resulting in radiation-induced MPNSTs [29].In our review series, 41.7% of patients harbouring a malignant transformation to MPNST received radiation in their history. Other studies haven't shown any positive effects of radiotherapy on patients outcome [30][31][32], while the recent literature indicates the beneficial role of the radiotherapy in local control of disease after a total or a near total resection of extracranial MPNSTs [14,[33][34][35][36][37][38]. Anghileri et al found adjuvant radiotherapy to be significantly related to longer OS on multivariate analysis, while no correlations with local recurrence or distant metastases were observed [14]. The radiation dosage administrated in the majority of the cases was 50 -60 Gy. Our review revealed the beneficial prognostic significance of adjuvant radiotherapy for OS in the univariate analysis. However, the multivariate analysis failed to show an independent influence of RT on OS. This could be related to the limited sample of patients. Considering the above findings and the highly malignant histological appearance of the tumour, in our patient we decided for adjuvant radiotherapy with stereotactic guidance due to its precise dosage delivery while sparing the adjacent healthy brain tissue. This strategy provides the possibility to apply an adequate high dose of 60 Gy despite of nearby sensitive risk structures like the brainstem. Thus, we were able to take advantages of both stereotactic radiotherapy and conventional fractionation while minimising the risks of RT-inducing brain injury like radiation necrosis and cognitive decline.
The optimal radiation dose has not yet been defined. We decided for a total dose of 60 Gy balancing the relatively high radiation dose to the highly malignant histological tumour appearance.
Some authors consider MPNSTs to be chemotherapyresistant [28] while others suggest that surgery followed by combined radiochemotherapy results in improved survival [39]. Two recent studies of large series of peripheral MPNSTs failed to show any benefit of chemotherapy [7,34]. Therefore, in our patient, chemotherapy was decided to be spared for the case of tumour relapse or metastatic disease.
In the present patient the MPNST seems to have resulted from the malignant transformation of a preexisting benign schwannoma. 36.1% of the review cases experience a progression of benign tumor to malignancy, having a worse OS compared to MPNSTs arising de novo. The latter difference though did not reach statistic significance (8.46 vs 22.95 months, p = 0.140). These observations point out the importance of a thorough long-time follow-up of all benign intracranial schwannomas and neurofibromas that have not been resected. However, it is not clear whether MRI followup can reliably indicate the exceptional transition of a schwannoma to a MPNST. Approximately, 25 to 50% of MPNSTs are associated with NF-1. The overall lifetime risk of genesis of MPNST in patients with NF-1 is estimated to be from 8 to 13% [14,40]. In the present review 17.1% of intracranial MPNSTs were related to NF-1.
It is noteworthy, that the female gender is less likely to present with intracranial MPNST and that females harbouring this tumour have a significant longer OS than men. Further studies are needed to enlighten the background of these observations.

Conclusion
In conclusion, we propose as therapeutic strategy for intracranial MPNST consisting of the maximal surgical resection feasible with preservation of neurological function, followed by adjuvant stereotactically guided radiotherapy. This strategy minimises the possible complications of surgery as well as of brain radiation. Chemotherapy should probably be spared for relapsed or metastasized disease. radiotherapy (in another clinic), wrote the part of the paper concerning radiotherapy and followed up the patient at his out-patient clinic. AK was the radiologist performing the preoperative and postoperative CT and MRI scans and wrote the part of the paper concerning the illustrations. PN was the pathologist who examined the tissue and wrote the part of the pathology evaluation. IW performed the ETN examination preoperatively and postoperatively, as well as performed with KG the relevant literature research. RK was the neurosurgeon who operated the patient, was the supervisor of the clinic admitted the patient, decided for the therapy procedures and revised the manuscript. All authors read and approved the final draft.