PNETs comprise a heterogeneous group of malignant tumors that mainly affect the pediatric population, which are composed of neuroepithelial cells that are able to differentiate along the neuronal, astrocytic, muscular, or melanocytic lines, with high index of proliferation [8]. Supratentorial PNETs are very rare and aggressive tumors (WHO IV), and account for approximately 2.5–3 % of all pediatric tumors [9]. LGA in adults are primary (CNS) neoplasms with a variable biological behaviour. While some remain dormant for several years, others grow slowly, infiltrating surrounding normal brain and showing an intrinsic tendency for transformation into a higher grade astrocytoma [10].
Generally, adult PNETs usually appear as lobulated, purple-grayish or pinkish masses. On microscopic examination, the tumor cell population consists of undifferentiated small cells with ill-defined, scanty cytoplasm, and round or oval cells with hyperchromatic nuclei [11]. Mitotic activity is variable. Microscopic calcifications, necroses and Homer-Wright rosettes are also observed in a number of cases. This article summarize a total of 40 cases of PNET in the past decade, most pathological diagnosis is confirmed by tumor cell morphology. Pathology revealed sPNET that was characterized by dense cellularity, marked nuclear pleomorphism, nuclear molding, brisk mitotic activity, and abundant apoptotic cell death. The sPNET is diagnosed by Homer-Wright rosettes of morphological and immunohistochemical positive expression such as CD99 and GFAP, Syn, CD56 which alleviate the diagnostic rate. The Ki-67 index is an important basis of pathological diagnosis.
Nowdays, the genetic and biological features of pediatric PNETs are becoming more and more popular. There are several theories that have been postulated to describe the pathogenesis of radiogenic tumors: (a) radiation could cause genetic alterations by disrupting the deoxyribonucleic acid (DNA) of specific genes, such as PTEN or the p53 tumor suppressor gene [12, 13]; (b) karyotypic instability results from the radiation-induced multiple chromosome aberrations [14]; (c) irradiation may induce the release of oncogenic factors, such as platelet-derived and vascular endothelial growth factors [15], beta chain of fibroblast growth factor [16], and raised tissue level of basic fibroblast growth factor [17] by alterating vascular parenchymal. In a patient with radio-induced glioma reported by Gessi et al. [18], the genetic alterations were p53 mutation, in which the loss of heterozygosity of 17p and 19q, and O6-methylguanine-DNA methyl-transferase (MGMT) promoter methylation occurred, and without amplification of epidermal growth factor receptor (EGFR). An unusual somatic p53 mutation of a type specific to radiation-induced DNA damage in a secondary glioblastoma identified by Tada [12]. Also IL-1βm-RNA was seven-fold higher in the recurrent tumor of PNET and it could have triggered the fatal vasospasm [19]. Signaling pathways, sonic hedgehog (Shh), Wnt, and Notch pathways, especially the Shh pathway, are considered playing an important role in the pathogenesis and biological behavior of desmoplastic PNET [20]. Recent studies indicated inappropriate activation of Shh pathway promotes tumorigenesis and increases the risk of tumor development by ionizing radiation [21]. Mutations in genes in the Shh pathway appear to contribute to PNET development.
This case is extremely rare in the same parts of the two different tumors, PNET and LGA. Clinical features as well as pathological and molecular studies can discriminate PNET from astrocytoma, as these tumors are clearly different in terms of their origin, typical features and common characteristics. The pathogenesis of secondary sPNET is unknown. Whether they represent transformation of LGA, radiotherapy induced or de novo tumor is not clear. The possible explanations for occurrence of multiple diverse primary brain tumors in cases that contain two different tumors include secondary brain tumor induction by radiation, tumorigenetic activities of certain oncogenes, and defects in tumor suppressor genes [22]. According to the modified Cahan’s criteria for radiation-induced tumors: the onset of secondary tumors is sited in the previously irradiated field, both the primary and secondary tumors are histologically different from each other, and a sufficiently long latency period [23]. There were two cases tranlated from low grade gliomas to PNET with pathologic supports, which might be related to radiotherapy directly, but still not enough [24]. D. PAL analyzed 12 cases which were secondary pathological certification in PNET patients retrospectivly, including 10 cases directly related to radiation therapy, and 2 cases of a direct tumor level change [17]. The pathological type of this case is also a direct shift, but it still doesn’t have enough reasons to be prove the origin. Regardless, the strategy of treatment was still following the current one. After the secondary surgery, we performed radiotherapy focused on the tumor position. Now 5 months later, we did not see enough evidence of recurrence.
sPNETs are considered as a challenging clinical entity for neurosurgeons due to its poor prognosis it confers. Intrinsic biological difference of these tumors has been suggested by sPNETs patients’ worse outcome in terms of survival rates in comparison with high risk medulloblastoma patients treated by the same therapeutic strategy [9]. The overall 5-year progression-free survival rate for patients who do not undergo resection ranges from 18 to 47 % [9]. The reported overall survival of PNETs in the pediatric population varied between 29 and 57 % [11]. Follow-up ranged from 48 h to 204 months (mean 22.7 months; median 18.7 months) for this summary. 25 patients (62.5 %) were dead, 15 patients were alive. For all patients, the estimated OS rates were 63.6 % at 1 year, 21.2 % at 2 years, 12.1 % at 3 years, and 6.1 % at 5 years (Fig. 3, Additional file 1: Table S1). The mean duration of overall survival (OS) was 22 months. The survival time differences could be probably attributed to differences of the tumor sites, growth ways, proliferation differentiation degrees and the specific treatments. Maybe a larger sample-controlled trials is needed. Moreover, the prognosis for adults with supratentorial PNETs seems to be worse than that of their pediatric counterparts [25]. In adults the most important prognostic factor is considered to be the Ki-67 index. Kim et al. [4] reported that adult patients with a Ki-67 index greater than 30 % demonstrated very poor outcome, with a mean postoperative survival time of eight months [26]. Mutations in IDH1 were identified in 6 % primary sPNET [27]. More importantly, in a study by Hayden et al., it has been suggested that mutations in IDH1, which encodes for cytoplasmic isocitrate dehydrogenase, are among the most frequent mutations in sPNET in adult cases [28]. This case is that no radiotherapy one was administrated but transforming from another tumor.
It is not clear whether the occurrence of sPNET in our patient represent transformation or de novo tumors. Given the similarity to few previous cases, we postulate that they are probably radiation induced which should be considered in differential diagnosis. Molecular abnormalities, oncogenes or defective tumor-suppressor genes of medulloblastomas and astrocytomas patients have recently been investigated; however, corresponding molecular data of the present patients are unknown. The potential factors responsible for the development of multiple primary brain tumors remain unknown, further studies in elucidating the relevant factors are wanted. In addition, the more complex multifactors inducing tumorigenesis have to be clarified in order to resolve the mechanisms of multiple primary brain tumors. Future studies in large series may provide reliable data on the origin and prognostic significance of this component.