Children with DIPG are typically diagnosed at the 5th to 10th year of their life, with tumors being more frequently located in the pons rather than the midbrain or medulla oblongata [14-16]. Most of the remaining 80% of tumours are diffuse brainstem gliomas (DBG) [11]; usually a fibrillary (World Health Organization (WHO) grade 2) or malignant (WHO grade 3 or 4) astrocytoma with typically an aggressive disease course and grim prognosis [3,17-19]. DBGs are the most aggressive subgroup of brainstem tumours, and typically occur in the pons (part of the brainstem) [20]. MRI constitutes the imaging modality of choice for DIPG with unique imaging characteristics [21,22]. Although the imaging criteria used to define a classical or 'typical' DIPG (tDIPG) vary to some extent, and there are inconsistencies in the interpretation of the images [23], the general consensus radiographic features of tDIPG include a T1-hypointense and T2-hyperintense tumor involving at least 50% of the axial pontine diameter [24-26]. Typical DIPG was defined radiographically as a poorly defined tumor with mass effect occupying ≥75% of the axial pontine diameter that was hypointense on T1-weighted MR images and hyperintense on T2-weighted images [27]. DIPGs may typically demonstrate mild heterogeneous enhancement or no enhancement, nevertheless, increased enhancement or ring enhancement may be suggestive of poorer prognosis [28-30]. In many centers in the United States, biopsy is reserved for patients with a clinical diagnosis of 'atypical' DIPG (aDIPG), i.e., pontine tumors in which the above imaging features are absent or incomplete [31]. In the three clinical cases presented, the classical MRI criteria meet in the first and third clinical case (Figure 1, Figure 13), but the second accounts for hypointensity on T1 FLAIR and T2 FLAIR images and hyperintensity on T2 FSE images (Figure 4). This atypical MR image and the subsequent clinical disease development is the basis for such pontinе tumors without the possibility of biopsy, to be extremely careful. Because of the vitally important brainstem functions, the brain tumors localized there are extremely difficult to treat, especially with the application of the radiation method, which can cause a sudden stopping of respiratory and cardiac activity. Surgery of DIPG to relieve hydrocephalus, is not indicated, because the typical appearances on MRI are characteristic and reduction in tumor size only adds to the child’s worsened overall condition [32]. There is currently no indication for image guided stereotactic brain biopsy in children with a short history and typical MR appearances, because it currently will not alter management strategy [33]. During the past three decades, numerous clinical trials have been conducted in newly diagnosed DIPG patients including RT with or without chemotherapy, targeted agents and immunological approaches. None of these trials have demonstrated a clear improvement in median overall or progression-free survival (PFS) [34].

Radiotherapy remains the gold-standard treatment for diffuse intrinsic pontine gliomas (DIPG); by contrast, chemotherapy has not shown any benefit [35]. Several studies have concluded that conventional external-beam RT up to TD 54 Gy in 30 fractions in 6 weeks was the treatment mainstay [36-42] with median time to progression and overall survival of 6 and 9 months [8,41,43]. In the clinical cases presented with the MP image corresponding to classic DIPG, we conducted a conventional fractionated IMRT with DD 1.8 Gy up to TD 50.4-54 Gy (Figure 2, Figure 3, Figure 14, Figure 15). As a general principle, the treatment volume of the radiation field should encompass all the site(s) of disease with a defined margin to allow for non-imageable tumour spread into adjacent brain (1 cm for low grade and 2 cm for high grade tumours) [44]. To increase the overall and disease-free survival, other dose fractionation modes such as hyperfractionated and hypofractionated RT was applied [45,46]. Given the limited life expectancy of patients with DIPG, hypofractionated RT regimens have been considered [35,47-51]. Hypofractionated RT has potential radiobiological advantages over standard conventional fractionated therapy in newly diagnosed glioblastoma multiforme [52]. Some institutions have employed hypofractionated RT to decrease the length of therapy in an effort to reduce the patient's burden, with clinically similar levels of disease control reported [35,51,53-55]. Hypofractionated RT is well tolerated with the advantage of decreasing the treatment burden on children and their families [45], with nearly comparable results to conventional fractionation, though not fulfilling the non-inferiority assumption [55,56]. In hypofractionated schemes a total dose of 45 Gy is delivered in daily fractions of 3 to 5.5 Gy over three weeks [35]. Due to the shorter healing period of the hypophractionated RT, we considered to realize it after gradually rising daily doses due to critical neurological structures localized on the brainstem (Figure 5, Figure 6). Despite the poor prognosis of DIPG with limited survival durations, re-irradiation may be considered. With advances in treatment techniques, it is feasible to re-irradiate select patients with progressive disease [15]. Re-irradiation (Re-RT) can safely be delivered for progressive DIPG. Utility analysis suggests that a schemes of 24 Gy in 12 fractions are preferred [57]. 25 patients were enrolled, and 11 out of 16 patients with local relapse received Re-RT to a dose of 19.8 Gy delivered over 11 days [58]. The majority of patients with DIPG, responding to upfront RT, do benefit of Re-RT with acceptable tolerability [59]. Based on the attempt of the above-mentioned authors, in the presented clinical case with tumor progression after 8 months of the hypofractional RT (Figure 7), we decided to carry out a re-irradiation in brainstem target volume with a 0.2 cm insurance zone with DD 2 Gy up tu TD 20 Gy (Figure 8, Figure 9). Despite our concerns about the tolerability of this re-irradiation, the child experienced it very well without acute side neurological toxicity. After two months, in the pontine tumor with an atypical MRI image, we observe the development of liquid metastases on the lateral cerebral ventricle walls, despite the hypofractionated RT, followed by re-irradiation (Figure 10). Such a find is rarely observed and most likely corresponds to glioblastoma in childhood, which is very rare. Another histological possibility is that it is a pontine meduloblastoma, in which if a biopsy was performed, primary RT should include the entire craniospinal axis. The conclusion of this clinical case is that the atypical MRI image of the pontine tumors requires a stereotaxic biopsy or spectroscopic MRI.