**3.1. DIPG and tissue donation**

identify molecular subtypes increased, biopsy has been increasingly performed to not only identify the type of tumor present but to delineate potential molecular targets that could be therapeutically explored. In experienced hands, the permanent morbidity after stereotactic

In histologic diagnosis, DIPGs are defined as a fibrillary astrocytoma, WHO grades II–IV, and in most of the cases resemble malignant gliomas in other locations [28]. However, the prognosis for DIPGs is not associated with the histological classification. DIPGs harboring the histone 3 mutation classified as WHO II and III have a poor OS, similar to WHO IV patients [29]. In addition, significant histopathological variability has been reported in DIPGs, where a single biopsy may not be representative of the histological classification of the entire tumor [23,29,30].

Important biological information obtained from biopsies may be used in future clinical trials, guiding new treatment regimens and allowing for advances in surgical and molecular analytical techniques [31]. The use of tissue obtained from pretreatment biopsies combined with antibodies to detect driver mutations gives the opportunity to identify the genomic mutational landscape of DIPG and provides opportunities to improve diagnosis, prognosis,

**3. New advances: the future of genomics, epigenomics, and proteomics**

Taking into consideration that DIPG may represent a biologically distinct subclass of glioma, there is a great need for the comprehensive investigation of tumor biology. Therefore, studies in this rare type of cancer cannot be performed without the knowledge of genomics and proteomics. The development of new technologies that can rapidly analyze DNA, RNA, and proteins and the progress in bioinformatics area are substantial advances that have largely been achieved in the past years. Analysis of mRNA, methylation, and proteomic profiling of DIPGs compared to healthy brain tissue identified two distinct subgroups characterized by upregulation of N-Myc and Hedgehog signaling pathways [15]. Combinatory analysis of whole-genome and whole-exome sequencing, copy number alterations, methylation, and gene expression profiling revealed three molecular subgroups in DIPG, highlighting novel therapeutic targets [18]. The three molecular subgroups consisted of upregulation of N-Myc (histone 3 wild-type DIPGs), silent genomes with fewer copy number alterations, and histone 3 K27M mutant DIPGs with *ACVR1* and *TP53* mutations. DIPGs of silent and H3-K27M molecular subtypes would benefit from therapies targeting altered histone modifications, while patients of the N-Myc subtype would benefit from therapy targeting N-Myc or ID2. Furthermore, DIPGs of the N-Myc and silent subgroups lacked amplification of receptor tyrosine kinases, indicating the inefficacy of inhibitors targeting these kinase pathways [18]. Therefore, numerous combinatory analyses of DIPG have identified the importance of the

biopsy has been found to be less than 5%.

406 Neurooncology - Newer Developments

and better understanding of the potential drug targets.

synergistic genetic and epigenetic basis of this fatal childhood cancer.

A primary requirement for genomic analysis of cancer is tumor material. Much of the histological and prognostic information that we have about DIPG is from biopsies that were frequently performed until the early 1970s, before any of the current genomic techniques were available, and when CT/MRI were not widely accessible. After this period, the frequency of biopsies significantly decreased and histological information from pretreatment samples has not been available. Over the past 40 years, most DIPG patients participated in clinical trials without prior genomic profiling of their tumors. Therefore, the reason why these treatments failed is not clear.

Over the years, the lack of tissue samples and biological information on DIPG caused many research groups to explore other ways to collect tissue samples. Among these, autopsy procurement of brain samples began to have a great meaning in the understanding of DIPG. Programs for postmortem specimen donations from research groups throughout the world in a variety of tumors had positive results. The contribution of autopsy tissue donation in DIPG is relatively new and yields promise to facilitate genome-wide studies in this disease.

A variety of research teams have been working in the recent years with postmortem tissue collection and several important publications show a number of potential targets for new treatments [18,32–35]. These studies also revealed that DIPG cannot be considered a single entity, and according to the underlying biology of the tumor, different types of treatment may be needed. Findings from preclinical drug testing conducted on accurate *in vitro* and *in vivo* models of DIPG will provide direction to future clinical trials.

### **3.2. Preclinical models**

*In vitro* and *in vivo* models of pontine gliomas are helping to guide the understanding of DIPG and key genomic changes that help maintain the tumor's growth and resistance to therapy. Different approaches have been used to generate primary neurosphere cultures and allog‐ raft and xenograft mouse models to elucidate the biology of DIPG; however, they are unlikely to provide all the answers. Allograft models mimicking brainstem gliomas have been used to unravel expression signatures and to serve as a platform to test the efficacy of novel thera‐ peutic agents, such as small molecule multi-kinase inhibitors [36,37]. Primary neurosphere cultures and xenograft models from DIPG tissue obtained at autopsy have provided remark‐ able advances in understanding tumor biology. Some of these include the identification of a cell of origin, methods of effective drug delivery, and identification of potential therapeutic targets [4,38–41]. A pitfall in these models is the exposure of autopsied tissue to chemothera‐ peutic agents. Therefore, research groups are increasingly focusing on deriving preclinical models from biopsied tissue [42]. Biopsy-derived preclinical models have been utilized for identification of genomic expression profiles and for testing potential therapeutic agents [40,43,44].

Effectively treating cancer in mouse models may not always yield similar results in humans. On the other hand, animal models can represent an alternative for screening of novel agents and combination of drugs, leading to the discovery of the most promising drugs for human DIPG trials. A recent study identified a FDA-approved epigenetic drug, Panobinostat, to be a leading therapeutic candidate by testing a plethora of promising anticancer drugs in biopsyand autopsy-derived preclinical models of DIPG [44].

A better knowledge of the genomic aberrations that are considered drivers in DIPG is essential to treat accurate animal models. Research is improving and several studies are focusing on the discovery of these important mutations and agreeing that novel combinations should be tested in genetically and histologically accurate preclinical models prior to their translation to the clinic [5,42]. This collaborative effort will elucidate many of the unanswered questions in DIPGs.
