High-grade gliomas in adolescents and young adults reveal histomolecular differences vis-à-vis their adult and pediatric counterparts

See article by Roux et al. in this issue, pp. 1190–1202.

Malignancies in adolescents and young adults (AYA) pose a challenge for pediatric oncologists and, less frequently, for adult oncologists when it comes to patients’ correct clinical management and access to the best possible treatments all over the world.¹ Since the beginning of this millennium, renewed efforts to deal with these issues have been made by lay and medical organizations.² After sarcomas, brain tumors are the most common solid tumors in this patient population between childhood and adulthood.³ More than any other malignancies, brain tumors demand a focused approach in AYA because of their morbidity, complexity, and prognosis.

High-grade gliomas (HGGs) are a paradigmatic example of such complexity. They can occur at any time of life, from infancy to old age. They are devastating tumors, and the vast majority of patients affected succumb to them within 5 years. Conventional treatments, such as surgery, radiation, and chemotherapy, are certainly important in their management, but are rarely curative by themselves.

Recent integrated molecular analyses have shown that pediatric HGGs are histologically similar, but biologically distinct from their adult counterparts. Subgroups of cases feature recurrent mutations but different ages of incidence, anatomic locations, and clinical outcomes. Characteristically, isocitrate dehydrogenase 1 and 2 (IDH1/2) mutations with or without 1p/19q codeletions identify diffuse hemispheric glioma in adults, while mutations of the histone genes (H3.3, H3.1) are predominant in pediatric HGG, and diffuse midline gliomas with the histone mutation H3K27 define a type of tumor listed in the World Health Organization 2016 classification of central nervous system tumors.⁴ In the midst of this molecular divide between pediatric and adult HGGs there are those occurring in AYA.

In the present issue of Neuro-Oncology, the paper by Roux et al devoted to HGGs in AYA marks a substantial advance in this age group’s diagnosis and care.⁵ An important point regarding this study is that although it is retrospective, the authors conducted their analyses as a frontline protocol. This means that after patients’ samples and records were collected by major tertiary neuro-oncology institutions, the data were cleaned up front, under the supervision of 3 experienced neuropathologists who conducted a centralized histological and radiological review. This strict protocol led to the exclusion of more than 20% of cases misdiagnosed as HGGs, which left an important collection of 80 cases available for the study.

An integrated diagnosis, according to the WHO 2016 classification, was obtained in 75% of cases using “routine techniques,” ie, specific antibodies and standard sequencing procedures. Whole exome sequencing (WES) confirmed the diagnosis in all cases, refining it by detecting rare mutations in 6 tumors H3/IDH wild-type. Interestingly, DNA methylation profiling proved useful in only one case (1.2%), in which it identified an HGNET-MN1 (high-grade neuroepithelial tumor–meningioma 1 gene) that would have gone undetected using standard techniques.

These results seem to reinforce the notion that morphological aspects and immunophenotyping with specific antibodies as molecular surrogates,⁶ associated—if necessary—with standard molecular techniques such as targeted/next-generation sequencing, suffice to enable an adequate histomolecular diagnosis in a large majority of HGGs. More complex molecular analyses (not currently available at many institutions), such as WES and DNA methylation profiling, seem to be useful in only a minority of particularly complex cases.

As for the distribution of the molecular subgroups in AYA (Figure 1), the study shows a prevalence of the “pediatric type” histone-mutant tumors (40%) over the “adult-type” IDH-mutant types (27.5%), but with important distinctions with respect to the 2 reference age groups.

Fig. 1.

Adolescent high-grade glioma subtypes and their distribution.

The frequency of H3K27 mutant gliomas was 26.3% in AYA with a topographical distribution different from that observed in children, ie, a marked prevalence of non-brainstem diffuse midline glioma (NB-DMG) over classical diffuse intrinsic pontine glioma (DIPG) associated with a statistically relevant difference in mean age (16 y for DIPG, and 21 y for NB-DMG). The occurrence of 2 cases of H3K27-mutant diffuse non-midline glioma emphasizes the importance of routinely screening for these mutations in all HGGs regardless of their location.

The H3G34-mutant HGGs (13.7%) emerge as the most typical hemispheric supratentorial malignant gliomas in AYA patients, with a mean age of 18 years (range, 15–25) and histological features ranging from astrocytoma-like to embryonal-like.

Two important elements came to light in the second most frequent molecular subgroup of “adult -type” IDH-mutant tumors (27.5%): (i) non-canonical IDH mutations are relatively common, which means they are not detectable by immunohistochemistry using the IDH1-32H antibody, so sequencing is necessary in such cases; and (ii) oligodendrogliomas with 1p/19q codeletions are very rare (2.5%).

The study also confirms the limited prognostic value of grading (WHO grade III vs grade IV)—as in pediatric HGGs⁷—and underscores the role of age, performance status, extent of resection, and integrated molecular classification as independent determinants of final outcome. The authors are also to be congratulated for their multidisciplinary/integrated statistical analysis, which confirms the need for neurosurgeons and neuroradiology referral centers for HGG in AYA to: (i) excise as much tumor as possible without damaging the patient and (ii) correctly assess tumor extent before and after surgery. This is already a rule in the currently running SIOP protocols for children and adolescents with medulloblastoma and ependymoma.^8,9

In conclusion, Roux et al clearly show that in AYA with HGGs: (i) case assignment and stratification for treatment should follow clinical and molecular integrated diagnoses; (ii) older age should not prevent inclusion in pediatric trials, given the proportion of histone gene mutations typical of pediatric age in this age group too, and (iii) the only way to satisfy the previous two conditions is to make inclusion in clinical protocols (wherever available) mandatory with a view to improving the prognosis for AYA with HGG.