Are there any unique considerations for treating pediatric high grade glioma?

Trial availability, disease extent, age, specific molecular alterations, and family history should likely all play a role in choosing the most appropriate treatment paradigm.

It’s expected that most future studies in pediatric high grade glioma will be stratified by their nascent biology, given that it's expected that this will have an impact on targeting strategies, clinical behavior, and prognosis. 

Early results from the prior generation of studies which are well annotated with patient, treatment, and molecular information are illustrating who does well and who does poorly. The HERBY trial (PMID: 29412784, PMID: 29763623) is probably the first study to nicely annotate patients to the extent that is needed to truly understand the independent prognostic potential of a molecular signature or key molecular alteration. While the therapeutic approach is not personalized or necessarily novel, it gives us a window into how to stratify future studies. 

No one agrees on what the bins of patients for therapeutic combinations should be at present, and this will be the case for some time. For now, it’s important to recognize developing avenues of interventions, which at least have the potential to meaningfully stratify patients into relevant groups. 

The most substantially divergent schema is that of therapy for malignant gliomas in very young children (PMID: 17554787). We’ve known for some time that these patients are adequately treated with chemotherapy and surgery alone, but we are now to the point where we can potentially supplant chemotherapy with more gentle, effective, targeted agents which go after ALK, NTRK, ROS1, and MET (PMID: 32238360). The first generation of inhibitors have their own problems/side effects which we are learning about now, but it’s likely that agents targeting these key drivers in young children will become a reality in the front line setting in the coming decade (Robinson G, JCO 37, 2019 (suppl; abstr 10009)). Focal radiation is largely just used for salvage therapy following repeat resection (if possible) in these patients. 

All other groups of high grade glioma likely warrant radiotherapy in the front line setting. 

Probably the largest, well-defined subgroup of supratentorial pediatric HGG, are those with H3.3 G34R alterations. This subset is usually older, and supratentorial in location, although brainstem involvement is more common than one might expect from a supratentorial adult HGG. At present, no targeted approaches are home-runs for this group—but there are a number of trials which are addressing this population: 

-ACNS1721 Stratum 1 -NCT03581292- which is a study evaluating the use of RT+concurrent veliparib (PARP inhibitor) followed by maintenance veliparib and TMZ. The argument for this approach was that PARP inhibition could lead to increased DNA damage when combined with RT and TMZ. Unfortunately this arm of the study is now closed, but the IDHmut arm is still accruing. 

-PNOC13 - NCT03690869 - allows enrollment of these patients, but results from the HERBY study have shown that H3 G34R patients are “immune cold”. Advocates of the use of hypofractionation argue that this could increase mutational burden and result in at least a “warm” tumor that when paired with ICI could result in a meaningful improvement in EFS, but this is clearly a study question and only time will tell. Some have argued that PXA-like cases have higher immune infiltrates and would be of interest for ICI, but no data has confirmed this assertion.

PNOC008 - NCT03739372- is attempting to use a personalized treatment approach in combination with radiotherapy to improve outcomes based on known targetable mutations. A similar version of this approach was tried with DIPG (PNOC003, NCT02274987, PMID: 30861105, now closed). This trial design, while innovative and potentially informative, is controversial. Advocates tout the personalization of therapy, while critics remain frustrated with the infinitely small bins that result and how to assess patient specific performance with the infinite number of combinations that can result following review of the sequencing results. 

Probably one of the smallest subsets of pediatric high grade glioma is the IDH mutant population. Some have argued that the proportion of cases can be as much as ~30% in the adolescent/teenage population, but it’s more likely that this population represents <15% depending on the center that the data is coming from (primary pediatric vs. mixed peds/adult hospital). This population is the one clear strong performer besides the infant HGG population but surgery, chemotherapy, and radiation are all still recommended. 

-ACNS1721 Stratum 2- NCT03581292 - is addressing this population with PARP inhibition, RT, and TMZ given that IDH mutant tumors have more “BRCA-ness” given the production of the oncometabolite 2HG. Given early adult data suggesting the BRCA mutant cancers can be more susceptible to PARP inhibition (PMID: 20609467), this therapeutic strategy is exciting and could yield meaningful long term survival. 

There are a number of small subsets of pediatric high grade glioma with targetable alterations that also likely have promise: 

BRAF inhibition in a subset of cases of both LGG and HGG have shown promise.

ACNS1723-NCT03919071 is evaluating the doublet of trametinib and dabrafenib in HGG patients as a part of frontline therapy following surgery and radiation. While BRAF alterations are known to represent a subset of bad performers in LGG patients, BRAF alterations in HGG yield a subset of patients with a targetable alteration that can have long term sustained EFS (PMID: 24725538). This study is still accruing and APEC14B1 can facilitate screening for your patient if testing isn’t available at your center.

EGFR alterations have also been discovered in a subset of thalamic HGG (w/out K27M alterations) by a collaborative effort led by the UCSF group (PMID: 32303840), and early results with limited use have shown durable responses. Candidate agents for this population are plentiful as EGFR inhibition has been utilized in a variety of adult malignancies for ~15+ years. No trials utilizing this approach in the frontline setting are available, but historic studies utilizing 1st and 2nd gen agents have been completed, although this was long before we knew which groups were enriched for these alterations. 

Patients with Mismatch repair deficiencies should be considered for immune checkpoint blockade. Always take a family history... Data from Eric Bouffet et al has shown that these “immune hot” tumors are primed for consideration of ICI given the high neoantigen load. While these patients are prone to numerous malignancies (GI, colon, melanoma, lung), this is a major advance for this challenging population (PMID: 27001570). 

Patients with H3.1 or 3.3 K27M alterations should be managed as a different disease and enrolled on study whenever possible, as there is nothing but RT for this population.

While Gliomatosis is no longer a defined entity by the WHO anymore (PMID: 27157931), it still poses a treatment dilemma. Typically large field radiotherapy is the rule (WBRT) to doses of 45-54 Gy, although cone downs can be considered if it’s GC type 2 (focal contrast enhancement) after 45 Gy. 

Cases with metastatic disease should be managed with CSI +/- TMZ after max safe resection. 

Historic approaches like ACNS0423 (PMID: 27006176) are reasonable for nearly everyone else. Investigators may argue about the incremental effectiveness of 0423 vs. 0126 vs. radiation alone is small given the low completion rate of the maintenance phase of chemotherapy, and low predictive utility of MGMT status in peds relative to adults—but these are still some of the best outcomes on record with an “all-comer” unselected population. It’s still unclear whether TMZ benefits any particular subset of HGG in kids, although analysis of the HERBY data is fairly conclusive (PMID: 29763623).