Establishment of xenografts and methods to evaluate tumor burden for the three most frequent subclasses of pediatric-type diffuse high grade gliomas.

PURPOSE

We aimed to expand and refine the experimental models for pediatric-type diffuse high grade gliomas (pHGG) and the methods to follow up disease progression in mouse pHGG xenografts.

METHODS

Using whole exome sequencing and immunoassays we characterized pHGG primary cultures and xenografts established at hospital SJD Barcelona. We obtained tumor samples and serial CSF samples from mouse xenografts. To assess tumor progression, we evaluated: (1) mouse weight, (2) human cell counts in brain paraffin sections, and (3) tumor DNA amount, quantified through droplet digital polymerase chain reaction (ddPCR) in paraffin sections and cerebrospinal fluid (CSF).

RESULTS

We established 15 experimental models of three pHGG subclasses, four of which engrafted in mice. Xenografts HSJD-DIPG-007 and HSJD-DMG-005 are diffuse midline glioma (DMG) H3 K27-altered, HSJD-GBM-002 is an H3 G34-mutant diffuse hemispheric glioma, and HSJD-GBM-001 is an H3-wildtype and IDH-wildtype pHGG. ddPCR quantification of human H3F3A K27M, H3F3A G34R, and ACVR1 R206H in paraffin samples is linear and sufficiently sensitive. We required a preamplification step to detect H3F3A K27M in CSF. In HSJD-DIPG-007 xenografts, human cell counts correlated with H3F3A amounts in paraffin for the whole engraftment period. Weight loss correlated with human cell counts and H3F3A amounts in paraffin. Serial collection of CSF was feasible, but H3F3A amounts in the CSF correlated only with weight loss.

CONCLUSION

The developed methods contribute to the preclinical field of pHGG and introduce for the first time the concept of liquid biopsy in mice, which still needs improvement regarding its use as a preclinical biomarker.